JP7572933B2 - Case management device, case management method, and program - Google Patents

Description

The present invention relates to technology for managing various cases in business. Cases include projects such as system construction.

Currently, various projects such as system construction are being carried out in the course of business. In carrying out such projects, various issues may arise. For example, there may be a deterioration in the quality of the performance of the resulting system. In order to deal with such issues, past projects (past cases) may be referred to for response. To do this, it is necessary to identify past cases similar to the project currently being handled.

For example, Patent Document 1 aims to realize a device that can accurately measure the similarity of attributes between cases expressed in natural language and can accurately search for similar cases. To solve this problem, Patent Document 1 describes a configuration that includes a database in which past cases, case numbers, category numbers, and weights for each keyword are registered, a similarity generation means that extracts keywords from problem parts of a given new case, searches the database to extract attribute information consisting of category numbers of past cases in which the same keyword exists and weights for each keyword, and generates a similarity, and a means for sorting the generated similarities in descending order and outputting the categories, similarities, etc. of the past cases.

Japanese Patent Application Publication No. 9-73464

Here, when it comes to issues in project execution, such as quality deterioration, it is necessary to evaluate the difficulty of the project. This is because the higher the difficulty (the more difficult it is), the more likely it is that issues will arise. However, in Patent Document 1, the difficulty is not taken into consideration. Also, in order to identify similar cases, materials for both the current project and the previous cases must be used. When using materials in this way, it is necessary to perform searches with high accuracy.

In view of the above, the present invention aims to evaluate the difficulty of a case by searching for related materials more accurately.

In order to solve the above problem, the present invention uses reflection materials of past quality deterioration cases to determine the difficulty of a target case. More specifically, a case management device for managing cases for creating a predetermined deliverable includes a storage unit for storing reflection materials including concern keywords of past cases, an intermediate file creation unit for processing the case materials of the case to be evaluated to create an intermediate file composed of text data, a natural language processing unit for determining the appearance position of each term included in the intermediate file using a keyword position determination dictionary that defines the appearance position of the term in a sentence and for dividing the intermediate file into words using the determined appearance positions, and a similar material search unit for creating case keywords based on the divided words and searching for similar materials from the reflection materials using the case keywords, thereby enabling evaluation of the case to be evaluated.

According to the present invention, it becomes possible to evaluate the difficulty of a case by searching for materials related to the case more accurately.

1 is a system configuration diagram of a case management system according to an embodiment of the present invention; FIG. 2 is a hardware configuration diagram of a case management server according to an embodiment of the present invention. FIG. 2 is a diagram showing an item difficulty determination rule used in one embodiment of the present invention. FIG. 2 is a diagram showing a keyword position determination dictionary used in one embodiment of the present invention. FIG. 2 is a diagram showing reflection materials used in one embodiment of the present invention. FIG. 2 is a diagram showing target case materials used in one embodiment of the present invention. FIG. 13 is a diagram showing a countermeasure/measure list used in one embodiment of the present invention. 3 is a flowchart showing an overall processing flow in one embodiment of the present invention. 4 is a flowchart showing details of natural language processing in one embodiment of the present invention. FIG. 11 is a diagram showing an example of steps S42 and S43 in one embodiment of the present invention. 1 is a flowchart showing two methods of difficulty determination in one embodiment of the present invention. FIG. 13 is a diagram showing a result report in one embodiment of the present invention.

One embodiment of the present invention will be described below with reference to the drawings. In this embodiment, a system construction project will be described as an example of a project. In a system construction project, quality deterioration such as delays in development and delivery deadlines, rework, and failure to meet required specifications can occur. When such quality deterioration occurs, reflection materials are often created that analyze the details.

Furthermore, in cases such as system construction projects, work is often done in stages. For example, a project may proceed in stages from upstream to downstream, such as pre-study, proposal, order, planning, construction, and completion. At each of these stages (tasks), project documents such as pre-study briefing materials, proposals, and specifications are created.

Here, in such cases, the further upstream the difficulty of the case can be evaluated, the easier it is to prevent and suppress quality deterioration. Furthermore, even if quality deterioration does occur, recovery is easier. For this reason, if the difficulty of the case is evaluated at a more upstream stage, it becomes possible to suppress quality deterioration, etc. In this embodiment, using a system construction project as an example, reflection materials from past cases are used to evaluate the difficulty of the target case in an ongoing case. At this time, it is desirable to perform this evaluation by comparing the reflection materials with target case materials from further upstream, such as pre-examination meeting explanatory materials. Details of this embodiment will be explained below in the order of <Configuration> <Information> <Processing flow>. Note that the materials in this embodiment include documents, papers, etc.

<Configuration>
1 is a system configuration diagram of a case management system according to the present embodiment. In FIG. 1, the case management system includes a case management server 1, a database system 20, and terminal devices 100-1 and 100-2, which are connected via a network 30. Each of these devices will be described below.

First, the case management server 1 is a device that executes the main processing of this embodiment, and can also be called a case management device. The case management server 1 has, as functional blocks, an intermediate file creation unit 11, a natural language processing unit 12, a similar document search unit 13, a matching unit 14, a case difficulty determination unit 15, a memory unit 16, and an interface unit 17. Here, the intermediate file creation unit 11 creates intermediate files composed of text data from various documents. The natural language processing unit 12 also divides text data such as intermediate files into predetermined units such as words.

The similar material search unit 13 uses keywords to search for similar materials that are similar to the target material. Here, it is preferable that the similar material search unit 13 uses keywords created based on a predetermined unit of the natural language processing unit 12. The matching unit 14 matches the above-mentioned keywords with keywords of concern contained in the searched similar materials. Furthermore, the case difficulty determination unit 15 evaluates the difficulty of the target case according to the matching result.

The storage unit 16 also stores various information, programs, and the like. In this embodiment, the storage unit 16 stores a case difficulty determination rule 161 and a keyword position determination dictionary 162. The case difficulty determination rule 161 and the keyword position determination dictionary 162 will be described later. Furthermore, the interface unit 17 communicates and connects with other devices via a connection means such as the network 30. In this embodiment, the interface unit 17 is connected to the database system 20 and the terminal devices 100-1 and 100-2.

Next, the database system 20 stores various information, and can be realized by a so-called file server. The database system 20 stores reflection materials 21, target case materials 22, and a countermeasures and measures list 23. These will be explained later, along with the information stored in the storage unit 16. The information of the database system 20 may be stored in the storage unit 16, and the case difficulty determination rules 161 and keyword position determination dictionary 162 of the storage unit 16 may be stored in the database system 20.

Next, the terminal devices 100-1 and 100-2 can be realized by computers operated by users. The terminal devices 100-1 and 100-2 access the case management server 1 to send various instructions and receive and display the processing results. Furthermore, the terminal devices 100-1 and 100-2 can also access the database system 20 to read, modify, and newly register various stored information. These terminal devices 100-1 and 100-2 can be realized by PCs, tablets, smartphones, etc. Furthermore, although two terminal devices are shown in FIG. 1, the number of terminal devices is not limited to these. Furthermore, the terminal devices 100-1 and 100-2 are installed in organizations such as companies that process cases, and their users are assumed to be case managers, task executors, etc. Furthermore, the network 30 is for communication between the devices, and can be realized by the Internet or an intranet.

Next, an implementation example of the case management server 1 that executes the main processing of this embodiment will be described. Figure 1 shows the functional blocks of the case management server 1, which can be realized by a so-called computer. This implementation example will be described below. Figure 2 is a hardware configuration diagram of the case management server 1 in this embodiment.

In the case management server 1, the processing device 101, the storage device 160, and the input/output device 170 are connected to each other via a communication path such as a bus. Here, the processing device 101 is a so-called processor, and performs processing according to each program. For this purpose, it can be realized by a so-called CPU (Central Processing Unit) or the like.

The storage device 160 also stores various programs such as the intermediate file creation program 110, the natural language processing program 120, the similar material search program 130, the matching program 140, and the case difficulty determination program 150. The storage device 160 also stores case difficulty determination rules 161 and a keyword position determination dictionary 162. In this way, it corresponds to the storage unit 16 in FIG. 1. It is also preferable that the storage device 160 is composed of a memory and a storage. In this case, it is preferable that the various programs are stored in the storage, and are expanded into the memory when processing is executed. In this way, the various programs are stored in a storage medium such as a storage.

1 have the following correspondence relationship: That is, programs and units that execute similar processes are listed below.
Intermediate file creation program 110: intermediate file creation unit 11
Natural language processing program 120: natural language processing unit 12
Similar material search program 130: Similar material search unit 13
Matching program 140: Matching section 14
Case difficulty determination program 150: Case difficulty determination unit 15
Such a program causes the case management server 1, which is a computer, to function as each of the above-mentioned units.

Next, the input/output device 170 has a function equivalent to the interface unit 17. In other words, it connects to other devices via the network 30. As described above, the case management server 1 can be realized by a computer that functions according to a program, but each unit can also be configured with dedicated hardware, etc. This concludes the explanation of the configuration of this embodiment.

<Information>
Next, various information used in the above-mentioned embodiment will be described with reference to Fig. 3 to Fig. 7. First, Fig. 3 is a diagram showing the case difficulty determination rule 161 used in the embodiment. In the embodiment, a concern keyword list 161-1 and a similarity determination list 161-2 are illustrated as examples of the case difficulty determination rule 161. Note that, of these, either one or both may be used.

First, FIG. 3(a) shows the keyword of concern list 161-1. The keyword of concern list 161-1 is used to determine the difficulty of a case in the case difficulty determination unit 15. For this purpose, the keyword of concern list 161-1 records keywords of concern that are thought to be related to quality deterioration. More specifically, the keyword of concern list 161-1 records the weight (points) and past cases for each keyword of concern. Here, the weight (points) indicates the degree of contribution of the keyword of concern to quality deterioration. Furthermore, the past cases are information that identifies past cases corresponding to the materials (reflection materials) in which the keyword of concern is recorded.

Figure 3(b) also shows the similarity judgment list 161-2. The similarity judgment list 161-2 is used to identify the degree of similarity between a case and past cases and to judge the difficulty of the case. For this purpose, the similarity judgment list 161-2 records, for each past case, the characteristics of the corresponding reflection material and the difficulty of the past case. For this purpose, the similarity judgment list 161-2 records, for each past case, the reflection material of the corresponding past case, the concern keywords contained in the reflection material, the frequency with which the concern keywords appear in the reflection material, and the difficulty of the past case. Of these, the concern keywords and the frequency of appearance indicate the characteristics of the past case.

Then, using the characteristics of the past cases, past cases similar to the target system construction project are identified, and the difficulty of the system construction project can be determined based on the difficulty of the identified past cases. The difficulty determination described here will be described later in the "Process Flow".

Next, FIG. 4 is a diagram showing the keyword position determination dictionary 162 used in this embodiment. The keyword position determination dictionary 162 is information indicating the appearance position in a sentence for each keyword, that is, each "word/symbol/phrase." The keyword position determination dictionary 162 can determine the appearance position of a keyword in the target material by using the natural language processing unit 12. For this reason, the keyword position determination dictionary 162 records the beginning, middle, and end of a sentence as the appearance position. Note that beginning of a sentence means the beginning position of a sentence, and middle/end of a sentence means the end position of a sentence. And, middle of a sentence means a position other than the beginning or end of a sentence.

In this embodiment, the keyword position determination dictionary 162 is defined for each "word/symbol/phrase", but it may also be defined by an abstract part of speech classification. In addition to general parts of speech such as nouns and verbs, more detailed parts of speech may be used for the part of speech classification. For example, proper nouns, numbers, pronouns, suffixes, etc. may be used. Furthermore, it may be defined not only by a single part of speech, but also by a combination of these, such as noun-suffix-person's name.

Next, FIG. 5 is a diagram showing reflection materials 21 used in this embodiment. Reflection materials 21 are information showing the contents of reflections in past cases, particularly in past cases of quality deterioration. In past cases, when the construction of a system is completed or delivered, materials may be created for reviewing and reflecting on the quality deterioration. Reflection materials 21 show this content, and for this purpose, reflection meetings, completion report meetings, etc. are held, and the minutes of these meetings can be used, or completion report materials can also be reused.

In FIG. 5, the reflection materials 21 are composed of "1. Project overview" which shows an overview of past cases, "2. Problem" which includes the causes of quality deterioration, and "3. Analysis and measures to prevent recurrence" which includes future policies. Note that the reflection materials 21 shown in FIG. 5 are merely an example and are not limited to this. The reflection materials 21 may be any information that shows the contents of past cases. For this reason, the reflection materials 21 are not limited to materials when the system construction is completed or delivered, but may also be related materials such as specifications from past cases. Furthermore, the reflection materials 21 can be created by the user on the terminal device 100-1 or 100-2.

Next, FIG. 6 is a diagram showing the target project documents 22 used in this embodiment. The target project documents 22 are documents used in the target project, that is, the system construction project in this embodiment. In the example shown in FIG. 6, the documents are specifications for Bank A, and the system overview and project overview are recorded. The target project documents 22 are then processed by the intermediate file creation unit 11, and an intermediate file is created. It is preferable that the target project documents 22 are stored in the database system 20 for each project, such as a system construction project. It is also assumed that there are one or more target project documents 22 for each project. The target project documents 22 can be created by the user on the terminal device 100-1 or 100-2.

Next, FIG. 7 shows the countermeasures/measures list 23 used in this embodiment. The countermeasures/measures list 23 records countermeasures/measures against quality deterioration according to keywords of concern. Using this, it is possible to propose countermeasures and proposed measures to prevent quality deterioration for the target case to the user. This concludes the explanation of the information used in this embodiment.

<Processing flow>
Next, the processing flow of this embodiment will be described mainly with reference to Fig. 8 to Fig. 12. In the following, the processing will be mainly performed by each unit shown in Fig. 1. First, Fig. 8 is a flowchart showing the overall processing flow in this embodiment.

First, in step S1, the interface unit 17 inputs the target project documents 22 for the currently underway system construction project from the database system 20. This is preferably done in response to an instruction from a user to the terminal device 100-1 or 1000-2. The terminal device 100-1 or 1000-2 may also use an optical reading device such as a scanner to read the target project documents in paper form, which are then accepted by the interface unit 17. The database system 20 may also store the target project documents 22 read by the optical reading device.

Next, in step S2, the intermediate file creation unit 11 determines whether the input target case materials 22 are text data. As a result, if the input target case materials 22 are text data (YES), the process proceeds to step S4. If the input target case materials 22 are not text data (NO), the process proceeds to step S3. At this time, it is preferable for the intermediate file creation unit 11 to determine YES if the input target case materials 22 are composed of text data only, but it may also determine YES if text data is included.

Next, in step S3, the intermediate file creation unit 11 performs text conversion on the input target case materials 22 to create an intermediate file. Note that if the determination in step S2 is NO, the input target case materials 22 are referred to as an intermediate file. Also, character recognition processing may be performed when reading with the optical reading device, and steps S2 and S3 may be omitted.

In addition, in step S4, the natural language processing unit 12 performs natural language processing on the intermediate file, which is text data. As part of this natural language processing, the natural language processing unit 12 determines the appearance position of each term contained in the intermediate file, and divides the intermediate file into words using the determined appearance positions. The keyword position determination dictionary 162 is used to determine the appearance positions.

Here, the details of the natural language processing in step S4 will be described with reference to FIG. 9. FIG. 9 is a flowchart showing the details of the natural language processing in this embodiment. In step S41, the natural language processing unit 12 accepts input of an intermediate file.

Next, in step S42, the natural language processing unit 12 unifies the received intermediate file. Then, in step S43, the natural language processing unit 12 divides the unified intermediate file into sentences. Here, the reason for executing steps S42 and S43 and the details thereof will be explained.

Figure 10 is a diagram showing an example of steps S42 and S43. Figure 10(a) shows an accepted intermediate file. As shown in Figure 10(a), this intermediate file has line breaks in the middle of sentences, dividing the sentences. For example, in line 3 of Figure 10(a), the sentence is divided by "extract." If this continues, there is a risk that case keywords for searching for similar documents from the target case documents 22 or the intermediate file will not be accurately extracted. Therefore, in steps S42 and S43, the intermediate file is formatted and cleansed. A specific example of this will be described below with reference to Figure 10.

In step S42, the natural language processing unit 12 performs the following on the intermediate file of FIG.
Line breaks are deleted. In addition, periods, commas, symbols, etc. are left as they are. As a result, the unified intermediate file shown in FIG. 10(b) is created.

Then, in step S43, the natural language processing unit 12 determines the position of occurrence of each word in the unified intermediate file shown in FIG. 10(b), and divides it into sentences. This division can be achieved by distinguishing between each sentence, inserting line breaks between sentences (between the end and beginning of a sentence), or identifying the beginning or end of a sentence (full stop). As a result, an intermediate file divided into sentence units, as shown in FIG. 10(c), is created.

For this purpose, the natural language processing unit 12 uses the keyword position determination dictionary 162 to determine the position where each word appears. In the example of FIG. 10(b), by using the keyword position determination dictionary 162 of FIG. 4, "." is determined to be the end of the sentence, and "..", "(1)," and "(2)" are determined to be the beginning of the sentence. As a result, the sentence is separated before the ".." on line 1 and before the (2) on line 5 in FIG. 10(b). In other words, a line break is inserted here, and an intermediate file divided into sentence units as shown in FIG. 10(c) is created. This concludes the explanation of steps S42 and S43.

Finally, in step S44, the natural language processing unit 12 divides the intermediate file divided into sentence units into word units. This only requires that each word can be extracted from the intermediate file divided into sentence units. This concludes the detailed explanation of step S4, and we will return to Figure 8 to continue the explanation.

Next, in step S5, the similar material search unit 13 searches for similar materials in the reflection materials 21 based on the words extracted in step S44. To this end, the similar material search unit 13 identifies case keywords from the words extracted in step S44. For this identification, the extracted words may be used as they are, or some of the words may be used as case keywords. When some words are used as case keywords, a specific type of word such as an autonomous word may be used, or a pre-recorded word may be used. The similar material search unit 13 then searches for the reflection materials 21 using the identified case keywords. This search searches for reflection materials 21 that satisfy a specified condition, such as searching for reflection materials 21 that contain a specified number or more of case keywords. The search results are referred to as similar materials.

Next, in step S6, the matching unit 14 executes a matching process. To this end, the matching unit 14 matches the case keywords with the concern keywords contained in the searched similar materials. In other words, the matching unit 14 identifies words contained in the similar materials that match the case keywords among the concern keywords recorded in the concern keyword list 161-1 and the similarity determination list 161-2.

Next, in step S7, the project difficulty determination unit 15 uses the results of step S6 to determine the difficulty of the system construction project, which is the target project. The details of this determination are described with reference to FIG. 11.

Figure 11 is a flowchart showing two methods of difficulty determination in this embodiment. First, the first method shown in Figure 11 (a) will be described. In step S71, the case difficulty determination unit 15 counts the number of concern keywords identified in step S6.

In addition, in step S72, the case difficulty assessment unit 15 uses the concern keyword list 161-1 to identify the weight (points) of the concern keyword identified in step S6. For example, for the concern keyword "new customer," a weight (point) of "2" is identified.

Then, in step S73, the case difficulty determination unit 15 multiplies the number counted in step S71 by the weight (points) identified in step S72, and calculates the result as the difficulty of the target case. Note that number x weight (points) is just an example, and the difficulty of the target case is not limited to this. For example, other parameters may be included, and the calculation is not limited to accumulation. Furthermore, the number or weight (points) itself may be used as the difficulty of the target case.

Next, the second method shown in FIG. 11(b) will be described. In step S74, the case difficulty determination unit 15 determines the occurrence frequency of the concern keywords identified in step S6 using the similarity determination list 161-2. That is, the occurrence frequency corresponding to the concern keywords that correspond to the reflection materials, which are similar materials, is determined in the similarity determination list 161-2.

Next, in step S75, the case difficulty determination unit 15 identifies, from among the occurrence frequencies identified in step S74, occurrence frequencies that are similar to the occurrence frequency of the keyword of concern in the target case materials 22 input in step S1. Similar occurrence frequencies include occurrences where the difference is equal to or less than a certain value, including coincidences.

Next, in step S76, the case difficulty determination unit 15 identifies the difficulty level corresponding to the occurrence frequency identified in step S75 from the similarity determination list 161-2. Then, in step S77, the case difficulty determination unit 15 determines the difficulty level of the target case based on the difficulty level identified in step S76. For example, the case difficulty determination unit 15 calculates a representative value of the difficulty levels identified in step S76. Representative values include sums and averages, and these are determined as the difficulty level. This completes the explanation of the second method. Note that, in this method, the case difficulty determination unit 15 may use the similarity determination list 161-2 to identify the most similar reflection material 21 based on the occurrence frequency, and determine the difficulty level of the corresponding past case as the difficulty level of the target case.

This concludes the detailed explanation of step S7, and we will return to Figure 8 for further explanation. In step S8, the case difficulty determination unit 15 identifies countermeasures and measures using the countermeasures and measures list 23. In other words, the case difficulty determination unit 15 identifies countermeasures and measures that correspond to the concern keywords identified in step S6.

Then, the case difficulty determination unit 15 can create a result report 200 using the results of steps S7 and S8. FIG. 12 is a diagram showing the result report 200 in this embodiment. In FIG. 12, the result report 200 records, for each system construction project that is the target case, its difficulty level, the keywords of concern identified in step S6, and the measures and policies identified in step S8. In addition, it is preferable that the case difficulty determination unit 15 stores the result report 200 in the storage unit 16 or the database system 20.

Furthermore, the case difficulty determination unit 15 can also display the result report 200 on the terminal device 100-1 or 100-2 via the interface unit 17. In this case, it can be displayed in the format shown in FIG. 12. As a result, the users involved in the target case can understand the difficulty level and the measures and policies. Note that in step S8, the identification of the measures and policies may be omitted. In this case, the measures and policies in the result report 200 can be omitted. Furthermore, the measures and policies in the result report 200 may be retained, and the difficulty level and concern keywords may be omitted.

As described above, in this embodiment, the difficulty level of the target system construction project is determined as one evaluation based on the similar documents searched for in step S5. Therefore, it is possible to omit steps S6 and S7. In this case, it is possible to realize evaluation of the number of similar documents and the difficulty level determined for each similar document.

This concludes the explanation of this embodiment, but the present invention is not limited to this embodiment. In particular, the case is not limited to a system construction project, but can be applied to a variety of business operations.

1...Case management server, 11...Intermediate file creation unit, 12...Natural language processing unit, 13...Similar material search unit, 14...Matching unit, 15...Case difficulty determination unit, 16...Storage unit, 161...Case difficulty determination rules, 162...Keyword position determination dictionary, 17...Interface unit, 20...Database system, 21...Reflection materials, 22...Target case materials, 23...Countermeasures and measures list, 30...Network, 100...Terminal device

Claims (15)

In a case management device for managing a case for creating a predetermined deliverable,
A storage unit that stores reflection materials including keywords of concern from past cases;
an intermediate file creation unit that processes case materials of a case to be evaluated and creates an intermediate file composed of text data;
a natural language processing unit that determines an appearance position of each term included in the intermediate file using a keyword position determination dictionary that defines an appearance position of the term in a sentence, and divides the intermediate file into words using the determined appearance positions;
a similar material search unit that creates case keywords based on the divided words and searches for similar materials from the reflection materials using the case keywords;
A case management device for implementing evaluation of the case to be evaluated. 2. The case management device according to claim 1,
the keyword position determination dictionary defines the position of the term as being at the beginning, middle, or end of the sentence;
The natural language processing unit divides the intermediate file into sentence units using the keyword position determination dictionary, and divides the divided intermediate file into word units. 3. The case management device according to claim 2, further comprising:
a matching unit for matching the case keyword with a keyword of concern contained in the searched similar material;
A case management device having a case difficulty determination unit that determines the difficulty of the case to be evaluated based on the result of the matching. 4. The case management device according to claim 3,
The case difficulty determination unit determines the difficulty of the case to be evaluated using at least one of the number of matches between the case keywords and the concern keywords and a weight. 5. The case management device according to claim 3,
The storage unit further stores measures and policies corresponding to the keywords of concern,
The case management device further comprises an interface unit that outputs a result report that associates the degree of difficulty of the case to be evaluated with measures and policies. A case management method using a case management device for managing a case for creating a predetermined deliverable,
The memory unit stores reflection materials including keywords of concern from past cases,
The intermediate file creation unit processes the case materials of the case to be evaluated to create an intermediate file consisting of text data;
a natural language processing unit determines an appearance position of each term included in the intermediate file using a keyword position determination dictionary that defines an appearance position of the term in a sentence, and divides the intermediate file into words using the determined appearance positions;
A case management method in which a similar document search unit creates case keywords based on the divided words, and searches for similar documents from the reflection documents using the case keywords, thereby enabling evaluation of the case to be evaluated. 7. The case management method according to claim 6,
the keyword position determination dictionary defines the position of the term as being at the beginning, middle, or end of the sentence;
The case management method further comprises the steps of: dividing the intermediate file into sentence units by the natural language processing unit using the keyword position determination dictionary; and dividing the divided intermediate file into word units. 8. The case management method according to claim 7,
A matching unit matches the case keyword with a keyword of concern contained in the searched similar materials;
A case management method in which a case difficulty determination unit determines the difficulty of the case to be evaluated based on the result of the matching. 9. The case management method according to claim 8,
The case management method further comprises the case difficulty determination unit determining the difficulty of the case to be evaluated using at least one of the number of matches between the case keywords and the concern keywords and a weight. 10. The case management method according to claim 8, further comprising:
The storage unit further stores measures and policies corresponding to the keywords of concern,
The case management method includes outputting a report by an interface unit that correlates the degree of difficulty of the case to be evaluated with the countermeasures and measures. A case management device that has a storage unit that stores reflection materials including concern keywords of past cases and manages cases for creating a predetermined deliverable,
an intermediate file creation unit that processes case materials of a case to be evaluated and creates an intermediate file composed of text data;
a natural language processing unit that determines an appearance position of each term included in the intermediate file using a keyword position determination dictionary that defines an appearance position of the term in a sentence, and divides the intermediate file into words using the determined appearance positions;
creating case keywords based on the divided words, and functioning as a similar material search unit that searches for similar materials from the reflection materials using the case keywords;
A program for implementing evaluation of the evaluation target project. The program according to claim 11,
the keyword position determination dictionary defines the position of the term as being at the beginning, middle, or end of the sentence;
The natural language processing unit is a program that divides the intermediate file into sentence units using the keyword position determination dictionary, and divides the divided intermediate file into word units. 13. The program according to claim 12, further comprising:
a matching unit that matches the case keyword with a keyword of concern contained in the searched similar material;
A program that functions as a case difficulty determination unit that determines the difficulty of the case to be evaluated based on the result of the matching. The program according to claim 13,
The case difficulty determination unit is a program that determines the difficulty of the case to be evaluated using at least one of the number of matches between the case keywords and the concern keywords and a weight. 15. The program according to claim 13 or 14,
The storage unit further stores measures and policies corresponding to the keywords of concern,
Further, the program causes the case management device to function as an interface unit that outputs a result report that associates the degree of difficulty of the case to be evaluated with measures and policies.

Images