JP7590645B2 - Output program and information processing device - Google Patents

Description

The present invention relates to an output program and an information processing apparatus.

Database systems may store linked data, which is sometimes called linked open data (LOD). Linked data has a data format suitable for information processing and is highly reusable. Linked data is generally described according to the Resource Description Framework (RDF).

Linked Data includes multiple entities, each identified by a Uniform Resource Identifier (URI). An entity may be called a resource or a node. An entity represents an entity such as a person, organization, or event. Linked Data expresses a fact as a triple of a subject, predicate, and object. The subject and object each correspond to an entity, and the predicate corresponds to a link between two entities. The predicate is sometimes called an attribute of the subject entity. Linked Data, which includes multiple entities and multiple links, may be represented as a graph. Linked Data is an example of graph data (knowledge graph) described by triples that indicate nodes (entities) and relationships (edges) between the nodes.

A database system may search Linked Data for entities that meet certain criteria. For example, the database system may execute a query written in a query language such as SPARQL (SPARQL Protocol and RDF Query Language). A Linked Data search may involve selecting a base entity and following one or more links that meet the search criteria specified by the query to extract other entities.

A geographic information system (GIS) has been proposed that combines different types of RDF data to provide geographic information services. The RDF data that can be combined includes topographical data, road data, administrative district data, place name data, building data, and commercial data. The proposed geographic information system, for example, detects the user's current location, searches for tourist attractions close to the current location, and presents the user with descriptions of the tourist attractions.

Noboru Shimizu, Kazue Mishima, Shohei Yamaguchi, Hiroshi Tsuda, and Terunobu Kume, "Linked Data and Geospatial Information," Information Processing, Vol. 52, No. 3, pp. 318-325, March 2011.

One entity included in graph data may have relationships with multiple other entities as attributes of the same type. From a grammatical perspective, such as entity type and attributes, it is difficult to distinguish between these multiple other entities when searching. As a result, all of these multiple other entities may be extracted for a given query.

However, there may be a hierarchical relationship between multiple entities that are connected as attributes of the same type. In such cases, the higher-level entities may not be necessary for the user, given the situation the user is in when searching. This creates the problem of data that is unnecessary for the user being extracted from the graph data.

For example, a knowledge database may contain graph data showing the relationships between problems, causes, and solutions. In the knowledge database, multiple causes may be connected to one problem, and multiple solutions may be connected to one problem. However, when the problem occurs, information on all causes or information on all solutions may not be useful to the user. If the user's work does not meet the lower-level requirements, the user may not be able to apply the higher-level causes or higher-level solutions to the work.

In one aspect, the present invention has an object to provide an output program and an information processing device that reduce data unnecessary for a user from a search result.

In one aspect, an output program is provided that causes a computer to execute a process of receiving information identifying a first piece of data, referencing a data group including a first plurality of pieces of data and information indicating a relationship between the first plurality of pieces of data, identifying a second plurality of pieces of data having the same attribute associated with the first piece of data, selecting a second piece of data associated with a level lower than the first level from among the second plurality of pieces of data based on a designation of a first level included in a plurality of levels having a hierarchical relationship, and outputting the second piece of data. In another aspect, an information processing device is provided that has a storage unit and a control unit.

On one hand, it reduces data that is unnecessary for users from search results.

FIG. 1 is a diagram illustrating an information processing apparatus according to a first embodiment. FIG. 11 is a diagram illustrating an information processing apparatus according to a second embodiment. FIG. 2 is a block diagram showing an example of hardware of an information processing device. FIG. 13 is a diagram illustrating an example of data representing a relationship between an event and a solution. FIG. 11 is a diagram showing a first example of a data file and a template. FIG. 13 is a diagram showing an example of data indicating a relationship between an event and a cause. FIG. 13 is a diagram showing a second example of a data file and a template. FIG. 1 illustrates an example of layering of entities representing solutions. FIG. 13 is a diagram showing a first example of adding layer information to a data file. FIG. 13 is a diagram illustrating an example of layering of entities representing causes. FIG. 13 is a diagram showing a second example of adding layer information to a data file. FIG. 1 illustrates an example of a database search and search result filter. FIG. 2 is a block diagram showing an example of software of the information processing device. 13 is a flowchart showing an example of a data registration procedure. 13 is a flowchart showing an example of a data search procedure. FIG. 13 is a diagram showing an example of a reference level selection screen.

The present embodiment will be described below with reference to the drawings. First, the first embodiment will be described. FIG. 1 is a diagram for explaining an information processing device of the first embodiment. The information processing device 10 of the first embodiment is a computer that searches for data from a database. The information processing device 10 may be a server device or a client device.

The information processing device 10 has a storage unit 11 and a control unit 12. The storage unit 11 may be a volatile semiconductor memory such as a random access memory (RAM). The storage unit 11 may also be a non-volatile storage such as a flash memory or a hard disk drive (HDD). The control unit 12 may be a processor such as a central processing unit (CPU), a graphics processing unit (GPU), or a digital signal processor (DSP). The control unit 12 may also include an electronic circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA). The processor executes a program stored in a memory such as a RAM. A set of multiple processors may be called a "multiprocessor" or simply a "processor."

The storage unit 11 stores the data group 13. The data group 13 is stored in a database. The data group 13 includes a first plurality of data and information indicating the relationship between the first plurality of data. The data group 13 may be graph data including a plurality of entities and a plurality of links indicating the relationship between the plurality of entities. The entities may be called resources or nodes. The entities may represent real entities such as people, organizations, and events. The links may represent that one entity corresponds to a specific attribute from the perspective of the other entity. The data group 13 may be expressed as a directed graph. The data group 13 may also be lined data or RDF data.

Level information is associated with at least some of the first plurality of data included in data group 13. The level information may be included in data group 13, or may be stored outside data group 13. Level information associated with certain data indicates one level among multiple levels having a hierarchical relationship. For example, when data group 13 is registered in a database, the level information is also registered.

As an example, data group 13 includes data 14a, 14b, 14c, and 14d. Data 14a, 14b, and 14c are associated with data 14d by the same attribute. Level information 15a is associated with data 14a. Level information 15b is associated with data 14b. Level information 15c is associated with data 14c. Level information 15a indicates that data 14a is at level 1. Level information 15b indicates that data 14b is at level 2, which is higher than data 14a. Level information 15c indicates that data 14c is at level 3, which is higher than data 14b.

Data group 13 may be a data group indicating an event and a cause or solution of the event. For example, data 14a is data indicating an event, and data 14a, 14b, and 14c are data indicating the cause or solution of the event. In this way, there may be multiple causes or multiple solutions for one event. Also, for example, the hierarchical relationship indicated by level information 15a, 15b, and 15c is a relationship in which the higher the level, the more prerequisites there are for applying the cause or solution. In a situation where the cause of a lower level has not been resolved, the user may not be able to consider the cause of a higher level. Also, in a situation where the solution of a lower level has not been implemented, the user may not be able to implement the solution of a higher level.

The control unit 12 searches for data from the data group 13. In the data search, the control unit 12 accepts information that identifies the first data. The first data is data that indicates the starting point of the search. The first data may be identified by an identifier such as a URI. The information that identifies the first data may be input by a user. Upon accepting the information that identifies the first data, the control unit 12 refers to the data group 13 stored in the memory unit 11, and identifies a second plurality of data that have the same attribute associated with the first data.

In identifying the second plurality of data, the control unit 12 may use a template indicating search conditions including attribute specification to search the data group 13 for data that satisfies the search conditions, starting from the first data. The control unit 12 may generate a search query from the information identifying the first data and the template, and execute the generated search query on the data group 13. As an example, the control unit 12 accepts information identifying data 14d, and identifies data 14a, 14b, and 14c that are associated with data 14d and have the same attribute.

The control unit 12 accepts the designation of a first level from among the multiple levels. The designation of the first level may be input by the user. The designation of the first level may also be performed by acquiring a value sensed by a sensor regarding the user. For example, the first level may be designated based on a temperature measurement value acquired from a thermometer. The control unit 12 may accept the designation of the first level before identifying the second plurality of data, or may accept the designation after identifying the second plurality of data. The control unit 12 may also cause the first level to be designated when the identified second plurality of data includes data associated with level information.

Based on the designation of the first level, the control unit 12 selects, from the second plurality of data, second data associated with a level equal to or lower than the designated first level. Then, the control unit 12 outputs the selected second data. The control unit 12 may display the second data on a display device connected to the information processing device 10. The control unit 12 may also transmit the second data to another information processing device. As an example, the control unit 12 accepts a designation of level 2 from levels 1 to 3. Then, the control unit 12 selects, from the data 14a, 14b, and 14c, data 14a and 14b associated with level information equal to or lower than level 2. At this time, the control unit 12 excludes data 14c associated with level information indicating level 3.

In this way, the information processing device 10 of the first embodiment identifies a second plurality of data having the same attribute relationship as the first data. Then, the information processing device 10 selects and outputs the second plurality of data having a level equal to or lower than the specified level. This allows the information processing device 10 to narrow down the data by focusing on the hierarchical relationship between the data, even when the data included in the search results cannot be sufficiently narrowed down by attributes alone. Therefore, data that is too high-level considering the user's situation at the time of the search is excluded from the search results, and data that is unnecessary for the user is reduced from the search results. For example, causes and solutions that are too high-level to be applied to the user's current work are reduced from the search results.

Next, a second embodiment will be described. Descriptions of contents similar to those of the first embodiment described above may be omitted. FIG. 2 is a diagram for explaining an information processing device of the second embodiment. The information processing device 20 of the second embodiment is a computer that registers data in a database. The information processing device 20 may be a server device or a client device. The information processing device 20 may be the same as the information processing device 10 of the first embodiment.

The information processing device 20 has a storage unit 21 and a control unit 22. The storage unit 21 corresponds to the storage unit 11 of the first embodiment. The control unit 22 corresponds to the control unit 12 of the first embodiment. The storage unit 21 stores a data group 23. The data group 23 corresponds to the data group 13 of the first embodiment. However, level information has not yet been associated with any of the first plurality of data included in the data group 23. As an example, the data group 23 includes data 24a, 24b, 24c, and 24d. The data 24a, 24b, and 24c are associated with the data 24d by the same attribute.

When the control unit 22 receives the data group 23, it selects, from the first plurality of data included in the data group 23, a second plurality of data having the same attribute associated with the first data. When a template indicating search conditions including an attribute exists, the control unit 22 may select, as the second plurality of data, two or more pieces of data that satisfy the search conditions indicated by the template and cannot be separately searched using the search conditions. As an example, the control unit 22 selects data 24a, 24b, and 24c from the data group 23.

The control unit 22 associates level information indicating one specified level with each of the selected second multiple data, based on the designation of one level among multiple levels having a hierarchical relationship. The designation of one level may be input by the user. Furthermore, the control unit 22 may present information indicating multiple levels as options to the user. The control unit 22 then registers the level information in the database in association with the data group 23. The control unit 22 may register the level information at the same time as registering the data group 23, or may additionally register the level information after the data group 23 is registered. Furthermore, the level information may be incorporated into the data group 23, or may be stored outside the data group 23.

As an example, the control unit 22 associates level information 25a with data 24a, level information 25b with data 24b, and level information 25c with data 24c. Level information 25a indicates that data 24a is at level 1. Level information 25b indicates that data 24b is at level 2, which is higher than data 24a. Level information 25c indicates that data 24c is at level 3, which is higher than data 24b.

In this way, the information processing device 20 of the second embodiment selects, from the data group 23, a second plurality of data that have the same attribute relationship as a certain data. Then, the information processing device 20 associates level information with each of the selected second plurality of data. As a result, the data group 23 in which at least some of the data is associated with level information is registered in the database. Therefore, the information processing device 20 can create a database that can narrow down data by focusing on the hierarchical relationships between data, even when the data cannot be sufficiently narrowed down by attributes alone. As a result, an efficient database is created that can reduce data that is unnecessary for the user when searching from the search results.

In the above description of the second embodiment, the data group 23 to which level information is assigned includes three relationships indicating a data pair, but it may also include only one relationship indicating a data pair. In this case, the control unit 22 may execute the following process. When the control unit 22 receives data including first data, second data, and information indicating the relationship between the first data and the second data, it selects at least one of the first data and the second data. Based on the designation of the first level among multiple levels having a hierarchical relationship, the control unit 22 associates level information indicating the first level with the one piece of data and registers it in a database including multiple data and information indicating the relationship between the multiple data.

For example, when the control unit 22 receives data including data 24a, data 24d, and information indicating the relationship between data 24a and 24d, the control unit 22 selects data 24a. Based on the designation of one level among a plurality of levels having a hierarchical relationship, the control unit 22 associates level information 25a indicating the one level with data 24a and registers the level information in the database.

Next, a third embodiment will be described. The information processing device 100 of the third embodiment is a database management device that registers and searches for data in response to a user request. The information processing device 100 may be a server device or a client device. Note that, although the information processing device 100 executes both data registration and data search in the third embodiment, different information processing devices may execute data registration and data search. Also, a single database may be managed by multiple information processing devices in a distributed manner.

Figure 3 is a block diagram showing an example of hardware of an information processing device. The information processing device 100 has a CPU 101, a RAM 102, a HDD 103, an image interface 104, an input interface 105, a media reader 106, and a communication interface 107. The CPU 101, the RAM 102, the HDD 103, the image interface 104, the input interface 105, the media reader 106, and the communication interface 107 are connected to a bus. The CPU 101 corresponds to the control unit 12 of the first embodiment and the control unit 22 of the second embodiment. The RAM 102 or the HDD 103 corresponds to the memory unit 11 of the first embodiment and the memory unit 21 of the second embodiment.

The CPU 101 is a processor that executes program instructions. The CPU 101 loads at least a portion of the programs and data stored in the HDD 103 into the RAM 102 and executes the programs. The information processing device 100 may have multiple processors. A collection of processors may be called a multiprocessor or simply a "processor."

RAM 102 is a volatile semiconductor memory that temporarily stores programs executed by CPU 101 and data used by CPU 101 for calculations. Information processing device 100 may have a type of memory other than RAM.

The HDD 103 is a non-volatile storage that stores software programs and data. The software may include an operating system (OS), middleware, and application software. The information processing device 100 may also have other types of non-volatile storage, such as flash memory.

The image interface 104 outputs images to a display device 111 connected to the information processing device 100 in accordance with instructions from the CPU 101. The display device 111 may be a CRT (Cathode Ray Tube) display, a liquid crystal display (LCD: Liquid Crystal Display), or an organic EL (OEL: Organic Electro-Luminescence) display. Other types of output devices, such as a printer, may also be connected to the information processing device 100.

The input interface 105 receives an input signal from an input device 112 connected to the information processing device 100. The input device 112 may be a mouse, a keyboard, or a touchpad. Multiple input devices may be connected to the information processing device 100.

The media reader 106 is a reading device that reads the program and data recorded on the recording medium 113. The recording medium 113 may be a magnetic disk, an optical disk, or a semiconductor memory. The magnetic disk may include a flexible disk (FD) and a HDD. The optical disk may include a compact disc (CD) and a digital versatile disc (DVD). For example, the media reader 106 copies the program and data read from the recording medium 113 to a storage such as the HDD 103. The recording medium 113 may be a portable recording medium. The recording medium 113 may be used for distributing the program and data. The recording medium 113 and the HDD 103 may be called computer-readable recording media.

The communication interface 107 is connected to the network 114. The communication interface 107 communicates with other information processing devices via the network 114. The communication interface 107 may be a wired communication interface connected to a wired communication device such as a switch or a router. The communication interface 107 may also be a wireless communication interface connected to a wireless communication device such as a base station or an access point.

Next, the structure of the database will be explained. The information processing device 100 holds a knowledge database that shows the relationships between events, causes, and solutions. The data contained in the knowledge database is Linked Data. Linked Data is described according to RDF.

The knowledge database includes multiple entities that indicate events, causes, and solutions, and multiple links that indicate relationships between the multiple entities. The entities are identified by URIs. The links have attributes that indicate the type of relationship between one entity and the other entity. The type of relationship is identified by the URI. The knowledge database includes as a relationship type a cause relationship that indicates that one of the causes of one entity is the other entity. The knowledge database also includes as a relationship type a solution relationship that indicates that one of the solutions for one entity is the other entity.

Linked Data expresses a fact using a triple with a subject, predicate, and object. Linked Data is a collection of triples. The subject and object each correspond to an entity. The predicate corresponds to a link. Linked Data is sometimes represented as a directed graph with nodes that correspond to entities and edges that correspond to links. There may be two or more causes for one event, and two or more solutions.

Figure 4 is a diagram showing an example of data that indicates the relationship between an event and a solution. As an example, the Linked Data includes entities 31 to 39. Entity 31 indicates an event. Entities 32 to 39 indicate solutions. Entities 32 to 39 are each connected to entity 31 by links that indicate a solution relationship. Therefore, the solutions indicated by entities 32 to 39 are candidate solutions for the event indicated by entity 31.

Entity 31 represents the event "hunger". Entity 32 represents the solution "find edible food". Entity 33 represents the solution "fill your stomach with water". Entity 34 represents the solution "wash". Entity 35 represents the solution "check the expiration date". Entity 36 represents the solution "find a restaurant that has your favorite food". Entity 37 represents the solution "invite a friend to dinner". Entity 38 represents the solution "find trendy foods". Entity 39 represents the solution "eat recommended dishes".

Figure 5 is a diagram showing a first example of a data file and a template. Data file 131 contains RDF data representing the Linked Data of Figure 4. Data file 131 contains eight triples indicating relationships between entity 31 and entities 32 to 39. Template 132 is used to search for a desired entity from the Linked Data of data file 131. Template 132 contains triples indicating search conditions. The subject of the triple is a parameter indicating the base point of the search. The predicate of the triple is a URI indicating the solution relationship. The object of the triple is a variable indicating the entity to be extracted.

The information processing device 100 receives the URI of the base entity. The information processing device 100 then generates a query 133 by substituting the received URI into the parameters of the template 132. The query 133 is a search query indicating a search for entities that have a solution relationship with the entity of interest. The information processing device 100 executes the query 133 to extract entities that satisfy the search conditions from the Linked Data. When the URI of the entity 31 is specified, the information processing device 100 extracts entities 32 to 39 from the Linked Data. Note that the URI of the base entity may be embedded in the template 132.

In this way, multiple entities indicating potential solutions may be associated with an entity indicating one event. Even if a base entity and attributes are specified, all of these multiple entities satisfy the search conditions. Therefore, the information processing device 100 outputs search results that include all of these multiple entities. However, multiple entities included in the search results do not necessarily indicate the same level of solution. Therefore, all entities included in the search results are not necessarily useful to the user.

For example, entities 32 and 33 provide suitable solutions to the user's situation of "constant hunger." Entities 34 and 35 provide suitable solutions to the user's situation of "persistent hunger." However, entities 34 and 35 do not provide suitable solutions to a user whose "constant hunger" situation has not been resolved. Thus, entities 34 and 35 provide higher-level solutions than entities 32 and 33.

Similarly, entities 36 and 37 show suitable solutions for the user's situation of "usually satisfying appetite." However, entities 36 and 37 do not show suitable solutions for a user whose situation of "prolonged hunger" has not been resolved. Thus, entities 36 and 37 show a higher level of solution than entities 34 and 35. Entities 38 and 39 show suitable solutions for the user's situation of "usually eating favorite food." However, entities 38 and 39 do not show suitable solutions for a user whose situation is at the level of "usually satisfying appetite." Thus, entities 38 and 39 show a higher level of solution than entities 36 and 37.

In this way, search results generated using template 132 may contain multiple entities that represent solutions at different levels. On the other hand, entities that represent solutions that are too high level for the user's current situation are unnecessary information for the user.

Figure 6 is a diagram showing an example of data that indicates the relationship between an event and its cause. As another example, Linked Data includes entities 41 to 47. Entity 41 indicates an event. Entities 42 to 47 indicate causes. Entities 42 to 47 are each connected to entity 41 by links that indicate causal relationships. Therefore, the causes indicated by entities 42 to 47 are candidate causes for the event indicated by entity 41.

Entity 41 indicates the event of "decline in model accuracy." Entity 42 indicates the cause of "lack of data for some organizations." Entity 43 indicates the cause of "variation in expression." Entity 44 indicates the cause of "data preparation errors." Entity 45 indicates the cause of "failure to respond to organizational changes." Entity 46 indicates the cause of "changes in trends." Entity 47 indicates the cause of "release of new products."

Figure 7 is a diagram showing a second example of a data file and a template. Data file 135 contains RDF data representing the Linked Data of Figure 6. Data file 135 contains six triples indicating the relationships between entity 41 and entities 42 to 47. Template 136 is used to search for a desired entity from the Linked Data of data file 135. Template 136 contains triples indicating search conditions. The subject of the triple is a parameter indicating the base point of the search. The predicate of the triple is a URI indicating a causal relationship. The object of the triple is a variable indicating the entity to be extracted.

The information processing device 100 receives the URI of the base entity. The information processing device 100 then generates a query 137 by substituting the received URI into the parameters of the template 136. The query 137 is a search query indicating a search for an entity that has a causal relationship with the entity of interest. The information processing device 100 executes the query 137 to extract entities that satisfy the search conditions from the Linked Data. When the URI of the entity 41 is specified, the information processing device 100 extracts entities 42 to 47 from the Linked Data. Note that the URI of the base entity may be embedded in the template 136.

For simplicity, the Linked Data in Figures 4 to 7 describe a simple relationship between one event and multiple solutions or multiple causes. However, Linked Data can describe more complex relationships between various events, causes, and solutions.

In this way, multiple entities indicating possible causes may be associated with an entity indicating one event. Even if a base entity and attribute are specified, all of these multiple entities satisfy the search conditions. Therefore, the information processing device 100 outputs search results that include all of these multiple entities. However, multiple entities included in the search results do not necessarily indicate causes at the same level. Therefore, all entities included in the search results may not be useful to the user.

For example, entities 42 and 43 indicate likely causes for a business situation called "data deficiency." Data deficiency indicates a situation where not enough data is collected for machine learning. Entities 44 and 45 indicate likely causes for a business situation called "human problem." Human problem indicates a situation where the quality of collected data is low due to operational deficiencies in data collection. However, entities 44 and 45 do not indicate likely causes for business where the "data deficiency" has not been resolved. Thus, entities 44 and 45 indicate a higher level of cause than entities 42 and 43.

Similarly, entities 46 and 47 indicate suitable solutions to a business situation called "concept drift." Concept drift indicates a situation in which the characteristics of collected data have changed since the model was generated due to social and economic reasons. However, entities 46 and 47 do not indicate likely causes for business where the "human problem" has not been resolved. Thus, entities 46 and 47 indicate a higher level of cause than entities 44 and 45.

In this way, search results generated using template 136 may include multiple entities indicating causes at different levels. On the other hand, entities indicating causes that are too high in level given the user's current situation are unnecessary information for the user. Therefore, information processing device 100 allocates and places some of the entities included in Linked Data into multiple layers at different levels. Information processing device 100 excludes entities placed in layers at a level higher than the reference level indicating the user's current situation from the search results.

Figure 8 is a diagram showing an example of layering of entities representing solutions. Entities 32 to 39 included in the Linked Data in Figure 4 are arranged in layers 51 to 54. Layer 51 indicates the user's situation of "always hungry". The level of layer 51 is level 1. Layer 52 indicates the user's situation of "prolonged hunger". The level of layer 52 is level 2, higher than layer 51. Layer 53 indicates the user's situation of "usually satisfied appetite". The level of layer 53 is level 3, higher than layer 52. Layer 54 indicates the user's situation of "usually eating favorite food". The level of layer 54 is level 4, higher than layer 53.

Entities 32 and 33 are placed in layer 51. Therefore, entities 32 and 33 are associated with layer information indicating level 1. Entities 34 and 35 are placed in layer 52. Therefore, entities 34 and 35 are associated with layer information indicating level 2. Entities 36 and 37 are placed in layer 53. Therefore, entities 36 and 37 are associated with layer information indicating level 3. Entities 38 and 39 are placed in layer 54. Therefore, entities 38 and 39 are associated with layer information indicating level 4.

Figure 9 is a diagram showing a first example of adding layer information to a data file. Data file 134 is a data file in which layer information has been added to data file 131 shown in Figure 5. A layer group named "I'm hungry" is defined in data file 134. This layer group includes layers 51 to 54, which correspond to levels 1 to 4. Entities 32 to 39 are each assigned a label indicating that the layer group name is "I'm hungry" and that the level is one of 1 to 4. In Linked Data, this label is defined as an attribute of each entity.

Figure 10 is a diagram showing an example of layering of entities that represent causes. Entities 42 to 47 included in the Linked Data in Figure 6 are arranged in layers 55 to 57. Layer 55 indicates the business situation of "data deficiency." The level of layer 55 is level 1. Layer 56 indicates the business situation of "human-made problem." The level of layer 56 is level 2, which is higher than layer 55. Layer 57 indicates the business situation of "concept drift." The level of layer 57 is level 3, which is higher than layer 56.

Entities 42 and 43 are placed in layer 55. Therefore, entities 42 and 43 are associated with layer information indicating level 1. Entities 44 and 45 are placed in layer 56. Therefore, entities 44 and 45 are associated with layer information indicating level 2. Entities 46 and 47 are placed in layer 57. Therefore, entities 46 and 47 are associated with layer information indicating level 3.

Figure 11 is a diagram showing a second example of adding layer information to a data file. Data file 138 is a data file in which layer information has been added to data file 135 shown in Figure 7. A layer group named "reduced accuracy" is defined in data file 138. This layer group includes layers 55 to 57, which correspond to levels 1 to 3. Entities 42 to 47 are each assigned a label indicating that the layer group name is "reduced accuracy" and that the level is one of 1 to 3. In Linked Data, this label is defined as an attribute of each entity.

In the examples of Figures 9 and 11, the information processing device 100 inserts layer information into the Linked Data. Alternatively, the information processing device 100 may store layer information outside the Linked Data. Also, in the examples of Figures 8 and 10, the information processing device 100 defines one layer group for the Linked Data. Alternatively, the information processing device 100 may define multiple layer groups corresponding to different search perspectives. For example, the information processing device 100 may define multiple layer groups corresponding to different templates.

In this case, the same entity may belong to different layers. Also, different levels may be assigned to the same entity. For example, a solution may be a level 1 solution for one event, but a level 2 solution for another event. Also, a cause may be a level 1 cause for one event, but a level 2 cause for another event.

Figure 12 is a diagram showing an example of a database search and a search result filter. The Linked Data shown in Figure 12 is more complex than that shown in Figures 4 and 6. This Linked Data includes 31 entities, entities #1 to #31. Entities #15 to #20 belong to layer 61, level 1. Entities #21 to #26 belong to layer 62, level 2. Entities #27 to #31 belong to layer 63, level 3.

For example, the information processing device 100 accepts the specification of entity #1 as the base entity and searches for related entities according to a certain template. The information processing device 100 then extracts a path including entities #5 and #15, a path including entities #7 and #17, a path including entities #8 and #21, a path including entities #9 and #22, and a path including entities #11 and #27.

At this time, the information processing device 100 may accept level 1 as the reference level from the user. In that case, the information processing device 100 excludes from the search results, of the five paths, the paths that include entities #8 and #21, the paths that include entities #9 and #22, and the paths that include entities #11 and #27. As a result, two paths remain. The information processing device 100 may also accept level 2 as the reference level from the user. In that case, the information processing device 100 excludes from the search results, of the five paths, the paths that include entities #11 and #27. As a result, four paths remain. In this way, paths that include entities at a level exceeding the reference level are excluded from the search results.

Next, the functions and processing procedures of the information processing device 100 will be described. FIG. 13 is a block diagram showing an example of software of the information processing device. The information processing device 100 has a data storage unit 121, a template storage unit 122, a query execution unit 123, a data filter unit 124, and a data registration unit 125. The data storage unit 121 and the template storage unit 122 correspond to, for example, the storage areas of the RAM 102 or the HDD 103. The query execution unit 123, the data filter unit 124, and the data registration unit 125 correspond to, for example, the programs executed by the CPU 101.

The data storage unit 121 stores Linked Data that includes layer information. For example, the data storage unit 121 stores the above-mentioned data files 134 and 138. However, the information processing device 100 may have another storage unit for storing layer information. In that case, the data storage unit 121 stores Linked Data that does not include layer information. The template storage unit 122 stores templates for generating search queries.

The query execution unit 123 receives a search request from a user. The search request includes a key URI indicating a base entity and information for specifying a template to be used in the search. Upon receiving a search request, the query execution unit 123 generates a search query from the key URI and the template, and executes the search query on the data storage unit 121 to generate search results. The search results include the URIs of entities that satisfy the search conditions indicated by the search query.

The data filter unit 124 filters the search results generated by the query execution unit 123 and outputs the modified search results. In filtering, the data filter unit 124 extracts entities to which layer information has been assigned from the search results. If an entity to which layer information has been assigned is present, the data filter unit 124 presents the user with selection information for selecting a reference level. Then, the data filter unit 124 excludes entities at a level higher than the reference level from the search results.

The selection information presented to the user may include a layer group name and information indicating a plurality of levels that the user can select as a reference level. The selection information may also include a question corresponding to the layer group, which is related to the user's current situation. The question is created by an administrator of the information processing device 100 before the data search. In this case, the information processing device 100 may estimate a reference level indicating the user's current situation based on the user's answer to the question. In outputting the search results, the data filter unit 124 may display the modified search results on the display device 111, or may transmit them to another information processing device.

The data registration unit 125 accepts a data file from a user. The data registration unit 125 adds the Linked Data indicated by the accepted data file to the data storage unit 121. At this time, the data registration unit 125 extracts a template related to the added Linked Data from the template storage unit 122. The related template is a template indicating search conditions from which at least one entity included in the added Linked Data can be extracted.

If a related template exists, the data registration unit 125 searches the additional Linked Data for entities to which layer information should be assigned, based on the related template. The entities to which layer information should be assigned are two or more entities that can be reached by following the same type of link from the same entity. These two or more entities are searched for based on the search criteria indicated by the related template.

The template may include a variable indicating an entity that is expected to have layer information attached, and may specify as a search condition that the entity has an attribute indicating a layer group name (exp:layerName) or an attribute indicating a level (exp:layerLevel). For example, a search condition indicating that the variable corresponding to the object has a layer group name attribute may be added to the above-mentioned templates 132 and 136. In that case, the data registration unit 125 may search for an entity that matches the variable from the additional Linked Data.

If there is an entity to which layer information should be assigned, the data registration unit 125 presents the URI of the corresponding entity to the user and accepts the level designation of the entity from the user. The data registration unit 125 generates layer information including the designated level and assigns it to the entity. At this time, the data registration unit 125 may present the URIs of multiple entities to the user all at once, or may accept the level designation for the multiple entities all at once from the user. The data registration unit 125 may also present the URIs of the entities and accept the level designation for each entity one by one.

The data storage unit 121 may already define one or more layer groups and selectable levels for each layer group. The data registration unit 125 may then present information on the selectable layer groups and selectable levels to the user along with the entity information. The presentation of information to the user may be by displaying the information on the display device 111, or by transmitting the information to a client device used by the user.

Figure 14 is a flowchart showing an example of a procedure for registering data. (S10) The data registration unit 125 accepts a data file including new data that is Linked Data. The data registration unit 125 registers the new data included in the data file in a database. (S11) The data registration unit 125 extracts templates related to the new data from a group of existing templates. A related template is a template that indicates search conditions that may result in the extraction of at least one entity included in the new data.

(S12) The data registration unit 125 determines whether a template related to the new data exists. If a related template exists, the process proceeds to step S13, and if a related template does not exist, the data registration process ends.

(S13) The data registration unit 125 searches the new data for an entity to which layer information is to be assigned, based on the associated template. For example, the data registration unit 125 searches the new data for two or more entities that satisfy the search conditions indicated by the template and cannot be searched separately from each other using the search conditions. If the template includes a variable with a predefined condition that the entity is one that has a layer group name attribute or a level attribute, the data registration unit 125 may search for an entity that matches the variable.

(S14) The data registration unit 125 presents the list of entities extracted in step S13 to the user. The list includes the URI of each entity. The list is presented to the user, for example, by displaying the list of entities on the display device 111. (S15) In response to the presentation of the list in step S14, the data registration unit 125 accepts from the user a designation of a layer group name and level for each entity included in the list. The layer group name and level are input, for example, using the input device 112. The data registration unit 125 assigns the attributes of layer group name creation and level to each entity included in the list.

Figure 15 is a flowchart showing an example of a data search procedure. (S20) The query execution unit 123 accepts a key URI and a template specification from the user. The key URI and the template specification are input using, for example, the input device 112. (S21) The query execution unit 123 generates a query by substituting the key URI for a parameter included in the template, and executes the query to search the database.

(S22) The data filter unit 124 extracts entities to which layer information has been assigned from the search results obtained in step S21. Entities to which layer information has been assigned are entities that have a layer group name attribute and a level attribute. (S23) The data filter unit 124 determines whether an entity to which layer information has been assigned exists. If an entity to which layer information has been assigned exists, the process proceeds to step S24; if an entity to which layer information has been assigned does not exist, the process proceeds to step S26.

(S24) The data filter unit 124 presents the user with selection information for selecting a reference level, which is one of the multiple levels. The presentation of the selection information is, for example, displaying the selection information on the display device 111. The multiple levels that are candidates for the reference level are defined in the database in association with the layer group names of the entities extracted in step S22. The selection information is, for example, a list of selectable levels.

(S25) In response to the presentation of the selection information in step S24, the data filter unit 124 receives information for determining the reference level from the user. The information for determining the reference level is input, for example, using the input device 112. The data filter unit 124 determines the reference level according to the response from the user. The information for determining the reference level is, for example, information indicating a level selected by the user from among multiple levels.

The data filter unit 124 excludes from the search results those entities in step S22 whose levels exceed the reference level. The data filter unit 124 does not need to exclude entities whose levels are not specified from the search results. Thus, the corrected search results include entities whose levels are equal to or below the reference level and entities whose levels are not specified. (S26) The data filter unit 124 outputs the search results. However, if steps S24 and S25 are executed, the data filter unit 124 outputs the corrected search results instead of the search results of step S21. For example, the data filter unit 124 displays the search results on the display device 111.

Figure 16 is a diagram showing an example of a reference level selection screen. The reference level selection screen 141 is displayed on the display device 111 in step S24 above. The reference level selection screen 141 includes a display field for layer group names and a selection field for a reference level. As an example, the reference level selection field is a selection field that allows the user to select one of levels 1 to 4. The information for each level includes a brief description that indicates the expected situation of the user. For example, the user selects one of levels 1 to 4 using the input device 112. The level that the user should select is the highest level among the levels having a description that matches the user's current situation. This specifies the reference level.

In this way, the information processing device 100 of the third embodiment manages a knowledge database that shows the relationships between events, causes, and solutions, and provides a search service for the knowledge database. This promotes the sharing of know-how between users. Furthermore, the information processing device 100 holds the data of the knowledge database as Linked Data. This facilitates linking with other databases, and promotes the use of the knowledge database. Furthermore, the information processing device 100 generates a query based on the specification of a base entity and a template, and executes the query to search the Linked Data. Thus, the information processing device 100 can efficiently perform data searches by defining the types of frequently performed data searches in templates.

When registering Linked Data, the information processing device 100 detects multiple entities associated with the same entity through the same type of attribute, and allocates and places the detected multiple entities into multiple layers of different levels through user queries. As a result, each of the detected multiple entities is given layer information indicating the layer level. When searching Linked Data, the information processing device 100 excludes entities with a level higher than a reference level from the search results through user queries.

Therefore, entities that indicate causes or solutions that are too high level from the perspective of the user's current situation are excluded from the search results. In this way, the information processing device 100 can narrow down the search results by focusing on the hierarchical relationships between entities, even if the search results cannot be sufficiently narrowed down based on the attribute types. As a result, information that is unnecessary for the user is reduced from the search results, improving the usefulness of Linked Data search results.

10, 20 Information processing device 11, 21 Storage unit 12, 22 Control unit 13, 23 Data group 14a, 14b, 14c, 14d, 24a, 24b, 24c, 24d Data 15a, 15b, 15c, 25a, 25b, 25c Level information

Claims (5)

a data group including a plurality of data, the relationships between the plurality of data being represented by a plurality of links each having an attribute, and level information indicating any one of a plurality of levels having a hierarchical relationship being assigned to at least some of the plurality of data, using a query specifying first data indicating a base point of a search among the plurality of data and a search condition for the attribute, to search for a plurality of second data which satisfy the search condition among the plurality of data and are each associated with the first data by links having the same attribute;
accepting a designation of any one of the plurality of levels, and selecting, from the plurality of second data, second data to which level information indicating a level equal to or lower than the designated level is assigned ;
outputting the selected second data;
An output program that causes a computer to execute a process. the selecting process includes a process of accepting designation of one of the levels when the second data to which the level information is assigned is included in the plurality of second data.
2. The output program according to claim 1 . the searching process includes a process of searching the plurality of second data items that satisfy the search condition, based on the first data item, by using a template indicating the search condition;
2. The output program according to claim 1 . The first data is an entity indicating an event, and the plurality of second data are entities indicating a cause or a solution of the event, respectively, and the hierarchical relationship indicates a relationship in which a higher level has more prerequisites for applying the cause or the solution.
2. The output program according to claim 1 . a storage unit that stores a data group including a plurality of data, the plurality of data being represented by a plurality of links each having an attribute, and level information indicating one of a plurality of levels having a hierarchical relationship being assigned to at least some of the plurality of data ;
a control unit that searches the data group for a plurality of second data items that satisfy the search conditions among the plurality of data items and are associated with the first data items by links having the same attribute, using a query that specifies first data items indicating a search base point among the plurality of data items and search conditions for the attribute, accepts a specification of any one of the plurality of levels, selects, from the plurality of second data items, second data items that are assigned level information indicating a level equal to or lower than the specified level , and outputs the selected second data items;
13. An information processing device comprising:

Images