JP6880580B2 - Problem estimation device, problem estimation method and program - Google Patents

Description

The present invention relates to a task estimation device, a task estimation method and a program.

When developing solutions and products, it is preferable to grasp the needs of customers such as these solutions and products. In order to grasp the customer's request, it may be necessary to estimate a hypothesis about the problem that the customer has, verify the hypothesis, and grasp its validity.

Patent Document 1 describes a business problem analyzer and the like that quantitatively visualize business processes and support problem extraction and evaluation work in CRM (Customer Evaluation Management) consulting business. According to the apparatus described in Patent Document 1, the current process can be visualized and the business process to be improved can be evaluated based on a certain quantified value. Then, based on this quantified value, the current problem extraction, factor analysis, evaluation work, etc., which were determined by the experience and intuition of the consultant, are processed constantly and reliably, and the CRM consulting business is performed. It is possible to improve the quality of.

Further, Patent Document 2 describes a human error countermeasure candidate presenting device and the like that make it easy for a person without skills to formulate countermeasures. The device or the like described in Patent Document 2 extracts an object and a predicate from the contents of past measures, and presents the combination of the obtained object and the predicate as a new measure for preventing the recurrence of human error. ..

Further, Patent Document 3 describes a design support method and the like for determining the optimum improvement measures for the assembly time of the product and the assembly defect generated in the assembly process.

Japanese Unexamined Patent Publication No. 2005-071113 Japanese Unexamined Patent Publication No. 2012-128637 JP-A-2007-280366

When testing a hypothesis about a problem, it is necessary to understand the problem as a premise. However, it is generally difficult to comprehensively grasp an issue, for example, it is necessary to collect a lot of information about the industry related to the issue. Further, in the techniques described in the above-mentioned patent documents, sufficient consideration is not given to comprehensively grasping the problems.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a problem estimation device or the like capable of comprehensively grasping the problems.

The problem estimation device according to one aspect of the present invention includes a problem estimation means that extracts an expression indicating a measure and an expression indicating a relationship between the measure and the problem from a plurality of documents and estimates the problem based on the extracted expression.

In addition, the problem estimation method in one aspect of the present invention extracts an expression indicating a measure and an expression indicating a relationship between the measure and the problem from a plurality of documents, and estimates the problem based on the extracted expression.

Further, the program according to one aspect of the present invention extracts an expression indicating a measure and an expression indicating the relationship between the measure and the problem from a plurality of documents on a computer, and executes a process of estimating a problem based on the extracted expression. Let me.

According to the present invention, it is possible to provide a problem estimation device or the like that enables a comprehensive grasp of a problem.

It is a figure which shows the structure of the problem estimation apparatus in 1st Embodiment of this invention. It is a figure which shows the other structure of the problem estimation apparatus in 1st Embodiment of this invention. It is a figure about an example of the expression which shows a measure. It is a figure about an example of the expression which shows the relationship between a measure and an issue. It is a flowchart which shows the operation example of the problem estimation apparatus in 1st Embodiment of this invention. It is a figure which shows the structure of the problem estimation apparatus in the 2nd Embodiment of this invention. It is a figure which shows the other structure of the problem estimation apparatus in the 2nd Embodiment of this invention. It is a figure about an example of the expression which shows a problem. It is a figure which shows the relationship of the measure or the problem estimated in the problem estimation apparatus in the 2nd Embodiment of this invention. It is a flowchart which shows the operation example of the problem estimation apparatus in 2nd Embodiment of this invention. It is a figure which shows the structure of the problem estimation apparatus in 3rd Embodiment of this invention. It is a figure which shows the other structure of the problem estimation apparatus in 3rd Embodiment of this invention. It is a figure which shows the other structure of the problem estimation apparatus in 3rd Embodiment of this invention. It is a figure which shows an example concerning the novelty of the document which shows the problem in 2nd Embodiment of this invention. It is a figure which shows the example of the logical relation of the subject in the 2nd Embodiment of this invention. It is a flowchart which shows the operation example of the problem estimation apparatus in 3rd Embodiment of this invention. It is a figure which shows an example of the information processing apparatus which realizes the problem estimation apparatus and the like in each embodiment of this invention.

Each embodiment of the present invention will be described with reference to the accompanying drawings. In each embodiment of the present invention, each component of each device (system) represents a block of functional units. A part or all of each component of each device (system) is realized by an arbitrary combination of the information processing device 500 and the program as shown in FIG. 17, for example. The information processing device 500 includes the following configuration as an example.

-CPU (Central Processing Unit) 501
-ROM (Read Only Memory) 502
-RAM (Random Access Memory) 503
-Program 504 loaded into RAM 503
A storage device 505 that stores the program 504.
Drive device 507 that reads and writes the recording medium 506.
-Communication interface 508 that connects to the communication network 509
-I / O interface 510 for inputting / outputting data
-Bus 511 connecting each component
Each component of each device in each embodiment is realized by the CPU 501 acquiring and executing a program 504 that realizes these functions. The program 504 that realizes the functions of each component of each device is stored in, for example, a storage device 505 or a RAM 503 in advance, and is read by the CPU 501 as needed. The program 504 may be supplied to the CPU 501 via the communication network 509, or may be stored in the recording medium 506 in advance, and the drive device 507 may read the program and supply the program to the CPU 501.

There are various modifications in the method of realizing each device. For example, each device may be realized by any combination of the information processing device 500 and the program, which are separate for each component. Further, a plurality of components included in each device may be realized by any combination of one information processing device 500 and a program.

Further, a part or all of each component of each device is realized by a general-purpose or dedicated circuit including a processor or the like, or a combination thereof. These may be composed of a single chip or may be composed of a plurality of chips connected via a bus. A part or all of each component of each device may be realized by a combination of the above-mentioned circuit or the like and a program.

When a part or all of each component of each device is realized by a plurality of information processing devices and circuits, the plurality of information processing devices and circuits may be centrally arranged or distributed. May be good. For example, the information processing device, the circuit, and the like may be realized as a form in which each is connected via a communication network, such as a client-and-server system and a cloud computing system.

(First Embodiment)
First, the first embodiment of the present invention will be described. As shown in FIG. 1, the task estimation device 100 according to the first embodiment of the present invention includes at least a task estimation unit 110. The task estimation unit 110 extracts an expression indicating a measure and an expression indicating a relationship between the measure and the problem from a plurality of documents, and estimates the problem based on the extracted expression.

As described above, when developing new solutions, services, new products, etc., it is preferable to grasp the requests of customers and the like who use them. In order to grasp the needs of customers, etc., make a hypothesis about the problems that customers, etc. have, that is, problems that should be solved or improved for customers, etc., and based on the hypothesis, establish the validity of the hypothesis through communication with customers, etc. A method of verification is assumed. By identifying important issues for customers, etc. through communication with customers, etc., it is possible to develop promising solutions and products. Therefore, it may be necessary to comprehensively grasp the issues of customers and the like.

However, when verifying a hypothesis about a problem, it is necessary to grasp the problem as a premise. In order to comprehensively grasp the issues, various information such as the field to which the customer belongs may be required. In addition, it may be necessary to understand future issues in the field to which the customer belongs and issues in the field related to the customer. Furthermore, it may be difficult to grasp known issues. In other words, a large amount of effort may be required to comprehensively grasp the issues.

Therefore, the task estimation unit 110 of the task estimation device 100 estimates the task based on expressions indicating measures included in the plurality of documents. That is, even if the problem is unknown, the problem estimation device 100 makes it possible to comprehensively grasp the problem for which some measure has already been taken together with the measure. Further, the task estimation unit 110 of the task estimation device 100 may select a task related to the target based on the information of the customer or the like that is the target of the task. By comprehensively estimating the problem in this way, for example, it becomes possible to comprehensively grasp the problem to be verified related to the target of the problem.

Each of the plurality of documents is information including character information. The character information includes, for example, a sentence, a sentence composed of a plurality of sentences, and the like. However, the character information is not limited to these, and may be in the form of a table, or the explanations in the figures may be treated as character information. The task estimation unit 110 handles various information including tables, figures, images, and the like as a part of a plurality of documents.

The plurality of documents are, for example, web pages. The plurality of documents are not limited to web pages and the like as long as they include character information, and may be, for example, books and the like. Further, the format of the data when each of the plurality of documents is handled by the task estimation unit 110 is not particularly limited. That is, the plurality of documents are, for example, text data, but may be in another format as long as the information in characters can be recognized.

Subsequently, the details of the task estimation device 100 in the present embodiment will be described. The problem estimation unit 110 extracts a measure from a plurality of documents based on an expression indicating the measure, and estimates a plurality of problems by extracting an expression indicating the relationship between the measure and the problem. That is, the task estimation unit 110 extracts task candidates from a plurality of documents as task estimation. As one configuration example, the task estimation unit 110 includes a measure extraction unit 111 and a problem extraction unit 112 as shown in FIG. The measure extraction unit 111 extracts a measure based on an expression indicating the measure. The problem extraction unit 112 extracts an expression indicating the relationship between the measure extracted by the measure extraction unit 111 and the problem, and extracts the problem based on the expression.

Further, as shown in FIG. 2, the task estimation device 100 may include a general information storage unit 150, a policy expression storage unit 160, and a policy task expression storage unit 170. The general information storage unit 150 stores the plurality of documents described above. The measure expression storage unit 160 stores expressions related to the measure. The measure problem expression storage unit 170 stores an expression indicating the relationship between the measure and the problem. Details of the contents stored in these will be described later. These elements are connected to the other components of the task estimation device 100 via a bus connecting each element of the information processing device, a wired or wireless communication network, or the like.

The measure extraction unit 111 extracts a measure from a plurality of documents based on expressions indicating the measure. The measure extraction unit 111 searches for an expression indicating the measure included in the document, and estimates the means and the like appearing in relation to the expression indicating the measure as the measure. The search method used by the measure extraction unit 111 is not particularly limited, and a general character string search method is used. Further, the number of measures extracted by the measure extraction unit 111 is not particularly limited.

In each embodiment of the invention, the measure represents any means that is or can be done. The expression indicating a measure is, for example, an expression meaning that a means to be a measure is being implemented or can be implemented. The expression "focusing on ..." is an example of an expression that indicates a measure. In this case, the measure extraction unit 111 searches a plurality of documents for the expression "focusing on ...". When the search result includes, for example, the expression "we are focusing on the enhancement of sweets", the measure extraction unit 111 extracts "enhancement of sweets" as a measure.

When the task analysis device 100 includes the general information storage unit 150, the measure extraction unit 111 extracts, for example, an expression indicating a measure from a plurality of documents stored in the general information storage unit 150. Further, the measure extraction unit 111 may acquire a plurality of documents each time by searching an external web page or the like, and extract expressions indicating measures from the acquired plurality of documents. The measure extraction unit 111 may extract expressions indicating measures from both the document accumulated by the general information storage unit 150 and the document obtained by searching an external web page.

The expression indicating the measure is not limited to the above-mentioned expression "focusing on ...". FIG. 3 is an example of an expression showing a measure. In Fig. 3, in addition to the expression "we are focusing on", the expressions "investing in", "strengthening" and "measures are" are used. It is mentioned as an expression indicating measures.

The expressions indicating these measures are stored in the above-mentioned measure expression storage unit 160 as an example. When the measure expression storage unit 160 is provided, the measure extraction unit 111 extracts the measure by searching for the presence / absence of the expression stored in the measure expression storage unit 160 in the document, for example. In this case, the measure extraction unit 111 may search for all the expressions stored in the measure expression storage unit 160, or may search for some expressions. In addition, the measure extraction unit 111 may acquire an expression indicating the measure from the outside each time when extracting the measure.

In addition, the measure extraction unit 111 may extract measures related to the target based on the information indicating the target of the problem to be estimated. In the description of each embodiment below, the subject of the task may also be referred to as a domain.

The target of the task is, for example, a unit such as a company, an organization, or an individual who has a common problem or attribute. More specifically, if the subject is a company, the target is the industry, business field, company size, location, and the like. In addition, the unit of the subject of the task may be "a person (or a company) who is having trouble handling the elderly", or if the task is for an individual, it may be gender, age, etc. Good. The target of the task is designated, for example, via an input means (not shown).

As an example, the measure extraction unit 111 estimates the measure related to the domain by searching the name of the domain that is the target of the problem as a key phrase together with the estimation of the measure. For example, if the domain is "convenience store" and it is assumed that the expression "focusing on ..." as described above is an expression indicating a measure, the measure extraction unit 111 is in the document. Search for the expressions "convenience store" and "focusing on". Even in this case, the search is performed on a plurality of documents such as a web page included in the general information storage unit 150, for example. When the search result includes the expressions "convenience store" and "focusing on the enhancement of sweets", the measure extraction unit 111 indicates that the domain measure with "convenience store" is "enhancement of sweets". Presumed to be.

In addition, the measure extraction unit 111 may extract measures related to the target of the problem by another method. For example, when a plurality of documents are stored in the general information storage unit 150 for each domain, the task extraction unit 111 relates to the target by extracting a measure from the documents for the domain to be the target of the task. Extract certain measures. The problem extraction unit 111 may extract measures related to the target by extracting the measures without limiting the target domain of the problem and then narrowing down the measures based on the words and phrases included in the measures. ..

The issue extraction unit 112 extracts issues related to each of the measures extracted by the measure extraction unit 111. More specifically, the problem extraction unit 112 extracts expressions indicating the relationship between the measure and the problem from a plurality of documents, and extracts the problem related to the measure.

In each embodiment of the present invention, the problem is, for example, a problem that the target company or the like should tackle. The issue may also represent the purpose for which the relevant measures are taken. The number of tasks estimated by the task extraction unit 112 is not particularly limited. When a plurality of measures are estimated by the measure estimation unit 110, the problem extraction unit 112 estimates a problem related to each of the estimated plurality of measures as an example.

The task extraction unit 112 extracts from a plurality of documents that a specific expression is an expression indicating the relationship between the measure and the task, and extracts the expression corresponding to the task from the extracted expressions as the task. The problem extraction unit 112 searches a plurality of documents for expressions indicating the relationship between the measures estimated by the measure extraction unit 111 and the problem. The task extraction unit 112 estimates an expression related to a task or purpose that appears with the expression as a task related to the measure. The search method used by the problem extraction unit 112 is not particularly limited as in the measure extraction unit 111, and a general character string search method is used.

The expression indicating the relationship between the measure and the problem is an expression indicating that the measure is taken for the purpose of solving the problem. As an example, it is assumed that the expression "[measure] is the aim of ..." indicates the relationship between the measure and the problem. The problem extraction unit 112 performs a search by incorporating the measures estimated by the measure extraction unit 111 into the position of [measure] in this expression. For example, when the measure extraction unit 111 extracts "enhancement of sweets" as a measure as described above, the problem extraction unit 112 expresses "the aim is to enhance sweets" in the document. Search for.

As an example, if the search result includes the expression "enhancement of sweets is aimed at attracting female customers", the task extraction unit 112 states that the purpose of "enhancement of sweets" is "incorporation of female customers". Presumed to be. That is, it is presumed that the problem extraction unit 112 takes measures to "enhance sweets" with "capturing female customers" as a problem.

Similar to the measure extraction unit 111, the task extraction unit 112 extracts expressions indicating a problem from a plurality of documents stored in the general information storage unit 150, for example, when the task analysis device 100 includes the general information storage unit 150. To do. Further, the task extraction unit 112 may acquire a plurality of documents each time it is extracted by searching an external web page or the like, and extract an expression indicating a problem from the acquired plurality of documents. The task extraction unit 112 may extract an expression indicating a problem from both the document accumulated by the general information storage unit 150 and the document obtained by searching an external web page. The document referred to by the problem extraction unit 112 when extracting the problem may be the same as or different from the measure extraction unit 111.

The expression indicating the relationship between a measure and an issue is not limited to the above-mentioned expression "[measure] is aimed at." FIG. 4 is an example of expressions showing the relationship between measures and issues. In the example shown in FIG. 4, as expressions showing the relationship between the measure and the problem, in addition to the expression "[measure] is aimed at", "the purpose of [measure] is ~" and "~". Is regarded as an issue, and the expressions "[measures]" and "solve ... by [measures]" are mentioned. Similar to the above example, the task extraction unit 112 performs a search by incorporating the measures estimated by the measure extraction unit 111 into the position of [measure] in the above example.

The expression indicating the relationship between these measures and the issues is stored in the above-mentioned measure issue expression storage unit 170 as an example. When the measure task expression storage unit 170 is provided, the task extraction unit 112 searches for the presence or absence of expressions stored in the measure task expression storage unit 170 in the document and extracts the measure. In this case, the task extraction unit 112 may search for all the expressions stored in the measure task expression storage unit 170, or may search for some expressions. In addition, the problem extraction unit 112 may acquire an expression indicating the relationship between the measure and the problem from the outside via a communication network or the like each time when estimating the problem.

Further, the document to be searched by the task extraction unit 112 is, for example, a web page or the like, but is not particularly limited. The task extraction unit 112 searches, for example, a document stored in the general information storage unit 150. In addition, the task extraction unit 112 may acquire the document to be searched each time when estimating the task. The search target by the problem extraction unit 112 may be the same as or different from the search target by the measure extraction fixed unit 111.

Further, the problem extraction unit 112 may select a problem related to the domain to be the target of the problem based on the information indicating the target of the problem to be estimated, similarly to the measure extraction unit 111. That is, the task extraction unit 112 also performs a search using the name of the domain as a key phrase when estimating the above-mentioned task. For example, in the above example, when the domain is "convenience store", the task extraction unit 112 searches for the expressions "convenience store" and "the enhancement of sweets is aimed at ..." in the document. Similar to the measure extraction unit 111, the issue extraction unit 112 may extract issues related to the target domain by another method. In this way, the problem extraction unit 112 makes it possible to easily grasp a comprehensive problem related to a specific target by extracting a problem related to the target.

Subsequently, an operation example of the problem estimation device 100 in the present embodiment will be described with reference to the flowchart shown in FIG.

First, for example, the measure extraction unit 111 of the task estimation unit 110 refers to a web page published on an external website of the task estimation device 100, a general information storage unit 150, or the like, and implements a measure from a plurality of documents. Extract. Therefore, the measure extraction unit 111 first searches for an expression indicating the measure included in the plurality of documents (step S101). Then, the measure extraction unit 111 estimates as a measure a means or the like that appears in association with the expression indicating the measure acquired by the search in step S101 (step S102). The measure extraction unit 111 refers to the expression stored in the measure task expression storage unit 160 or the like as an expression indicating the measure as necessary.

Next, for example, the task extraction unit 112 of the task estimation unit 110 estimates a problem related to the measure extracted in step S101 from a plurality of documents. The task extraction unit 112 searches a plurality of documents for expressions indicating the relationship between the measures estimated by the measure extraction unit 111 and the tasks (step S103). The task extraction unit 112 estimates the task or the expression related to the purpose that appears together with the expression indicating the relationship between the measure and the task acquired in step S103 as the task related to the measure (step S104).

The task extraction unit 112 is stored as a plurality of documents, as a plurality of documents, in a web page or a general information storage unit 150 published on a website outside the task estimation device 100, as in the case of step S101. Refer to the documents etc. Further, the task extraction unit 112 refers to the expression stored in the measure task expression storage unit 170 as an expression indicating the measure as necessary.

As described above, in the task estimation device 100 according to the first embodiment of the present invention, the task estimation unit 110 extracts an expression indicating a measure and an expression indicating a relationship between the measure and the problem from a plurality of documents, and the extracted expression. Estimate the task based on.

In order to comprehensively grasp the issues, it may be difficult to grasp the known issues, and a large amount of labor may be required. In the problem estimation device 100 of the present embodiment, even when a known problem is not grasped, for example, the measure extraction unit 111 extracts a measure from a plurality of documents based on expressions indicating the measure. Then, for example, the problem extraction unit 112 extracts an expression indicating the relationship between the measure and the problem and extracts the problem. That is, the problem estimation device 100 in the present embodiment estimates a problem related to a predetermined target or the like from a plurality of documents such as a web page together with a measure even when a wit problem is not grasped. Therefore, the problem estimation device 100 in the present embodiment makes it possible to comprehensively grasp the problem.

(Second embodiment)
Next, a second embodiment of the present invention will be described. As shown in FIG. 6, the task estimation device 200 according to the second embodiment of the present invention includes at least a task estimation unit 210. Similar to the task estimation unit 110 described in the first embodiment, the task estimation unit 210 extracts an expression indicating a measure and an expression indicating a relationship between the measure and the problem from a plurality of documents, and is based on the extracted expression. Then, the issues related to the predetermined target are estimated. The task estimation unit 210 further extracts a task in a target different from the target based on the task from a plurality of documents, and among the tasks in the different target, a task related to the plurality of tasks related to the above-mentioned target. Is extracted.

That is, in the task estimation device 200 of the present embodiment, the task estimation unit 210 can comprehensively grasp the task by extracting a task related to a target different from the predetermined task target. Further, in the task estimation device 200 of the present embodiment, the task estimation unit 210 extracts the tasks related to the plurality of tasks related to the above-mentioned target from the tasks related to the different targets, thereby described above. It becomes possible to grasp issues that are highly relevant to the target.

Subsequently, the details of the task estimation device 200 in the present embodiment will be mainly described as different from the task estimation device 100 in the first embodiment. As one configuration example, the problem estimation unit 210 includes a measure extraction unit 111, a problem extraction unit 112, a problem expansion unit 213, and a problem narrowing unit 214, as shown in FIG. The measure extraction unit 111 and the problem extraction unit 112 are the same components as the measure extraction unit 111 and the problem extraction unit 112 in the first embodiment, respectively. The task expansion section 213 and the task narrowing section 214 are elements that are not included in the task estimation section 110 in the first embodiment. The task extension unit 213 extracts a task related to a target different from the predetermined target from the plurality of documents based on the task extracted by the task extraction unit 112. The problem narrowing unit 214 extracts the problems related to a plurality of measures related to the predetermined target from the problems in different targets.

Further, as shown in FIG. 7, the task estimation device 200 may include a general information storage unit 150, a policy expression storage unit 160, a policy task expression storage unit 170, and a task expression storage unit 180. The general information storage unit 150, the measure expression storage unit 160, and the measure task expression storage unit 170 are elements similar to the elements having the same reference numerals in the example shown in FIG. 2 with respect to the task estimation device 100 in the first embodiment. is there. In addition, the task expression storage unit 180 stores expressions related to the task. These elements are connected to the other components of the task estimation device 200 via a bus, a wired or wireless communication network, or the like.

The task extension unit 213 extracts a task in a target different from the predetermined target from a plurality of documents based on the task extracted by the task extraction unit 112. That is, the task expansion unit 213 expands the target to a domain that is different from the target of the task specified in advance based on the task extracted by the task extraction unit 112, and extracts the problem. The domain that is a different target is not particularly limited as long as it is different from the predetermined target. However, the task extension unit 213 may extract tasks for specific different objects.

For example, the task extension unit 213 extracts an expression indicating a task appearing together with the task estimated by the task estimation unit 112 from a plurality of documents relating to an object different from the designated object, and based on the expression, in a different object. Extract issues. By confirming that the document does not include words and phrases related to the specified target, the task extension unit 213 confirms that the document is a document related to a target different from the specified target. To do. Then, the task extension unit 213 searches for the task extracted by the task extraction unit 112 and the expression indicating the task for the documents related to different objects. That is, the task extension unit 213 extracts a task related to a target different from the target by extracting the task from a plurality of documents that do not include words and phrases related to the target.

As an example, the task extension unit 213 assumes that the expression "... is a task" indicates a task. In this case, the task extension unit 213 searches for the expression "... is a task". The expression indicating the task is not limited to the expression "... is the task", and other expressions may be used. FIG. 8 is an example of an expression showing a problem. In the example shown in FIG. 8, in addition to the expression "is a problem", "I am in trouble with", "the problem is", "needs", "is concerned" and "is a concern". The expression "anxiety" is mentioned as an expression indicating a problem.

The expressions indicating these tasks are stored in the task expression storage unit 180 described above as an example. When the task expression storage unit 180 is provided, the task extension unit 213 searches for the presence or absence of expressions stored in the task expression storage unit 180 in the document and extracts the task. In this case, the task extension unit 213 may search for all the expressions stored in the task expression storage unit 180, or may search for some expressions. Further, the task extension unit 213 may acquire an expression indicating the task from the outside each time when estimating the task of a different target.

The processing performed in the task expansion unit 213 will be described with reference to a more specific example. In this example, the domain that is the subject of the task is "convenience store", and the documents related to "convenience store" are expressed as "focusing on enhancing sweets" and "focusing on home delivery". Suppose it is included. Therefore, it is presumed that "enhancement of sweets" and "home delivery" are the measures by searching the measure extraction unit 111 as the expression "focusing on ..." as the expression indicating the measure. ..

Further, in this example, it is assumed that the documents related to "convenience store" include the expressions "the enhancement of sweets is aimed at attracting female customers" and "the delivery to home is aimed at attracting elderly people". Therefore, the issue related to the measure of "enhancement of sweets" is extracted as "incorporation of female customers", and the issue related to the measure of "home delivery" is extracted as "incorporation of elderly people". The problem is extracted by the problem extraction unit 112 searching for the expression "(measure) is aimed at ..." as an expression indicating the relationship between the measure and the problem.

In this case, the task extension unit 213 searches a plurality of documents for an expression indicating a problem related to a target different from the predetermined target "convenience store", assuming that the expression shown in FIG. 8 is an expression indicating the problem. .. The task extension unit 213 searches a plurality of documents for expressions that do not include the phrase "convenience store", which is a predetermined target, and includes the task extracted earlier and other tasks, thereby different targets. Search for expressions that indicate issues related to.

For example, in multiple documents, the phrase "convenience store" is not included, and the expression "capturing female customers", which is the problem extracted earlier, and "the problem is how to deal with tourists", which indicates the problem, are expressed. It is assumed that it is included. In this case, the task expansion unit 213 extracts the task "correspondence to tourists" as a different target task.

Similarly, the expression "convenience store" is not included in multiple documents, and the previously extracted issues "incorporation of the elderly" and "responding to tourists" indicating the issues are expressed. It is assumed that it is included. In this case, the task extension unit 213 presumes that "correspondence to tourists" is a task of a different domain as in the above example. In the above example, it was judged that the problem was related to a different object by whether or not a predetermined word related to the object such as "convenience store" was included in the document, but other methods were used. May be done.

When the task analysis device 100 includes the general information storage unit 150, the task expansion unit 213 extracts, for example, an expression indicating a problem from a plurality of documents stored in the general information storage unit 150. Further, the task extension unit 213 may acquire a plurality of documents each time it is extracted by searching an external web page or the like, and extract an expression indicating the task from the acquired plurality of documents. The document referred to by the task extension unit 213 may be the same as or different from other elements such as the measure extraction unit 111 and the problem extraction unit 112.

The task narrowing unit 214 narrows down the tasks extracted by the task expansion unit 213 as tasks in different targets. The issue narrowing unit 214 selects issues related to a plurality of measures from the issues extracted as issues related to different targets.

As described above, the task extension unit 213 extracts the tasks that appear together with the tasks extracted by the task extraction unit 112. The issues extracted by the issue extraction unit 112 are issues related to the measures extracted by the measure extraction unit 111. Therefore, it is considered that the issues in the different objects extracted by the issue extension unit 213 are related to the measures extracted by the measure extraction unit 111. In this case, the problem narrowing unit 214 selects a problem related to a plurality of measures extracted by the measure extraction unit 111 from among the problems in different targets extracted by the problem expansion unit 213.

In the above example, according to the task expansion unit 213, the task of "corresponding to tourists" in other targets is "capturing female customers" and "capturing elderly people" in the predetermined target "convenience store". It is extracted in relation to the issue of. In addition, in this example, a measure of "enhancing sweets" is taken for the issue of "acquiring female customers", and a measure of "home delivery" for the issue of "acquiring elderly people". Is being done. In other words, the issue of "corresponding to tourists" is related to the two measures of "enhancement of sweets" and "home delivery". Therefore, in this case, the issue narrowing unit 214 selects "correspondence to tourists" as an issue related to a plurality of measures.

With reference to FIG. 9, the selection of the task by the task narrowing unit 214 will be described in more detail. In the example shown in FIG. 9, it is assumed that the task estimation device 200 extracts measures and tasks related to domain A as the target of the task.

In the example shown in FIG. 9, the measure extraction unit 111 estimates four measures 1 to 4 as the measures of the domain A which is the target of the problem. In addition, the task extraction unit 112 estimates tasks 1 to 4 as tasks of domain A for each of measures 1 to 4. For each of the tasks 1 to 4, the task extension unit 213 extracts a task related to a target different from the domain A which is the target of the predetermined task (that is, a domain other than the domain A). Each of the "domains other than domain A" may be the same or different. In the example shown in FIG. 9, it is assumed that the task expansion unit 213 extracts task 5 for tasks 1 and 2, task 6 for task 3, and task 7 for task 4.

In this case, the task narrowing unit 214 selects a task related to a plurality of measures from the tasks 5 to 7 which are the tasks of the domain different from the domain A which is the target of the predetermined task. In the example of FIG. 9, the problem 5 is a problem for the measure 1 and the measure 2. That is, the problem 5 is a problem common to a plurality of measures. On the other hand, the problem 6 is a problem only for the measure 3. Problem 7 is a problem only for measure 4. Therefore, the task narrowing unit 214 selects task 5 in this case.

When the number of issues related to the plurality of measures is plural, the issue narrowing unit 214 may select the issues using various criteria. For example, the issue narrowing unit 214 may select all the issues related to a plurality of measures. In addition, the issue narrowing unit 214 may select some issues related to a plurality of measures based on the number of common measures. For example, when there are a problem related to three measures and a problem related to two measures as problems related to different targets, the problem narrowing unit 250 may select any of the three problems. Only issues related to the measure may be selected.

The issue expansion unit 213 and the issue narrowing unit 214 are not limited to the measures and issues extracted by the measure extraction unit 111 and the issue extraction unit 112, respectively, but are different targets based on the measures or issues extracted by other means. You may extract the problem in.

Subsequently, an operation example of the problem estimation device 200 according to the second embodiment of the present invention shown in FIG. 7 will be described with reference to the flowchart shown in FIG.

The operations from steps S201 to S204 are the same as the operations from steps S101 to S104 of the task estimation device 100 in the first embodiment. That is, the measure extraction unit 111 first searches for an expression indicating the measure included in the document (step S201). Then, the measure extraction unit 111 estimates as a measure a means or the like that appears in association with the expression indicating the measure acquired by the search in step S101 (step S202). The measure extraction unit 111 refers to the expression stored in the measure task expression storage unit 160 or the like as an expression indicating the measure as necessary.

Next, the task extraction unit 112 searches the plurality of documents for expressions indicating the relationship between the measure estimated by the measure extraction unit 111 and the problem (step S203). The problem extraction unit 112 estimates the problem or the expression related to the purpose that appears together with the expression indicating the relationship between the measure and the problem acquired in step S203 as the problem related to the measure (step S204).

Next, the task extension unit 213 extracts a task in a target different from the predetermined target from a plurality of documents based on the task extracted by the task extraction unit 112 in step S202. In this case, the task extension unit 213 extracts an expression indicating the task that appears together with the task estimated by the task estimation unit 112 in step S204 from a plurality of documents relating to different objects (step S205). In step S205, the task extension unit 213 confirms that the plurality of documents are documents relating to a target different from the designated target by the fact that words and phrases related to the designated target do not appear in the document. Then, the task extension unit 213 extracts a task in a different target based on the expression extracted in step S205 (step S206). As a plurality of documents, the task extension unit 213 refers to a web page published on a website outside the task estimation device 100, a document stored in the general information storage unit 150, or the like, as in the above steps. .. In addition, the task expansion unit 213 refers to the expression stored in the task expression storage unit 180 as an expression indicating the task, if necessary.

Next, the task narrowing section 214 selects a task related to a plurality of measures estimated by the measure extraction section 111 in step S201 from among the tasks of different targets estimated by the task expansion section 213 in step S206 (step). S207).

As described above, the problem estimation device 200 according to the second embodiment of the present invention is a problem in a target different from the target of the task specified in advance from a plurality of documents based on the problems extracted for the target specified in advance. To estimate. Therefore, as compared with the problem estimation device 100 in the first embodiment, it is possible to grasp a more comprehensive problem.

Further, the task estimation device 200 selects a task related to a plurality of measures in a predetermined target from among the tasks extracted as a task in a different target from a plurality of documents. Therefore, the task estimation device 200 makes it possible to comprehensively grasp the tasks that are important and it is preferable to verify the hypothesis of the tasks at an early stage.

(Third Embodiment)
Next, a third embodiment of the present invention will be described. As shown in FIG. 11, the task estimation device 300 according to the third embodiment of the present invention includes a task estimation unit 310 and a priority determination unit 320. The task estimation unit 310 extracts at least an expression indicating a measure and an expression indicating a relationship between the measure and the problem from a plurality of documents, and estimates a problem related to a predetermined target based on the extracted expression. The task estimation unit 310 is the same element as either the task estimation unit 110 in the first embodiment or the task estimation unit 210 in the second embodiment. 12 and 13 are examples of the configuration in this case. Further, the priority determination unit 320 determines the priority for each of the tasks estimated by the task estimation unit 310.

Once a task is estimated, a hypothesis about the task may be tested. In order to test the hypothesis efficiently, it is preferable that the validity of the hypothesis can be quickly and surely performed. Therefore, in the task estimation device 300 of the present embodiment, the priority determination unit 320 determines the priority for each of the tasks estimated by the task estimation unit 310. Therefore, for example, the validity of the hypothesis about the task can be verified from the task with high priority. That is, by using the priority of the task estimated by the task estimation device 300, it is possible to efficiently verify the validity of the hypothesis about the task.

The priority determination unit 320 determines the priority based on, for example, information related to the certainty of the task. The priority determination unit 220 acquires information related to the task as needed, and determines the priority based on the information.

As an example, the priority determination unit 320 determines the priority based on the plausibility of the task by using the information related to the certainty of the task. That is, the priority determination unit 320 determines the priority so that the task that is assumed to be plausible for the target of the task is given a high priority. Issues that are assumed to be plausible are issues that the company, etc. that is the target of the issue actually has and that is likely to be appropriate for the target.

In this case, the priority determination unit 220 determines the priority based on the creation time of the document including the extracted task. As a specific example, the priority determination unit 220 determines the priority based on the novelty of the document including the extracted task. The priority determination unit 220 determines the priority so that, for example, the creation time (for example, creation date) of the document including the extracted task gives the highest priority in order from the new task. The document creation time may be another date related to the description of the subject, such as the last update date of the document.

The issues contained in the new document can be interpreted as having recently begun to be perceived as issues for the subject. Then, when the target of the problem is a company or the like, there is a possibility that the business related to the problem is promising. Therefore, in this case, the priority determination unit 320 determines the priority so as to give a high priority, assuming that the task included in the new document is a relatively plausible task.

FIG. 14 is a diagram showing an example in which the priority determination unit 320 determines the priority based on the novelty of the document including the extracted task. In the example shown in FIG. 14, it is assumed that the task estimation unit 310 estimates issues such as “capture of tourists”, “capture of elderly people”, and “capture of customers of other companies”. Further, FIG. 14 shows the creation date of the document including each task.

In the example of FIG. 14, the document including the issue of "acquiring tourists" is the newest document with the creation date of July 15, 2015, and the document including the issue of "acquiring customers of other companies". Is the oldest document created on February 4, 1995. Therefore, the priority determination unit 320 determines the priority so as to give the highest priority to the task of "acquiring tourists" and the lowest priority to the issue of "acquiring customers of other companies". In addition, the priority determination unit 320 determines the priority so as to give a priority between the other two tasks to the task of "taking in the elderly".

The priority determination unit 320 may use other criteria to determine the plausibility of the task described above. As another example, the priority determination unit 320 has a hit rate of a search for a document containing an issue when a plurality of documents are searched for when extracting an issue, a ranking of a search ranking for a document containing the issue, and the like. The priority may be determined based on. That is, the priority determination unit 320 considers that the issues included in the documents having a high search hit rate and the documents having a high search ranking are attracting a lot of attention and are promising issues for the subject of the assignment. Determine priorities to give priority.

In addition, when a word indicating uncertainty appears together with an expression indicating a problem in a document (that is, an expression indicating a problem and an expression indicating uncertainty co-occur), the problem is accompanied by these words. It is assumed that the degree of plausibility is low compared to the problems that do not appear. Therefore, as yet another example, the priority determination unit 320 may determine the priority based on the ratio (co-occurrence rate) at which the expression indicating the problem and the word indicating uncertainty appear together. For example, the priority determination unit 320 may determine the priority so as to lower the priority of the task that appears with the word indicating uncertainty. The word indicating uncertainty is, for example, a word such as "probably" or "may", but it may be a word different from these.

Further, the priority determination unit 320 may determine the priority based on the logical relationship of the tasks.

When there is a logical relationship between multiple issues, if one issue is not a valid issue for the target of the task (that is, it is not an issue actually held by the target company, etc.), the other The task may also not be a valid task for the subject. If there is such a relationship and one of the multiple tasks is not a valid task, the other tasks are not valid tasks for the subject, so it is not always necessary to test the hypothesis about the other tasks. there is a possibility. In other words, by verifying the hypothesis of one task first, it may not be necessary to verify the hypothesis of another task.

FIG. 15 is an example of the logical relationship between the tasks. There are three issues, "Issue 1" and "Issue 2", and "Upper issues" above them. It is assumed that the validity of the "upper task" depends on the validity of "task 1" and "task 2". That is, when both "problem 1" and "problem 2" are valid tasks for the subject of the task, it is assumed that the "upper task" is a valid task for the target of the task. If at least one of "task 1" and "task 2" is not appropriate for the subject of the task, it is assumed that the "upper task" is not a valid task for the subject of the task. Therefore, by first verifying the hypothesis of either "problem 1" or "problem 2", the unverified hypothesis of "upper task" and "problem 1" or "problem 2" can be verified. It may be omitted.

The priority determination unit 320 determines the priority of "task 1" and "task 2" so as to give a higher priority than the "higher-level task". Further, the priority determination unit 320 gives priority to "problem 1" and "problem 2" so as to give a high priority to a task having a low degree of plausibility (that is, not plausible) in this case. decide. By determining the priority in this way, the priority determination unit 320 can omit the verification of the remaining unverified task hypothesis when the result of verifying the task hypothesis is not valid for the task target. Become. That is, when either "problem 1" or "problem 2" is not a valid task for the subject, at least verification of the hypothesis of "upper task" can be omitted.

The priority determination unit 320 may determine the priority of "problem 1" and "problem 2" based on a standard different from the above. That is, the priority determination unit 320 may give the same priority to "task 1" and "task 2", or may give different priorities.

Subsequently, the operation of the task estimation device 300 in the third embodiment shown in FIG. 9 will be described with reference to the flowchart shown in FIG.

The operations of steps S301 to S307 are the same as the operations of steps S201 to S207 in the second embodiment, and the description thereof will be omitted.
In the task estimation device 300 of the present embodiment, following the operations up to step S307, the priority determination unit 320 performs the operation of step S308. That is, the priority determination unit 320 determines the priority of the plurality of tasks estimated up to step S307 based on the information related to the tasks (step S308).

As described above, in the task estimation device 300 according to the third embodiment of the present invention, the priority determination unit 320 determines the priority for the task estimated by the task estimation unit 310. By verifying the hypothesis of the task based on the priority, for example, it becomes possible to grasp the promising task that the target of the task is likely to actually have at an early stage. Alternatively, by verifying the hypothesis of the task based on the priority, it is possible to suppress the verification of the hypothesis of the unnecessary task.

Therefore, the task estimation device 300 makes it possible to comprehensively grasp the problem, similarly to the task estimation device 100 and the like in the first embodiment described above. In addition, the task estimation device 300 makes it possible to improve the efficiency of verification of hypotheses about the task.

Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the structure and details of the present invention within the scope of the present invention. In addition, the configurations in each embodiment can be combined with each other as long as they do not deviate from the scope of the present invention.

100, 200, 300 Problem estimation device 110, 210, 310 Problem estimation unit 111 Measure extraction unit 112 Problem extraction unit 213 Problem expansion unit 214 Problem narrowing unit 320 Priority determination unit 150 General information storage unit 160 Measure expression storage unit 170 Measures issue Expression memory 180 Task expression memory

Claims (12)

An expression indicating a measure and an expression indicating the relationship between the measure and the issue are extracted from a plurality of documents, and a task estimation means for estimating the issue based on the extracted expression, and a task estimation means .
A priority determination means for determining priorities for a plurality of the above-mentioned tasks is provided.
The priority determination means is a task estimation device that determines the priority based on the creation time of each of the plurality of documents including the plurality of tasks. The problem estimation means is described in claim 1, wherein a measure is extracted from a plurality of documents based on an expression indicating the measure, and a plurality of problems are estimated by extracting an expression indicating the relationship between the measure and the problem. Problem estimation device. The task estimation device according to claim 1 or 2, wherein the task estimation means extracts an expression indicating the measure related to the target of the task based on the information indicating the target of the task. The problem estimation means is
Measure extraction means to extract measures based on the expression indicating the above measures,
The problem estimation device according to any one of claims 1 to 3, further comprising a problem extraction means for extracting an expression indicating a relationship between the measure and the problem and extracting the problem. The task estimation device according to any one of claims 1 to 4 , wherein the priority determining means determines the priority based on a logical relationship between the plurality of tasks. An expression indicating a measure and an expression indicating the relationship between the measure and the problem are extracted from a plurality of documents, and a problem estimation means for estimating the problem based on the extracted expression is provided .
The task estimation means extracts an expression indicating the measure related to the target of the task based on the information indicating the target of the task.
The task estimation means further extracts a task related to a target different from the target of the task from a plurality of documents that do not include words and phrases related to the target of the task, based on the task. A task estimation device equipped with. The problem estimation means is
The task estimation device according to claim 6 , further comprising a task narrowing-down means for extracting a plurality of tasks related to the measure from among the tasks in a target different from the target of the task. The problem estimation means estimates a plurality of problems by extracting a measure from a plurality of documents based on an expression indicating the measure and extracting an expression indicating the relationship between the measure and the problem, claim 6 or 7. The problem estimation device described in. The computer
From a plurality of documents, an expression indicating a measure and an expression indicating the relationship between the measure and the problem are extracted, and the problem is estimated based on the extracted expression .
A task estimation method for determining a priority for a plurality of the tasks based on the creation time of each of the plurality of documents including the plurality of tasks. On the computer
A process of extracting an expression indicating a measure and an expression indicating a relationship between the measure and a problem from a plurality of documents and estimating a problem based on the extracted expression .
A program for executing a process of determining a priority of a plurality of the tasks based on the creation time of each of the plurality of documents including the plurality of tasks. The computer
From a plurality of documents, an expression indicating a measure and an expression indicating the relationship between the measure and the problem are extracted, and the problem is estimated based on the extracted expression .
Based on the information indicating the target of the task, an expression indicating the measure related to the target of the task is extracted.
A task estimation method for extracting a task related to a target different from the target of the task based on the task from a plurality of documents that do not include words and phrases related to the target of the task. On the computer
A process of extracting an expression indicating a measure and an expression indicating a relationship between the measure and a problem from a plurality of documents and estimating a problem based on the extracted expression .
Based on the information indicating the target of the task, the process of extracting the expression indicating the measure related to the target of the task, and
A program that executes a process of extracting a task related to a target different from the target of the task based on the task from a plurality of documents that do not include words and phrases related to the target of the task.

Images