JP7111354B2 - Search system and search method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a search system and search method using a database, and is particularly suitable for searching relationships between multiple physical property parameters.

The goal of prediction and design in materials research is to identify materials with desired properties. A method that has been widely used for this purpose from the past is a method that aims to identify materials that have the desired properties from a condition-property chart. This involves observing changes in characteristics when changing only one specific condition out of multiple conditions, creating a chart, and finding the conditions with the desired characteristics by interpolating or extrapolating the chart. , is a method of identifying matching materials. The term "chart" used here has the same meaning as the term "graph" representing a line graph or the like, but a different term is used for the purpose of distinguishing it from the later-described "graph" consisting of nodes and edges.

At this time, the change in characteristics when changing only one specific condition out of a plurality of conditions is often obtained by conducting an experiment on one's own. This is because it is difficult to obtain a large amount of data with all the same conditions other than the above-mentioned specific conditions, even after searching a large number of documents.

Patent Literature 1 discloses a search system capable of objectively searching for information on constituent substances of new materials having desired properties. The search system disclosed in the document includes a database having a plurality of pieces of physical property parameter information for each of a plurality (a large number) of substances. At this time, the database may include physical property parameters for which actual data are not given depending on the substance. Using a physical property parameter to be searched as one axis and a part of other physical property parameters as another axis, a two-dimensional or three-dimensional space or more is created, and each substance in the database is mapped. At this time, physical parameters for which there is no actual data are supplemented with virtual data predicted using multivariate analysis, calculation based on a predetermined logical formula, or first-principles calculation. In a search map obtained by mapping real data and virtual data, substances having desired properties are identified based on predetermined rules.

Patent Document 2 discloses a search system capable of searching for an unknown combination of physical property parameters having a significant relationship based on already known relationships among arbitrary combinations of multiple physical property parameters. and search methods are disclosed. This search system includes a database, a graph generation unit, and a graph search unit, and is configured as follows. The database stores a plurality of pairs of physical property parameters that have a relationship with each other, and the graph generator uses the plurality of physical property parameters stored in the database as nodes, and the nodes corresponding to the physical property parameter pairs that have a relationship are edges. to generate a graph. The graph search unit searches the graph generated by the graph generation unit based on given search conditions, and outputs search results.

In Patent Document 3, similar to Patent Document 2, in searching for a combination of unknown physical property parameters having a significant relationship among arbitrary combinations of a plurality of physical property parameters, it is possible to perform a search in consideration of priority. A capable search system and search method are disclosed. Similar to Patent Document 2, the relationship between a plurality of physical property parameters is a graph having a plurality of physical property parameters as nodes and edges between nodes corresponding to physical property parameter pairs having a relationship. attached to nodes and/or edges as their attributes. The attributes are used to obtain the priority of the routes extracted as a result of the search, and the results are output in descending order of priority.

JP 2007-18444 A International publication WO2017/221444 International publication WO2018/159237

As a result of examining patent documents 1, 2 and 3, the present inventor found that there are new problems as follows.

The technique described in Patent Literature 1 uses relationships among a plurality of physical property parameters to predict virtual data, but the relationships are limited to already known relationships.

On the other hand, the inventor of the present application is able to extract significant relationships that were unknown due to the huge number of physical property parameters when viewed across many technical fields. , and proposed a search system described in Patent Document 2 as a solution.

According to the search system and search method shown in Patent Documents 2 and 3, it is possible to cross-search all fields including fields with low mutual relationships, thereby finding unknown physical property parameters with significant relationships. Combinations can be extracted.

The inventors have found that conventional search systems assume that the user knows exactly the name of the physical property parameter involved in the search. That is, in the search system disclosed in Patent Literature 1, it is necessary to designate a plurality of physical property parameters as axes in order to configure the two-dimensional or three-dimensional space. Also in the search systems described in Patent Documents 2 and 3, the parameters that can be specified as search conditions are limited to the physical property parameters stored in the database. Therefore, the user must accurately specify the physical property parameter name stored in the database on the system side in advance. Since physical property parameter names are usually established in the technical field, there is no problem if the user is an expert in the technical field. However, the inventors have realized that the user may not have sufficient expertise in the art, such as not knowing the exact name of the physical property parameter. Even if the user knows the desired function, he/she may not know what kind of mechanism works and what kind of physical properties are involved in realizing that function. In addition, even if it is a search for knowing what kind of mechanism works and what kind of physical property is involved, in the conventional search system, the physical property corresponding to either the starting point or the end point of the mechanism that expresses the function is at least It must be entered as a search condition by the user himself.

An object of the present invention is to provide a search support function that assists users who do not have sufficient expertise in the technical field in which they are trying to search for physical properties in a search system in which search conditions including physical property parameters are input. That is.

Means for solving such problems will be described below, but other problems and novel features will become apparent from the description of this specification and the accompanying drawings.

According to one embodiment, it is as follows.

That is, the search system includes a physical property search unit and a user interface, and is configured as follows.

The user interface supplies the physical property search unit with search conditions including at least one physical property parameter, and the physical property search unit performs a physical property search based on the supplied search conditions and outputs the search results to the user interface.

The user interface includes a mechanism estimator and a relevant factor identifier. The mechanism estimating unit estimates, from the input specifications, candidate mechanisms for expressing the function indicated by the specifications, and the related factor identifying unit determines that at least one of the estimated mechanism candidates contributes to the mechanism. Identifies the presence or absence of extension parameters to be used and the extension parameters. Here, the extended parameter is a concept including a physical property parameter and an influencing factor, and the influencing factor is a factor that affects the physical property although it is not a physical property parameter. , morphology factors such as porosity, and size factors such as diameter and volume.

The search conditions supplied by the user interface to the physical property search unit are created based on the mechanism candidates estimated by the mechanism estimation unit or the specified extended parameters.

A brief description of the effects obtained by the above embodiment is as follows.

That is, in a search system in which search conditions including physical property parameters are input, even for users who do not have sufficient technical knowledge in the technical field of physical properties to be searched or who cannot accurately specify physical property parameter names, physical property A search aid can be provided to aid in the search.

FIG. 1 is a block diagram showing a configuration example of a search system according to Embodiment 1. FIG. FIG. 2 is a block diagram showing a configuration example of a search system according to the second embodiment. FIG. 3 is a block diagram showing a configuration example of a search system according to the third embodiment. FIG. 4 is a flow chart showing a configuration example of the search method of the present invention. FIG. 5 is a flow chart showing a modification of the search method of the present invention. FIG. 6 is a flow chart showing a further variant of the search method of the invention. FIG. 7 is an explanatory diagram showing an example of a function expression mechanism database that can be applied to the search system of the present invention. FIG. 8 is a block diagram showing an example of a hardware system in which the search system of the present invention is implemented.

1. Outline of Embodiment First, an outline of a representative embodiment disclosed in the present application will be described. Reference numerals in the drawings, which are referenced in parentheses in the general description of representative embodiments, are merely illustrative of the concept of the component to which they are attached.

[1] <Prediction of the mechanism that determines the expression of function>
A representative embodiment of the present invention is a search system (10) comprising a physical property search section (20) and a user interface (1), and is configured as follows (Fig. 1).

The user interface supplies search conditions including at least one physical property parameter to the physical property search unit, and the physical property search unit performs a physical property search based on the supplied search conditions, and provides search results to the user. It is configured so that it can be output to the interface.

The user interface comprises a mechanism estimator (2) and a related factor identifier (3).

The mechanism estimating unit is configured to be able to estimate one or a plurality of mechanism candidates for manifesting the function meant by the specification from the input specification. Here, specifications are sentences, phrases, or words related to physical property search, and do not necessarily include exact physical property parameter names.

The related factor identification unit can identify the presence or absence of an expansion parameter that contributes to the mechanism and the expansion parameter for at least one of the one or more candidates for the mechanism estimated by the mechanism estimation unit. configured as Here, the extended parameter is a concept including a physical property parameter and an influencing factor, and the influencing factor is a factor that affects the physical property although it is not a physical property parameter. , morphology factors such as porosity, and size factors such as diameter and volume.

The search conditions that the user interface supplies to the physical property search unit are created based on the one or more candidate mechanisms estimated by the mechanism estimation unit or the specified expansion parameters.

As a result, in a search system in which search conditions including physical property parameters are input, even for users who do not have sufficient technical knowledge in the technical field of the physical property to be searched or who cannot accurately specify the physical property parameter name, It is possible to provide a search support function that helps the physical property search. For example, even if the user lacks knowledge about what mechanism exists to express the function that the user has entered as a specification, and what physical properties contribute to that mechanism, it is possible to provide the user with appropriate options for searching. can provide. The influence of the extended parameters specified by the related factor specifying unit can be reflected in the search conditions supplied to the physical property search unit.

[2] <Presentation and Selection of Estimated Mechanism Candidates and Related Information (Comments)>
In the search system described in [1], the user interface further comprises a mechanism selection section (4) (Fig. 2). The mechanism selection unit is configured to present information related to the one or more candidate mechanisms estimated by the mechanism estimation unit to the user and allow the user to select one mechanism in a plurality of cases. The related factor identification unit identifies presence/absence of expansion parameters contributing to the selected mechanism and the expansion parameters. A preferred form of "presentation" is "display", but it is not particularly limited to this, and any means for informing the user may be used.

As a result, when there is a mechanism that the user does not expect to express the function expressed in the specification, it is possible to provide the user with an opportunity to notice it and guide him/her to an appropriate search. Additional relevant information (comments) may be presented to the user to help them understand the mechanism.

[3] <Referring to User History for Selecting Function Expression Mechanism>
In the search system described in item [2], the user interface further comprises a search history holding unit (5) that holds a history of searches performed in the past (Fig. 3). When the mechanism estimation unit estimates a plurality of mechanisms, the mechanism selection unit refers to the search history storage unit and preferentially presents a mechanism related to the history searched in the past. configured to

As a result, mechanism candidates that are highly likely to be selected by the user can be preferentially presented, and the user's burden can be reduced. A user here is not limited to an individual, and may be a user group consisting of a plurality of people.

[4] <Referring to the user's own history when selecting a mechanism for expressing a function>
In the search system described in item [3], the search history holding unit (5) is configured to be able to hold a search history for each user. When the mechanism estimation unit estimates a plurality of mechanisms, the user interface may refer to the search history storage unit and preferentially present mechanisms related to the user's past searches. configured to allow

As a result, it is possible to preferentially present mechanism candidates that are more likely to be selected by the user, thereby reducing the burden on the user. Here, the user himself/herself may be an individual user, or may be a plurality of users belonging to the same user group as the user. In addition to the user himself/herself, it may be configured such that the mechanism related to the search history of the user group to which the user belongs is preferentially presented.

[5] <Referring to the history of users other than yourself when selecting a function expression mechanism>
In the search system described in item [4], when the mechanism estimation unit estimates a plurality of mechanisms, the user interface refers to the search history storage unit and records the user's past searches. , and then preferentially present mechanisms related to histories searched by other users.

As a result, it is possible to preferentially present candidate mechanisms that are highly likely to be selected by the user, and to provide the user with an opportunity to become aware of a powerful function expression mechanism that the user has not anticipated. .

[6] <Function Expression Mechanism Database>
In the search system according to any one of [1] to [5], the mechanism estimating unit includes a function to be expressed, a mechanism that determines the expression of the function, and an extended parameter that contributes to the mechanism and a function expression mechanism database that associates with. The function expression mechanism database may include mechanisms that have no contributing extended parameters.

The mechanism estimating unit identifies a function corresponding to the specification by performing a semantic analysis of the specification, refers to the function expression mechanism database, and identifies one or more mechanisms that determine the expression of the identified function. configured to be identifiable.

As a result, it is possible to estimate candidate mechanisms that determine function expression at a higher speed than executing natural language processing of specifications in real time.

[7] <Search System Equipped with Physical Property Parameter Relationship Database and Graph Search Unit>
In the search system (10) according to any one of items [1] to [6], the physical property search unit includes a physical property parameter relationship database (21), a graph generation unit (22), and a graph a searching unit (24) (FIGS. 1-3).

The physical property parameter relationship database stores a plurality of parameter pairs of physical property parameters having a relationship. The graph generating unit is configured to generate a graph (23) having a plurality of physical property parameters included in the parameter pair as nodes and nodes corresponding to the parameter pairs as edges. The graph search unit is configured to search the graph based on given search conditions via the user interface and output a search result.

Here, physical property parameters among the extended parameters that contribute to the mechanism that determines function expression are configured to be included in the plurality of physical property parameters stored in the physical property parameter relationship database.

As a result, in a search system that enables cross-disciplinary physical property searches by using graph searches, search support functions that enable searches even for users who cannot accurately specify physical property parameter names. can be provided.

[8] <Expansion of physical property parameter relationship database and graph search unit>
In the search system (10) of item [7], the physical property parameter relationship database further stores extended pairs in addition to the plurality of parameter pairs. Here, an extended pair is a pair of an extended parameter having a relationship with a plurality of physical property parameters stored in the physical property parameter relationship database and the physical property parameter. The graph generator adds, to the graph, edges between nodes corresponding to the expansion parameters forming the expansion pair and nodes corresponding to physical property parameters forming the pair.

As a result, even in a search system that can search for physical properties including the contribution of influencing factors other than physical property parameters, it is possible to provide a search support function for users who do not have sufficient expertise in the technical field of the physical properties to be searched.

[9] <Search method; Estimation of the mechanism that determines the expression of function>
A representative embodiment of the present invention is implemented by software operating on a computer having a storage device, and includes a specification input step (S1), a mechanism estimation step (S2), a related factor identification step (S3), and a search condition input. A search method comprising a step (S4), a physical property search step (S5), and a search result output step (S6) is configured as follows (FIG. 4).

In the mechanism estimation step, from the specifications input in the specification input step (S1), one or a plurality of candidate mechanisms for expressing the functions implied by the specifications are estimated.

The relevant factor identifying step identifies presence/absence of an expansion parameter contributing to the mechanism and the expansion parameter for at least one of the one or more candidates for the mechanism.

The physical property search step performs a physical property search with search conditions based on the estimated one or more mechanism candidates or the specified extended parameters as input, and the search result output step outputs the search results.

As a result, in a search method in which search conditions including physical property parameters are input, even for users who do not have sufficient technical knowledge in the technical field of the physical property to be searched and who cannot accurately specify the physical property parameter name, A search support function can be provided that enables searching. More specifically, it has the same function and effect as the item [1].

[10] <Presentation and Selection of Inferred Mechanism Candidates and Related Information (Comments)>
[9] The search method further includes a mechanism selection step (S7) (Fig. 5).

The mechanism selection step presents information related to the one or more candidate mechanisms estimated in the mechanism estimation step to the user, and causes the user to select one mechanism in a plurality of cases. A mechanism selection step (S7) is inserted after the mechanism estimation step and before or after the relevant factor identification step.

As a result, when there is a mechanism that the user does not expect to express the function expressed in the specification, it is possible to provide the user with an opportunity to notice it and guide him/her to an appropriate search. Additional relevant information (comments) may be presented to the user to help them understand the mechanism.

[11] <Referring to User History for Selecting Function Expression Mechanism>
The search method described in item [10] further includes a search history holding unit (5) that holds a history of searches performed in the past in the storage device (FIG. 6).

When a plurality of mechanisms are estimated in the mechanism estimation step, the mechanism selection step can refer to the search history storage unit and preferentially present a mechanism related to the history searched in the past. configured as

As a result, mechanism candidates that are highly likely to be selected by the user can be preferentially presented, and the user's burden can be reduced.

[12] <Referring to the user's own history when selecting a mechanism for expressing a function>
[11] In the search method described in item [11], the search history holding unit is configured to be able to hold a search history for each user.

In the mechanism selection step, when a plurality of mechanisms are estimated in the mechanism estimation step, the search history holding unit is referred to and preferentially presents a mechanism related to a history of searches by the user himself/herself in the past. configured to allow

As a result, it is possible to preferentially present mechanism candidates that are more likely to be selected by the user, thereby reducing the burden on the user.

[13] <Referring to the history of users other than yourself when selecting a function expression mechanism>
In the search method described in [12], the mechanism selection step refers to the search history storage unit when a plurality of mechanisms are estimated in the mechanism estimation step, and the user himself/herself in the past It is configured to be able to preferentially present a mechanism related to a searched history, and then preferentially present a mechanism related to a history searched by another user.

As a result, it is possible to preferentially present candidate mechanisms that are highly likely to be selected by the user, and to provide the user with an opportunity to become aware of a powerful function expression mechanism that the user has not anticipated. .

[14] <Function Expression Mechanism Database>
The search method according to any one of [9] to [13] is characterized by associating a function to be expressed, a mechanism that determines the expression of the function, and an expansion parameter that contributes to the mechanism. A mechanism database (5) is further provided in said storage device (Fig. 6).

The mechanism estimation step identifies a function corresponding to the specification by performing a semantic analysis of the specification, refers to the function expression mechanism database, and identifies one or more mechanisms that determine the expression of the identified function. Identify.

As a result, it is possible to estimate candidate mechanisms that determine function expression at a higher speed than executing natural language processing of specifications in real time.

[15] <Search method including physical property parameter relationship database and graph search step>
In the search method described in any one of [9] to [14], the physical property parameter relationship database (S11) is provided in the storage device, and the physical property search step includes a graph generation step (S12 ) and a graph search step (S13) (FIGS. 4-6).

The physical property parameter relationship database stores a plurality of parameter pairs of physical property parameters that are related to each other. The graph generating step is configured to be able to generate a graph having a plurality of physical property parameters included in the parameter pairs as nodes and nodes between the nodes corresponding to the parameter pairs as edges, and the graph searching step includes , searches the graph based on given search conditions and outputs the search results.

Here, the physical property parameters among the extended parameters that contribute to the mechanism that determines the function expression are included in the plurality of physical property parameters stored in the physical property parameter relationship database.

As a result, in a search method that enables cross-disciplinary physical property searches by using graph searches, search support functions that enable searches even for users who cannot accurately specify physical property parameter names. can be provided.

[16] <Expansion of physical property parameter relationship database and graph generation/search step>
In the search method of item [15], the physical property parameter relationship database further stores extended pairs in addition to the plurality of parameter pairs. Here, an extended pair is a pair of an extended parameter having a relationship with a plurality of physical property parameters stored in the physical property parameter relationship database and the physical property parameter. The graph generator adds, to the graph, edges between nodes corresponding to the expansion parameters forming the expansion pair and nodes corresponding to physical property parameters forming the pair.

As a result, even in a search method that can perform a physical property search including the contribution of influencing factors other than physical property parameters, it is possible to provide a search support function for users who do not have sufficient expertise in the technical field of the physical property to be searched.

2. Details of the Embodiments The embodiments are further described in detail.

[Embodiment 1]
FIG. 1 is a block diagram showing a configuration example of a search system according to Embodiment 1. FIG.

A search system 10 according to the first embodiment includes a physical property search unit 20 and a user interface 1 .

The user interface 1 supplies search conditions including at least one physical property parameter to the physical property search unit 20, and the physical property search unit 20 performs a physical property search based on the supplied search conditions, and sends the search results to the user interface 1. configured for output.

A user interface 1 includes a mechanism estimator 2 and a related factor identifier 3 .

The mechanism estimating unit 2 estimates one or a plurality of mechanism candidates for expressing the function implied by the specification from the input specification. Here, the input specifications are sentences, phrases, or words that express, for example, queries or requests related to physical property search, and do not necessarily include accurate physical property parameter names. It doesn't have to be.

The related factor identifying unit 3 can identify the presence or absence of an expansion parameter that contributes to the mechanism and the expansion parameter for at least one of the one or more candidate mechanisms estimated by the mechanism estimating unit 2. configured to

The search conditions supplied from the user interface 1 to the physical property searching unit 20 are created based on the mechanism candidate estimated by the mechanism estimating unit 2 and/or the extended parameters contributing to the mechanism. The user interface 1 may present the user with candidate mechanisms and/or extended parameters that contribute to the mechanism, to help the user create search conditions, or based on the deduced mechanism, automatically A search condition may be created and supplied to the physical property search unit 20 .

As a result, in a search system in which search conditions including physical property parameters are input, it is possible to provide a search support function that enables searches even for users who cannot accurately specify physical property parameter names. If the user lacks knowledge about what kind of physical properties contribute to the mechanism that manifests the function that the user entered as a specification, or if there is a mechanism unexpected by the user that manifests the function can also provide users with appropriate choices for their search. Thus, the mechanism estimated by the mechanism estimation unit 2 and/or the extended parameters specified by the related factor specification unit 3 are reflected in the search conditions supplied to the physical property search unit 20 .

In the present invention, the extended parameter is a concept including a physical property parameter and an influencing factor, and the influencing factor is a factor that affects the physical property but is not a physical property parameter, for example, environmental factors such as temperature and pressure, Form factors such as spherical and porous and size factors such as diameter and volume are included. The specifications input by the user describe the functions desired by the user, but the factors contributing to the mechanism for expressing the functions are not limited to the physical property parameters. In addition, there are many cases in which the physical property parameters change depending on the influencing factors. In the present invention, although not a physical property parameter, a factor that affects physical properties is called an "influence factor" and is treated in accordance with a physical property parameter. there is For users who search for physical properties, it is sufficient to find something that satisfies the desired specifications. I don't care whether or not. Therefore, in the user interface 1, by handling it collectively as an extended parameter, even for users who do not have sufficient expertise in the technical field of the physical property to be searched or who cannot accurately specify the physical property parameter name, A function that can provide various support is provided.

If the physical property searching unit 20 does not support factors other than physical property parameters as search conditions, such influencing factors cannot be involved in the physical property search. Because it is information, it is worth presenting to the user. Alternatively, if the search range of the physical property searching unit 20 includes influencing factors, search conditions (search formulas) to be input to the physical property searching unit 20 can be created from all extended parameters related to the mechanism.

For example, if the search target is “a flat plate that is transparent and the inside can be seen, or the display inside can be seen from the outside, but functions as an electrode,” the above sentence can be entered as the specification. Alternatively, phrases or words such as "transparent plate that functions as an electrode", "transparent electrode", or "transparent electrode" may be input as specifications.

There are mainly two mechanisms for making an "electrode" exhibit the function of "transparency" or allowing a "transparent" substance to exhibit the function of "electrode".

One mechanism is when a "transparent electrode" is achieved by allowing the electrode material itself to transmit light. Another mechanism is that a "transparent electrode" is realized by mixing a conductive material with a fine structure that does not affect transparency into an insulator or a transparent material with high electrical resistance. be.

In the former case, the extended parameters contributing to this mechanism are 'optical transmittance' or 'optical absorption rate' and 'electrical conductivity', 'conductivity' or 'electrical resistance', and in the latter case , "grain size", "open area ratio", and "electrical conductivity", "conductivity" or "electrical resistance" that define the microstructure.

Here, "light transmittance", "light absorption rate", "electrical conductivity", "conductivity" and "electrical resistance" are physical property parameters. On the other hand, "particle size" and "aperture ratio" are not physical parameters, but they are still factors that contribute to the mechanism of manifesting the function of making the microstructure look transparent. It is treated according to the parameters.

Even if the user does not know at least one of the above two mechanisms, the search system 10 can present the user with two mechanisms that cause the "electrode" to exhibit the function of "transparency". So users can find solutions based on mechanisms they didn't know existed. In addition, since it is possible to know the extended parameters involved in the mechanism as described above, search conditions (search formulas) to be input to the physical property searching unit 20 can be created without knowing in advance the names of physical properties or physical property parameters that contribute to each. can do. As extended parameters involved in the mechanism, influencing factors that are not physical parameters, in this example "particle size" and "aperture ratio" can also be configured to be presented to the user.

Another example will be described. As a mechanism for expressing the function of "seeing black", there are cases where "the inherent color of the material is black" and where "the structural color is black". The physical parameter that contributes to the former is "light transmittance" or "light absorption". The latter is a mechanism in which the incident light cannot change its direction and cannot be emitted or reflected due to structural reasons such as the surface of the substance being porous, which is visually recognized as a black color by the observer. . Therefore, contributing factors are porous "pore diameter", "pore depth", "pore density" and the like. These are influencing factors rather than physical parameters.

The search system 10 causes the user interface 1 to allow the user to select one mechanism from one or more candidate mechanisms estimated by the mechanism estimating unit 2, and/or the mechanism estimated by the mechanism estimating unit 2 or An interface may be provided that allows the user to create search criteria (search formulas) based on the specified expansion parameters. On the other hand, a mechanism performed by the search system 10 may be provided without involving the user in the selection of the mechanism and the creation of search conditions (search formula).

Even if the user does not know the existence of such multiple mechanisms, it provides an opportunity for the user to notice, and even if the user knows the mechanism that expresses the function, the name of the physical property parameter that contributes to the mechanism can be accurately identified. Even if you don't know, you can enter search conditions, including the correct and appropriate property parameter names, and perform a property search.

As described above, according to the search system 10 of Embodiment 1, in a search system in which search conditions including physical property parameters are input, technical knowledge in the technical field of physical properties to be searched is insufficient, or physical property parameter names are It is possible to provide a search support function that assists the physical property search even for users who cannot accurately specify the property. In addition, it is possible to provide a search support function capable of providing the user with appropriate options for the search for a mechanism for manifesting the function input by the user as a specification.

<Configuration example of physical property search unit>
A physical property search unit 20 in the search system 10 performs a physical property search based on search conditions and outputs search results to the user interface 1 . For the physical property search unit 20, for example, the search system described in Patent Documents 1, 2, or 3 can be applied. A simple application example is shown in FIG.

The physical property search unit 20 includes a physical property parameter relationship database 21 that stores relationships of physical property parameters, a graph generation unit 22 and a graph search unit 24 .

A plurality of pairs of related physical property parameters are stored in the physical property parameter relationship database 21 . The pair of physical property parameters having a relationship at this time is not only based on a relationship based on scientific grounds, that is, based on a relationship that has been theoretically explained, but has not yet been theoretically explained, Moreover, even at the stage where no formalization has been made, a pair of physical property parameters whose relationship is known to exist can be included by recognizing a clear correlation from experimental data. In addition to relationships formulated like theorems and formulas, “relationships explained theoretically” include the presence or absence of correlation and the sign of the correlation coefficient (when one increases, the other also increases). or decrease) may be broadly included, even semi-quantitative or qualitative relationships. At this time, any relationship known in any field need not be particularly excluded, and physical property parameter pairs whose relationship is known in any field can be included.

The graph generation unit 22 generates a graph 23 in which a plurality of physical property parameters stored in the physical property parameter relationship database 21 are used as nodes, and nodes corresponding to physical property parameter pairs having a relationship are used as edges. The graph search unit 24 searches the graph 23 based on given search conditions and outputs the search result. Various route search methods based on mathematical algorithms such as graph theory can be applied to the graph search unit 24 .

Thereby, the physical property searching unit 20 can search for an unknown combination of physical property parameters having a significant relationship among arbitrary combinations of a plurality of physical property parameters, based on already known relationships. For example, if there is a causal relationship between two physical property parameters, and if there is a causal relationship, a path for explaining the relationship can be searched. The route for explaining the relationship between the two physical property parameters is not necessarily one, but a plurality of routes are extracted as search results. These may include causal relationships that cannot be explained solely by known causal relationships within a single technical field. The physical property parameter relationship database 21 includes physical property relationships known in a plurality of technical fields, and the graph 23 is exhaustive regardless of what technical field the relationship is known. , the result of the route search is also irrelevant to the technical field.

Further, the physical property search unit 20 is appropriately added with functions such as search considering conditional edges as described in Patent Document 2 and route search considering priority as described in Patent Document 3. • Enhanced search can be applied.

The search conditions include the nodes specified as the start point, end point, and waypoint of the route search as described above, i.e., physical property parameters, and conversely, the nodes specified under the condition that they should not be routed, that is, physical property parameters. Instead of a route search specifying two or three or more nodes as a search condition, it is also possible to specify a single start point or end point node to search for nodes existing within a predetermined range. It is possible to search not only for nodes that are close to each other, ie, that have a strong causal relationship between physical property parameters, but also for nodes that are within a certain distance, or that are separated by a certain distance or more.

Any search condition must accurately specify at least one physical parameter. On the other hand, users who search physical properties do not necessarily have sufficient academic knowledge and do not necessarily know the names of physical property parameters accurately. In particular, when searching for a material having a certain function, it seems that there are not a few users who do not know the mechanism by which the function is expressed and the physical properties that determine the function expression.

In the search system 10 of Embodiment 1, the physical property parameters included in the extended parameters specified by the related factor identifying unit 3 are selected from the physical property parameter names stored in the physical property parameter relationship database 21 by the mechanism estimating unit 2. It is preferable to select the one that contributes to the mechanism estimated by

Even if the user inputs a specification that does not use an accurate physical property parameter name, the search system 10 estimates the mechanism that causes the function intended by the specification to be expressed, and accurately names the physical property parameter that contributes to it. That is, since the name of the physical property parameter stored in the physical property parameter relationship database 21 is specified, the search condition input to the graph search unit 24 is appropriate. As a result, it is possible to provide a support function for inputting search conditions including accurate physical property parameter names to the physical property search unit 20 even for users who do not know the mechanism that expresses the function and the physical properties that determine the expression of the function. can.

The physical property search unit 20 can adopt various search systems instead of the search system applying the graph search as exemplified here. No matter what search system is adopted, as long as the search system requires accurate input of physical property parameter names as search conditions, the mechanism estimating unit 2 and the related factor identifying unit of the first embodiment By adopting the user interface 1 including 3, the same effect can be achieved. The same is true for other embodiments of the invention.

The function of the physical property searching unit 20 may be expanded to enable searching not only for physical property parameters but also for extended parameters including influence factors.

For this purpose, the physical property parameter relationship database 21 further stores extended pairs in addition to a plurality of originally stored parameter pairs. Here, an extended pair is a pair of an extended parameter having a relationship with a plurality of physical property parameters stored in the physical property parameter relationship database 21 and the physical property parameter. For example, when the physical property parameters A and B have a relationship, the physical property parameter relationship database 21 stores the physical property parameters A and B and their relationship. In this case, the pair of the temperature T and the physical property parameter A, which are influential factors and one of the extended parameters, and their relationship are added to the physical property parameter relationship database 21 as extended pairs.

The graph generation unit 22 expands the graph 23 to be generated based on the expansion pair of the expansion parameter and the physical property parameter additionally stored in the physical property parameter relationship database 21 . That is, the edge between the node corresponding to the expansion parameter forming the expansion pair of the expansion parameter and the physical property parameter and the node corresponding to the physical property parameter forming the expansion pair is added to the graph 23 . Since the search range of the graph search unit 24 is the extended graph 23, the physical property search unit 20 can search not only for physical property parameters but also for extended parameters including influence factors.

As a result, even in a search system that can search for physical properties including the contribution of influencing factors other than physical property parameters, it is possible to provide a search support function for users who do not have sufficient expertise in the technical field of the physical properties to be searched.

<Configuration example of mechanism estimation unit and related factor identification unit>
The mechanism estimating unit 2 and the related factor identifying unit 3 in the search system 10, for example, correlate the function to be expressed, the mechanism that determines the expression of the function, and one or more physical property parameters that contribute to the mechanism. It is preferably configured with a mechanism database.

FIG. 7 is an explanatory diagram showing an example of a function expression mechanism database that can be applied to the search system of the present invention.

In the function expression mechanism database 11, keywords used in specifications are associated with mechanisms for expressing functions intended by the specifications, and comments are added as necessary. Furthermore, each mechanism is associated with extension parameters that contribute to the mechanism. The mechanism database 11 may contain mechanisms for which there are no contributing extended parameters.

The keyword "transparent electrode" is associated with the mechanism "transparent electrode material", with a comment indicating that "the conductive electrode material itself is transparent". Furthermore, as physical parameters contributing to this mechanism, "light transmittance" and "electrical conductivity" are associated. In addition, the same keyword "transparent electrode" is associated with another mechanism called "microstructure electrode", which means that "microscopic electrodes that do not affect the transparency of an insulator or a transparent material with high electrical resistance are formed. There is a comment stating that the structure is "mixed with conductive material". Further, "particle size", "opening ratio" and "electrical conductivity" are associated as physical property parameters contributing to this mechanism.

The keyword "substance that looks black" is associated with two mechanisms: "material specific color" and "structural color." A comment is attached. Further, as physical parameters contributing to each of these mechanisms, the "optical absorptance" is associated with the porous "pore diameter", "pore depth" and "pore density".

In addition to the single keyword shown in the example, specifications can have different expressions using various synonyms and synonyms, so such synonyms and synonyms are added to the "keyword" item of the function expression mechanism database. can do.

On the other hand, the mechanism estimating unit 2 may be provided with a means for interpreting the meaning of the specification to estimate which "keyword" in the function expression mechanism database corresponds to the meaning of the specification. Various known techniques can be applied to the means for interpreting the meaning of natural language.

Although FIG. 7 illustrates a tabular database, databases of various other formats can be employed as the function manifestation mechanism database 11 . Also, a function equivalent to the function expression mechanism database 11 can be incorporated into the mechanism estimation unit 2 and the related factor identification unit 3 .

<Method for creating a functional expression mechanism database>
The function expression mechanism database 11 can be created by experts such as researchers based on their findings. Some or all of its creation work can be supplemented or replaced by artificial intelligence. When the scope of the search target of the physical property search unit 20 extends to a large number of fields, the knowledge of many experts will be collected and created. According to artificial intelligence, learning data to be input can be collected from a wide range of fields and machine-learned to easily aggregate knowledge from a wide range of fields.

For example, technical Q&As, explanations, papers, and other documents posted on websites on the Internet are input as learning data, and words are clustered. Clustering is a type of natural language processing that vectorizes the words that appear in documents input as learning data, and divides them into groups (clusters) of words that are close to the same concept based on their direction and size. is.

In the present invention, keywords that may be used in the specification input by the user, words that express the function intended by the specification, words that express the mechanism that determines the expression of the function, and contributions to the mechanism Machine learning is designed such that one or more extension parameters to be clustered for each mechanism. The physical property parameters included in the extended parameters that contribute to the mechanism are common to the physical property parameter names stored in the physical property parameter relationship database 21 . A word or phrase to be input to the "mechanism" of the function expression mechanism database 11 shown in FIG. 7 is selected from among the clustered words. In addition, by referring to them, phrases and sentences to be input in the "comment" are created. These operations may be performed manually. Words within the same cluster are registered as keywords.

If the function manifestation mechanism database 11 is created by the method using artificial intelligence as described above, it is possible to obtain an effect equivalent to consolidating the knowledge of experts in many fields.

<Hardware/software implementation form>
The search system 10 of the present invention is functionally constructed as software on a hardware system comprising a storage device and a computer.

FIG. 8 is a block diagram showing an example of a hardware system in which the search system 10 of the present invention is implemented.

A server 100 and user workstations 110 and 120 are connected to a network 200 such as the Internet. The server 100 has a computer 101 , a storage device 102 , a network interface 103 , an input section 104 and a display section 105 . User workstations 110 and 120 also have computers 111 and 121, storage devices 112 and 122, network interfaces 113 and 123, input units 114 and 124, and display units 115 and 125, respectively.

In the search system 10, for example, the physical property search section 20 is implemented in the server 100, and the user interface 1 is implemented in the user's workstations 110 and/or 120, respectively. In the case where the user interface 1 is installed on the user's workstation 110, the function expression mechanism database 11 is held in the storage device 112, and the mechanism estimation unit 2 and the related factor identification unit 3 operate on the computer 111. Implemented as software. Search conditions are supplied from the workstation 110 on the user side to the physical property search unit 20 mounted on the server 100 via the network 200, and the results of the physical property search executed on the server 100 are sent via the network 200. Then, it is output to the user interface 1 mounted on the work station 110 of the user who entered the specification. The same applies when the user interface 1 is implemented on another user-side workstation 120 .

Instead of such a mode, not only the physical property searching unit 20 but also the user interface 1 may be implemented in the server 100 . The workstations 110 and 120 on the user side may each be equipped with only a simple user interface function such as accepting specifications and outputting search results.

Furthermore, the entire search system 10 may be implemented on a single computer such as the server 100, and the network 200 and user workstations 110 and 120 may be omitted.

[Embodiment 2] <Presentation and selection of estimated mechanism candidates>
The search system 10 may include an interface that allows the user to select one mechanism from a plurality of candidate mechanisms estimated by the mechanism estimation unit 2 in the user interface 1 .

FIG. 2 is a block diagram showing a configuration example of the search system 10 according to the second embodiment.

The search system 10 includes a physical property search unit 20 and a user interface 1, similar to the search system 10 according to Embodiment 1 shown in FIG. is supplied, and the physical property search unit 20 is configured to perform a physical property search based on the supplied search conditions and output the search results to the user interface 1 .

Since the physical property searching unit 20 is the same as that of the first embodiment, detailed description thereof will be omitted.

A user interface 1 includes a mechanism selection unit 4 in addition to a mechanism estimation unit 2 and a related factor identification unit 3 . The mechanism estimating unit 2 and the related factor identifying unit 3 are the same as those described in the first embodiment, so detailed descriptions thereof will be omitted. The mechanism selection unit 4 is configured to output information on one or a plurality of mechanisms estimated by the mechanism estimation unit 2 to the user so that one mechanism can be selected in a plurality of cases. Here, the information about the mechanism estimated by the mechanism estimation unit 2 is information for informing the user of the content of the mechanism that expresses the function that the specification input by the user means. For example, the function shown in FIG. Information stored as "mechanism" in the expression mechanism database 11 can be used.

As a result, if there are multiple mechanisms for expressing the functions expressed in the specification, and if there is a mechanism that the user does not expect, it will be possible to provide awareness to the user and guide them to an appropriate search. can be done.

It is more preferable to configure the mechanism selection unit 4 so that it can also present to the user a "comment" associated with the mechanism. The user can more accurately understand the presented mechanism options through the presented comments, and can conduct a more appropriate search.

The related factor identification unit 3 identifies extended parameters that contribute to the mechanism selected by the mechanism selection unit 4 . The specified extended parameters may be presented to the user for reference when creating search conditions (search formulas). If the mechanism selected by the mechanism selection unit 4 does not have a contributing extension parameter, the fact is presented to the user. It is not always necessary to search for physical properties, and since the user can know the mechanism by which the function expressed by the input specification is expressed, it is sufficient to use that mechanism to satisfy the specification.

A search condition (search formula) may be created within the user interface 1 from the physical property parameters among the specified expanded parameters and supplied to the physical property searching section 20 . When the number of physical property parameters included in the identified extended parameter is one, for example, a subgraph containing nodes within a predetermined range from the node corresponding to the single physical property parameter is displayed as the search result. When the number of physical property parameters included in the identified extended parameters is two, for example, a route between two nodes corresponding to two physical property parameters is searched and displayed. If the number of physical property parameters included in the identified extended parameters is three or more, for example, search for and display a subgraph containing nodes corresponding to all physical property parameters included in the identified extended parameters, It is possible to automatically create various search conditions (search formulas), such as various searches for routes passing through all nodes to be searched while changing the combination of start points, end points, and waypoints.

When the physical property search unit 20 is configured to expand the function so as to enable searching for extended parameters including influencing factors as well as physical property parameters, the user interface unit 1 identifies related factors. For the expanded parameters specified by the unit 3, search conditions (search formulas) are automatically created and supplied to the physical property search unit 20 in the same manner as described above.

The mechanism selection unit 4 may be configured to present the contributing extension parameter together with the mechanism in addition to the comment associated with the mechanism selected by the mechanism selection unit 4, or instead of the "comment". good. Although FIG. 2 shows an example in which the related factor identification unit 3 is arranged after the mechanism selection unit 4, this is replaced to identify the related extended parameters for each of the mechanism candidates estimated by the mechanism estimation unit 2. and configure it so that it can be presented to the user and selected. The user can refer to the presented mechanism to create search conditions (search formulas) based on extended parameters, input them to the physical property search unit 20, and execute the physical property search.

[Embodiment 3] <Referring to User History for Selecting Function Expression Mechanism>
The search system 10 refers to the user's search history and preferentially presents a more appropriate mechanism candidate when the user selects one mechanism from a plurality of mechanism candidates estimated by the mechanism estimation unit 2. can be configured to allow

FIG. 3 is a block diagram showing a configuration example of the search system 10 according to the third embodiment.

The search system 10 includes a physical property search unit 20 and a user interface 1, similar to the search systems 10 according to the first and second embodiments shown in FIGS. A search condition including one physical property parameter is supplied, and the physical property search unit 20 performs a physical property search based on the supplied search condition, and is configured to be able to output the search result to the user interface 1. . Since the physical property searching unit 20 is the same as that of the first and second embodiments, detailed description thereof is omitted.

The user interface 1 includes a search history storage unit 5 in addition to the mechanism estimation unit 2, mechanism selection unit 4, and related factor identification unit 3 shown in FIG. The mechanism estimation unit 2 and the related factor identification unit 3 are the same as those described in the first embodiment, and the mechanism selection unit 4 is the same as that described in the second embodiment. omitted.

The search history holding unit 5 is configured to be able to hold the history of searches executed in the past. When the mechanism estimating unit 2 estimates a plurality of mechanisms, the mechanism selecting unit 4 refers to the search history holding unit 5 so as to preferentially present the mechanism related to the history searched in the past. configured to

As a result, it is possible to preferentially present mechanism candidates that are more likely to be selected by the user, thereby reducing the burden on the user. A user here is not limited to an individual, and may be a user group consisting of a plurality of people.

The search history holding unit 5 holds the search condition (search formula) supplied to the physical property search unit 20 as a search history, and also holds, as a search history, mechanisms presented to the user from the mechanism selection unit 4 and selected by the user. may The stored search history can be browsed, output, edited, and deleted by the user.

<Priority presentation of the user's own history when selecting a mechanism for expressing a function>
In the search system 10 according to the third embodiment, it is more preferable that the search history holding unit 5 be configured so as to be able to hold the search history for each user. When the mechanism estimation unit 2 estimates a plurality of mechanisms, the user interface 1 can refer to the search history storage unit 5 and preferentially present mechanisms related to the user's past searches. configured to allow

As a result, it is possible to preferentially present mechanism candidates that are more likely to be selected by the user, thereby reducing the burden on the user. Here, the user himself/herself may be an individual user, or may be a plurality of users belonging to the same user group as the user.

Additionally, mechanisms related to histories that other users have explored may be configured so that they may be preferentially presented next. As a result, mechanisms that have never been explored by any user are presented with the lowest priority. It is possible to provide the user with an opportunity to become aware of a powerful function expression mechanism that has not been developed before.

[Embodiment 4] <Search method>
The search system 10 of the present invention described in Embodiments 1 to 3 as described above includes a storage device and a computer, as described in the "hardware/software implementation" of Embodiment 1 with reference to FIG. Functions can be constructed as software on a hardware system. Therefore, the present invention can be positioned as a search method using a hardware system having a storage device and a computer.

FIG. 4 is a flow chart showing a configuration example of the search method of the present invention.

The search method of the present invention is implemented by software operating on a computer having a storage device, and includes a specification input step (S1), a mechanism estimation step (S2), a related factor identification step (S3), and a search condition input step (S4). , a physical property search step (S5), and a search result output step (S6), and is configured as follows.

A mechanism estimation step (S2) estimates, from the specifications input in the specification input step (S1), one or a plurality of candidate mechanisms for manifesting the function implied by the specifications. A related factor identifying step (S3) identifies, for at least one of the one or more candidate mechanisms estimated in the mechanism estimating step (S2), extended parameters that contribute to the mechanism. For implementation of the mechanism estimation step (S2) and the related factor identification step (S3), for example, software that includes and refers to the function expression mechanism database 11 as shown in FIG. 7 can be applied. The method of creating the function expression mechanism database 11, the method of estimating a mechanism with reference to it, and the method of specifying extended parameters that contribute to the mechanism can be the same as the method described in the first embodiment.

A search condition (search formula) input from the search condition input step (S4) or a search condition created including one or more physical property parameters among the extended parameters specified in the related factor specifying step (S3) (search formula) is supplied to the physical property search step (S5). The physical property search step (S5) performs a physical property search based on the supplied search conditions (search formula) and outputs the search results (S6).

The physical property searching step (S5) includes, for example, a graph generating step (S12) for generating a graph from the physical property parameter relationship database, and a graph searching step for searching the generated graph according to input search conditions. (S13). A plurality of pairs of related physical property parameters are stored in the physical property parameter relationship database. The physical property parameter relationship database may be read by the physical property parameter relationship database input step (S11). In addition, the search method described in the above-mentioned Patent Documents 1, 2, or 3 can also be applied to the physical property search step (S5). As a result, as in the search system 10 according to the first embodiment, even in a search method in which search conditions including physical property parameters are input, searches can be performed even for users who cannot accurately specify physical property parameter names. It is possible to provide a search support function that

The physical property searching step (S5) can be configured in the same manner as described in the “configuration example of the physical property searching unit” of the first embodiment. Further, the function may be expanded to enable searching not only for physical property parameters but also for expanded parameters including influence factors.

As a result, even in a search system that can search for physical properties including the contribution of influencing factors other than physical property parameters, it is possible to provide a search support function for users who do not have sufficient expertise in the technical field of the physical properties to be searched.

<Presentation and selection of inferred mechanism candidates>
A mechanism selection step (S7) can be added to the search method described above (FIG. 4).

FIG. 5 is a flow chart showing a modification of the search method of the present invention. In the flowchart shown in FIG. 4, a mechanism selection step (S7) is inserted between the mechanism estimation step (S2) and the related factor identification step (S3).

The mechanism selection step (S7) outputs information regarding one or more mechanisms estimated in the mechanism estimation step (S2) to the user, and prompts the user to select one mechanism in a plurality of cases. The relevant factor identification step (S3) identifies extended parameters that contribute to the mechanism selected in the mechanism selection step (S7). Other steps are the same as the steps in the search method described with reference to FIG. 4, so detailed description thereof will be omitted.

As a result, if there is a mechanism that the user did not expect to express the function expressed in the specification, the user can be notified and guided to an appropriate search.

FIG. 5 shows an example in which the mechanism selection step (S7) is inserted before the related factor specifying step (S3), but instead of this, it may be placed after the related factor specifying step (S3). good. In the relevant factor identification step (S3), an extended parameter contributing to each candidate mechanism is specified, and in the mechanism selection step (S7), the extended parameter contributing to each of the multiple candidate mechanisms is selected for reference when selecting the mechanism. , can be presented to the user.

<Referring to user history when selecting a mechanism for expressing a function>
A search history storage unit (5) can be added to the search method (FIG. 5) described above.

FIG. 6 is a flow chart showing a further variant of the search method of the invention. The search history storage unit (5) stores the search conditions (search formula) input from the search condition input step (S4) or the search conditions (search formula) including the physical property parameters specified in the related factor specifying step (S3). , the search condition (search formula) supplied to the physical property search step (S5) is held in the storage device.

In the mechanism selection step (S7), when a plurality of mechanisms are estimated in the mechanism estimation step, the search history storage unit (5) is referred to, and the mechanism related to the history searched in the past is preferentially presented. configured to be able to

As a result, mechanism candidates that are highly likely to be selected by the user can be preferentially presented, and the user's burden can be reduced.

Furthermore, the search history holding unit (5) is configured to be able to hold and manage the search history for each user, and the mechanism selection step (S7) is performed by the user himself/herself or a user group to which the user belongs in the past. It is more preferable to configure so that mechanisms related to searched histories can be preferentially presented. As a result, it is possible to preferentially present mechanism candidates that are more likely to be selected by the user.

At this time, it is preferable that the mechanism related to the search history of other users can be preferentially presented next to the mechanism related to the search history of the user. As a result, it is possible to preferentially present candidate mechanisms that are highly likely to be selected by the user, and to provide the user with an opportunity to become aware of a powerful function expression mechanism that the user has not anticipated. .

Although the invention made by the present inventor has been specifically described based on the embodiments, the present invention is not limited thereto, and it goes without saying that various modifications can be made without departing from the gist of the invention.

1 user interface 2 mechanism estimation unit 3 related factor identification unit 4 mechanism selection unit 5 search history storage unit 10 search system 11 function expression mechanism database 21 physical property parameter relationship database 22 graph generation unit 23 graph 24 graph search unit 100 server 110, 120 Workstations 101, 111, 121 Computers 102, 112, 122 Storage Devices 103, 113, 123 Network Interfaces 104, 114, 124 Input Units 105, 115, 125 Display Units 200 Network

Claims (16)

A search system comprising a physical property search unit and a user interface,
The user interface supplies search conditions including at least one physical property parameter to the physical property search unit, and the physical property search unit performs a physical property search based on the supplied search conditions, and provides search results to the user. configured to be able to output to the interface,
The user interface comprises a mechanism estimating unit and a related factor identifying unit,
The mechanism estimating unit is configured to be able to estimate one or a plurality of mechanism candidates for expressing the function meant by the specification from the input specification,
The relevant factor identifying unit is configured to be able to identify, for at least one of the one or more mechanism candidates, a physical property parameter that contributes to the mechanism ,
The search conditions supplied by the user interface to the physical property searching unit are created based on the physical property parameters specified by the relevant factor specifying unit ,
search system. In claim 1,
The user interface further comprises a mechanism selector,
The mechanism selection unit presents information related to the one or more mechanisms estimated by the mechanism estimation unit to the user, and allows the user to select one mechanism in a plurality of cases, and the related factor identification unit selects Configured to be able to identify the presence or absence of a physical parameter that contributes to the mechanism and the physical parameter ,
search system. In claim 2,
The user interface further comprises a search history holding unit that holds a history of searches performed in the past,
When the mechanism estimating unit estimates a plurality of mechanisms, the mechanism selecting unit refers to the search history holding unit and preferentially presents a mechanism related to the history searched in the past. consists of
search system. In claim 3,
The search history holding unit is configured to be able to hold a search history for each user,
When the mechanism estimation unit estimates a plurality of mechanisms, the user interface may refer to the search history storage unit and preferentially present mechanisms related to the user's past searches. configured to allow
search system. In claim 4,
When the mechanism estimating unit estimates a plurality of mechanisms, the user interface refers to the search history holding unit, preferentially presents mechanisms related to the user's past search history, and others. is configured to be able to preferentially present mechanisms related to the user's searched history to the next,
search system. In any one of claims 1 to 5,
The mechanism estimating unit comprises a function expression mechanism database that associates a function to be expressed, a mechanism that determines the expression of the function, and a physical parameter that contributes to the mechanism,
By performing a semantic analysis of the specification, a function corresponding to the specification is specified,
configured to be able to identify one or more mechanisms that determine the expression of the identified function by referring to the function expression mechanism database;
search system. In any one of claims 1 to 6,
The physical property search unit includes a physical property parameter relationship database, a graph generation unit, and a graph search unit,
The physical property parameter relationship database stores a plurality of parameter pairs of physical property parameters that are related to each other,
The graph generating unit is configured to generate a graph in which a plurality of physical property parameters included in the parameter pair are nodes and nodes corresponding to the parameter pair are edges,
The graph search unit is configured to search the graph based on given search conditions via the user interface and output search results,
The physical property parameters identified by the relevant factor identification unit are included in a plurality of physical property parameters stored in the physical property parameter relationship database.
search system. In claim 7,
In addition to the plurality of parameter pairs, the physical property parameter relationship database stores at least one factor that is different from the plurality of physical property parameters stored in the physical property parameter relationship database but is included in the plurality of physical property parameters. further storing an extended pair, which is a pair of an influencing factor that has a relationship with a physical property parameter and affects the physical property, and the physical property parameter;
The graph generation unit adds to the graph an edge between a node corresponding to the influencing factor constituting the extended pair and a node corresponding to the physical property parameter constituting the extended pair,
The physical property searching unit is configured to be able to search physical properties based on search conditions including influence factors, targeting the graph,
search system. A search method implemented by software operating on a computer having a storage device and comprising a physical property search step, a mechanism estimation step, and a related factor identification step,
The physical property searching step is configured to be able to supply a search condition including at least one physical property parameter, perform a physical property search based on the supplied search condition, and output a search result,
The mechanism estimating step estimates one or more candidate mechanisms for expressing the function implied by the specification from the input specification,
The related factor identifying step identifies the presence or absence of a physical property parameter that contributes to the mechanism and the physical property parameter for at least one of the one or more candidates for the mechanism ,
The physical property search step performs a physical property search with input of search conditions based on the specified physical property parameter , and outputs a search result.
exploration method. 10. The search method according to claim 9, further comprising a mechanism selection step,
The mechanism selection step outputs information related to the one or more mechanism candidates estimated in the mechanism estimation step to the user, and causes the user to select one mechanism in a plurality of cases.
exploration method. 11. The search method according to claim 10, further comprising a search history holding unit that holds a history of searches performed in the past in the storage device,
When a plurality of mechanisms are estimated in the mechanism estimation step, the mechanism selection step can refer to the search history storage unit and preferentially present a mechanism related to the history searched in the past. configured to
exploration method. In claim 11,
The search history holding unit is configured to be able to hold a search history for each user,
In the mechanism selection step, when a plurality of mechanisms are estimated in the mechanism estimation step, the search history holding unit is referred to and preferentially presents a mechanism related to a history of searches by the user in the past. configured to allow
exploration method. In claim 12,
In the mechanism selection step, when a plurality of mechanisms are estimated in the mechanism estimation step, the search history holding unit is referred to and preferentially presents a mechanism related to a history of searches by the user in the past. , configured to be able to present mechanisms related to histories searched by other users with priority to the next,
exploration method. 14. The search method according to any one of claims 9 to 13, wherein the search method includes a function expression mechanism database that associates a function to be expressed, a mechanism that determines the expression of the function, and a physical parameter that contributes to the mechanism. further prepared,
The mechanism estimation step identifies a function corresponding to the specification by performing a semantic analysis of the specification, refers to the function expression mechanism database, and identifies one or more mechanisms that determine the expression of the identified function. Identify,
exploration method. 15. In any one of claims 9 to 14, the search method comprises a physical property parameter relationship database,
The physical property search step includes a graph generation step and a graph search step,
The physical property parameter relationship database stores a plurality of parameter pairs of physical property parameters that are related to each other,
The graph generating step is configured to be able to generate a graph in which a plurality of physical property parameters included in the parameter pair are nodes and between nodes corresponding to the parameter pair are edges,
the graph search step searches the graph based on a given search condition and outputs a search result;
The physical property parameters that contribute to the mechanism that determines the expression of the function are included in a plurality of physical property parameters stored in the physical property parameter relationship database.
exploration method. In claim 15,
In addition to the plurality of parameter pairs, the physical property parameter relationship database stores at least one factor that is different from the plurality of physical property parameters stored in the physical property parameter relationship database but is included in the plurality of physical property parameters. further storing an extended pair, which is a pair of an influencing factor that has a relationship with a physical property parameter and affects the physical property, and the physical property parameter;
In the graph generation step , edges between nodes corresponding to the influencing factors constituting the extended pair and nodes corresponding to physical property parameters constituting the extended pair are added to the graph.
exploration method.

Images