JP6947155B2 - Information retrieval system - Google Patents

Description

The present invention relates to an information retrieval system that retrieves and outputs information requested by a searcher from a database that stores a plurality of pieces of information.

Conventionally, data in maintenance work in the manufacturing industry and the like has been data (paper data) written on paper. Although the digitization of materials has progressed due to the spread of personal computers (PCs), some materials such as drawings are still paper data. In an information retrieval system in maintenance work, it is important to digitize these paper data, store them together and create a database, but the amount of data is enormous and the labor required to aggregate the scattered data is also great. Therefore, it was unrealistic to digitize the paper data of the materials. However, advances in search technology have improved the accuracy of search results for search targets and increased the cost-effectiveness of digitizing paper data, making digitization practical.

For example, in the manufacturing industry, malfunctions and malfunctions of equipment, their causes, and countermeasures are stored in a database each time, and when similar malfunctions or malfunctions occur, they are referred to as past cases to avoid defects. We are trying to shorten the time to recovery. Information such as work standards and manuals is also stored in the database. Therefore, the information stored in the database becomes enormous.

Conventionally, as a method of obtaining necessary information from such a huge amount of data, there is a text search process in which a given character string is used as a keyword and a matching one is output as a search result.

Since the above-mentioned conventional method is a match search with a given keyword, there is a problem that a desired search result cannot be obtained unless the search keyword is properly input. For example, even in the case of maintenance work, the terms used there include technical terms, terms specific to the site, terms specific to the region, etc., and an expert can enter appropriate keywords. It is difficult for a person with little knowledge or experience to come up with an appropriate keyword, it is difficult for useful information to appear in the search results, and information other than useful information often appears.

An object of the present invention is to solve the above-mentioned problems of the prior art and to provide an information retrieval system capable of obtaining good search results even by a person who has little knowledge and experience and who has difficulty in inputting appropriate keywords.

Information retrieval system of the present invention, for achieving the above object, a question sentence for accepting a database for storing a plurality of information can be searched for work information in a plurality of manufacturing sites with similar equipment, the question sentence of a natural language form The reception unit, the input search keyword extraction unit that extracts the input search keyword from the question text received by the question sentence reception unit, the keyword database in which related keywords are associated with each other and registered, and the keyword dictionary. From the keywords registered in, the search candidate keyword reading unit that reads the keyword related to the input search keyword as the search candidate keyword, and the search candidate keyword read by the search candidate keyword reading unit are the input search keyword. In addition, as a search keyword, a search execution unit that performs a search process on the database is provided.

The information retrieval system of the present invention also includes a search keyword storage unit that cumulatively stores the search keywords.
When a question sentence is input a plurality of times in a series of search operations, the search candidate keyword read by the search candidate keyword reading unit is accumulated in the search keyword storage unit as a search keyword each time.
The search execution unit is characterized in that a search process is performed on the database using an arbitrarily selected part of a plurality of search keywords stored in the search keyword storage unit.

The information retrieval system of the present invention further includes an information type determination unit that determines the information type that is the search intention from the question text received by the question text reception unit.
The information type determined by the information type determination unit is configured to be changeable.
The database stores the plurality of pieces of information in a plurality of groups according to the information type.
The search execution unit uses an arbitrarily selected part of the plurality of search keywords stored in the search keyword storage unit to determine by the information type determination unit among the plurality of groups in the database. The feature is that the search process is performed on the group corresponding to the changed information type or the changed information type.

In the information retrieval system of the present invention, it is preferable that keywords having the same meaning as each other are registered in the keyword dictionary in association with each other.

In the information retrieval system of the present invention, it is preferable that keywords having synonyms with each other are registered in the keyword dictionary in association with each other.

In the information retrieval system of the present invention, the information stored in the database includes information on coping guidance for complaints, complains or inquiries from customers, and information for presenting optimal coping guidance from the database is provided. You may try to search.

In the information retrieval system of the present invention, the information stored in the database includes information on product design guidelines, related technical papers, and references for customer requests and inquiries, and the database provides appropriate customer answers and optimums. You may want to search for information to present a good product design.

According to the information retrieval system of the present invention, the database stores equipment maintenance work information at a plurality of manufacturing bases having similar facilities as a plurality of text-retrievable information divided into a plurality of groups according to the information type, and is a search candidate. The keyword reading unit reads the keywords related to the input search keyword extracted from the question text input by the searcher from the keyword database as search candidate keywords, and the search execution unit reads the keywords as the above search candidate keywords in addition to the input search keywords. Based on the read keywords, a match search is performed for the group corresponding to the information type determined by the information type determination unit or the changed information type among the plurality of groups in the database, so that an appropriate search keyword can be obtained. Even if it is not input, it is possible to supplement it and search for accurate equipment maintenance work information, it is easy to use even for inexperienced searchers, and it is possible to shorten the work time and efficiently use the information.
Moreover, since it is equipped with a search keyword storage unit that cumulatively stores search keywords, if a search cannot be performed with one question sentence in a series of search operations, input with a different question sentence is performed twice or three times. Repeatedly, every time a question sentence is input, the search candidate keywords read by the search candidate keyword reading unit are accumulated as search keywords in the search keyword storage unit, and the search execution unit accumulates a large amount in this way. A group corresponding to the information type determined by the information type determination unit or the changed information type among a plurality of groups in the database using a part of the searched keywords arbitrarily selected by the searcher. Since you can perform a search process with a large number of keywords including some of multiple search keywords , you can enter different question sentences multiple times and exclude unnecessary search keywords until you get the information you need. It will be possible to narrow down the search to the necessary equipment maintenance work information.

It is a block diagram which shows the schematic structure of the information retrieval system of one Embodiment of this invention. (A) is a diagram showing an example of a keyword stored in a keyword dictionary, and (b) is a diagram showing an example of an information type determination keyword stored in an information type determination dictionary. It is a flowchart for demonstrating operation of the information retrieval system of embodiment. It is a functional block diagram of a computer.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In this embodiment, equipment maintenance work information in the manufacturing industry, for example, information such as equipment failures and malfunctions and their causes, work standards such as countermeasures, and manuals, etc., is stored in a database as text searchable information. , When a failure or malfunction occurs, the searcher inputs the situation as a sentence, searches the database for information that can be used as a reference for past cases, and based on the searched information, the searcher performs equipment failure. A system for performing recovery work will be described as an example. However, the search application of the information retrieval system of the present invention is not limited, and it goes without saying that it can be used for purposes other than supporting equipment maintenance work such as equipment failure recovery work. As another application example, in the sales field, in order to improve customer satisfaction, it can be used to search past cases for the contents of inquiries from customers and to present guidance on optimal measures. In the field of product design, in order to respond to inquiries from customers and design products in response to customer requests, similar manufacturing results and documents such as technical treatises are searched as reference information, and the searcher is appropriate. It can be used when answering the question or designing the optimum product.

FIG. 1 is a block diagram showing a schematic configuration of an information retrieval system according to an embodiment of the present invention. The information retrieval system mainly includes an information retrieval device 10 that performs search processing, a database 12 that stores a plurality of (large amounts) of information to be searched, and an input device 14 such as a keyboard or mouse connected to the information retrieval device 10. A display device 16 such as a display connected to the information retrieval device 10, and a searcher-side terminal 20 connected to the information retrieval device 10 via the network 18 are provided. The network 18 is a communication network managed by a user's organization such as a LAN (Local Area Network). However, the network 18 is not limited to this, and the network 18 is a communication network in which a general public line such as WAN (Wide Area Network) or VAN (Virtual Private Network) is partially interposed in addition to a public communication network such as the Internet. You may. The information retrieval device 10 and the searcher-side terminal 20 are realized by a computer and a program, respectively. An example of the searcher side terminal 20 is a personal computer (PC) or a tablet computer, in which the input device 22 and the display device 24 are connected or integrated.

Documents, drawings, photographs and the like are stored in the database 12. The text may be not only the text input from the beginning, but also the text of the voice using the voice recognition engine, or the text of the text in the paper medium using OCR (Optical Character Reader). good. When storing a drawing or a photograph, a search process can be performed on a character or a file name embedded in the drawing. The information stored in the database 12 includes information such as work standards and manuals such as failures and malfunctions of devices and their causes, as well as countermeasures. In another application of the information retrieval system of the present invention, the information stored in the database 12 includes information regarding coping guidance for complaints, complains or inquiries from customers, in which case, according to the information retrieval system, the database. Information for presenting the optimum coping guidance can be searched from 12. In still another application of the information retrieval system of the present invention, the information stored in the database 12 includes information on product design guidelines, related technical papers, and references for customer requests and inquiries, and in this case, the information retrieval system. According to this, it is possible to search the database 12 for information for presenting appropriate customer answers and optimal product designs.

The information retrieval device 10 mainly includes a question sentence reception unit 26, a natural language processing unit 28, a keyword dictionary 30, a search candidate keyword reading unit 32, a search keyword storage unit 34, an information type determination unit 36, and information. It is composed of a type determination dictionary 38, a search execution unit 40, and a search candidate keyword change request reception unit 42. The natural language processing unit 28 includes an input search keyword extraction unit 44 and an information type determination keyword extraction unit 46.

Each of these functional blocks has the following functions. The question text reception unit 26 sequentially receives question texts in natural language format input by the searcher via the input device 22 of the searcher side terminal 20 via the network 18. On the searcher side terminal 20, for example, "A failure of the XX device has occurred in the XX equipment. Is there a similar example in the past?", "The type of failure is △△ failure", "Replacement of □□ parts" Natural language format question sentences such as "Procedure" can be input sequentially.

The natural language processing unit 28 performs natural language processing such as known morphological analysis on the question sentence received by the question sentence receiving unit 26, and extracts a word (independent word). In particular, the input search keyword extraction unit 44 of the natural language processing unit 28 extracts the input search keyword (search keyword input by the searcher) to be used later for the search from the input question text. For this extraction, the input search keyword extraction unit 44 can refer to, for example, the keyword dictionary 30 described in detail later, and extract words that match the words registered in this dictionary as input search keywords. Alternatively, it is also possible to refer to another dictionary (not shown) and extract words that match the words registered in this dictionary as input search keywords. In the above example, the input search keyword extraction unit 44 indicates "XX equipment" and "XX equipment" and "XX equipment" when the question sentence "A failure of the XX device has occurred in the XX equipment. Is there a similar example in the past?" Is input. Extract the word "XX device" as an input search keyword. Similarly, when the question sentence "The type of failure is △△ failure" is input, the word "△△ failure" is extracted as an input search keyword. In addition, when the question sentence "How to replace □□ parts" is input, the word "□□ parts" is extracted as an input search keyword.

On the other hand, the information type determination keyword extraction unit 46 of the natural language processing unit 28 selects the information type determination keyword that is later used to determine the information type that is the search intention (purpose) from the input question text. Extract. For this extraction, the information type determination keyword extraction unit 46 refers to the information type determination dictionary 38, which will be described in detail later, and uses words that match the words registered in the dictionary 38 as information type determination keywords. Can be extracted. Alternatively, it is also possible to refer to another dictionary (not shown) and extract a word that matches the word registered in this dictionary as a keyword for determining the information type. In the above example, the information type determination keyword extraction unit 46 indicates "failure" and "failure" when the question sentence "A failure of the XX device has occurred in the XX equipment. Is there a similar example in the past?" Is input. The word "similar example" is extracted as a keyword for determining the type of information. Similarly, when the question sentence "The type of failure is △△ failure" is input, the word "type of failure" is extracted as a keyword for determining the information type. In addition, when the question sentence "What is the replacement procedure for □□ parts?" Is input, the word "replacement procedure" is extracted as a keyword for determining the type of information.

In the keyword dictionary 30, related keywords are associated with each other and are registered and stored in advance. In this embodiment, keywords having the same meaning as each other are registered in the keyword dictionary 30 in association with each other. Further, in the keyword dictionary 30, keywords having synonyms with each other are registered in association with each other. However, in the keyword dictionary 30, only keywords synonymous with each other may be registered, or only keywords synonymous with each other may be registered. Alternatively or additionally, keywords that are considered to be highly related to each other can be associated with each other and registered in advance.

FIG. 2A shows an example of keywords registered in advance in the keyword dictionary 30. In FIG. 2B, keywords belonging to the same row (horizontal column) are related to each other and are associated with each other.

Returning to FIG. 1, the search candidate keyword reading unit 32 refers to the keyword dictionary 30 each time the input search keyword extraction unit 44 extracts the input search keyword, and reads out the keyword related to the input search keyword as the search candidate keyword. According to the above example, when the input search keyword is "○○ equipment", the search candidate keyword reading unit 32 performs a match search for the keyword dictionary 30 illustrated in FIG. 2A, and includes "○○ equipment". All other keywords belonging to the line (first line in the figure) are read out, or if specified (for example, only synonyms), the specified keyword is read as a search candidate keyword. In this example, "AA equipment", "A'A' equipment", "A" A "" ... Are read as synonyms for "○○ equipment", and "aaa", "a'a" are synonyms. 'a' "is read out. Similarly for "XX device", "△△ failure", and "□□ part" in the above example, related keywords are read out as search candidate keywords.

Returning to FIG. 1, the search keyword storage unit 34 stores the input search keyword obtained as described above and the keyword as the search candidate keyword as the search keyword. The search keyword storage unit 34 cumulatively stores the search keyword each time a question sentence is additionally input and an additional search keyword is extracted.

The information type determination unit 36 determines the information type that is the search intention based on the information type determination keyword extracted from the question text. The information type determination unit 36 refers to the information type determination dictionary 38 and determines the information type to be searched based on the information type determination keyword. FIG. 2B shows an example of the information type determination keyword stored in the information type determination dictionary 38. The information type determination dictionary 38 is registered in advance in a state in which an information type determination keyword that is assumed to be extracted from a question sentence and an information type corresponding to the keyword are associated with each other. For example, in the information type determination dictionary 38, "failure case" is used as the information type for the information type determination keywords "failure" and "similar example", and "fault type" is used for the information type determination keyword. The "manual" is registered as the information type, and the "manual" is registered as the information type in advance for the information type determination keyword "exchange procedure".

Further, in this embodiment, the information stored in the database 12 is divided into a plurality of groups A, B, C ... In advance according to the type corresponding to the information type determined by the information type determination unit 36. .. For example, in the example of FIG. 1, the group A of the first information type stores the information related to the "failure case", the group B of the second information type stores the information related to the "manual", and the group B of the third information type stores the information related to the "manual". Information about the "drawing" is stored in the group C.

The search execution unit 40 performs a text-based match search process on the information stored in the database 12 with the search keyword stored in the search keyword storage unit 34. In an advantageous form, the search execution unit 40 refers to the information in the group corresponding to the information type determined by the information type determination unit 36 among the plurality of groups A, B, C ... In the database 12. , It is configured to perform a text-based match search process with the above search keywords.

The search candidate keyword change request receiving unit 42 accepts the addition and / or deletion of words registered in the keyword dictionary 30 from the searcher.

The operation of this information retrieval system will be described below. FIG. 3 is a flowchart for explaining the processing operation of the information retrieval system.

First, the question text reception unit 26 of the information search device 10 receives the question text input by the searcher on the searcher side terminal 20 via the input device 22 (step S101). The question text reception unit 26 is a question text that the searcher inputs via the input device 22 on the searcher side terminal 20, for example, "A failure of the XX device has occurred in the XX equipment. Is there a similar example in the past?" Accept.

Subsequently, the natural language processing unit 28 performs natural language processing such as morphological analysis on the question sentence received by the question sentence receiving unit 26, and extracts a word (independent word) (step S102). In this example, from the question text "A failure of XX equipment has occurred in XX equipment. Is there a similar example in the past?", "XX equipment", "XX equipment", "Failure", "Occurrence", " "Past" and "similar examples" are extracted.

Subsequently, the input search keyword extraction unit 44 extracts (selects) the input search keyword from the words extracted by the natural language processing (step S103). In this example, the input search keyword extraction unit 44 refers to the keyword dictionary 30 illustrated in FIG. 2A, and inputs the words "○○ equipment" and "XX device" included in the dictionary 30. Extract as a search keyword. Further, the information type determination keyword extraction unit 46 extracts (selects) the information type determination keyword from the words extracted by the natural language processing (step S103). In this example, the information type determination keyword extraction unit 46 refers to the information type determination dictionary 38, and extracts the words "failure" and "similar example" contained in the information type determination dictionary 38 as information type determination keywords ( select.

Then, the search candidate keyword reading unit 32 refers to the keyword dictionary 30 and reads a keyword related to the input search keyword as a search candidate keyword (step S104). In this example, referring to the keyword dictionary 30 illustrated in FIG. 2A, "AA equipment" and "A" included in the first line to which the input search keyword "○○ equipment" extracted in step S103 belongs. "A'equipment", "A" A "", "aaa", "a'a'a'" ..., and the input search keyword "XX device" extracted in step S103 also belongs to the second. "BB device", "B'B' gear", "B" B "", "bbbb", "b'b'b'", etc. included in the second line are read out as search candidate keywords.

The search keyword storage unit 34 stores the input search keyword and the search candidate keyword read by the search candidate keyword reading unit 32 (step S105). In this example, the search keyword storage unit 34 uses "○○ equipment", "AA equipment", "A'A' equipment", "A" A "", "aaa", "a'a'a'", "XX device", "BB device", "B'B' gear", "B" B "", "bbbb", "b'b'b'" ... are memorized.

The information search device 10 causes the display device 24 to display the search keywords stored in the search keyword storage unit 34, and if necessary, the searcher searches for a search keyword that the searcher does not want among the displayed plurality of search keywords. Can be excluded (step S106).

Further, the information type determination unit 36 refers to the information type determination dictionary 38 illustrated in FIG. 2B, and bases the information type (search) based on the information type determination keyword extracted by the information type determination keyword extraction unit 46. The intention) is determined (step S107). In this example, since the information type "failure case" is associated with the combination of "failure" and "similar example" and registered in advance in the information type determination dictionary 38, the information type determination unit 36 Determines the type of information requested by the searcher as a "failure case". The information retrieval device 10 may be configured so that the information type can be manually changed or added by the searcher when the information type determined by the information type determination unit 36 is inappropriate or insufficient. Note that this step S107 may be performed before or at the same time as steps S104 to S106.

Then, the search execution unit 40 is selected for the information stored in the group corresponding to the information type determined by the information type determination unit 36 or the information type manually selected by the searcher in the database 12. A text-based match search process with the search keyword is performed (step S108). At this time, when the information type is a drawing, the search execution unit 40 can search for the file name attached to the drawing and / or the character information that can be embedded as text data in the drawing. The information retrieval device 10 displays the search result on the display device 24 on the searcher side (step S109).

By the way, it is also possible that the keywords registered in the keyword dictionary 30 are insufficient or the registered keywords are associated with each other incorrectly. In such a case, the search candidate keyword change request receiving unit 42 of the information retrieval device 10 can accept such a request and notify the administrator of the information retrieval system of the request through the display device 16. The administrator can update the keyword dictionary 30 after determining the eligibility of the request.

The information retrieval device 10 and the searcher-side terminal 20 are computers as shown in FIG. 4, respectively, and are connected to a memory (storage device), a CPU (processing device), a hard disk drive (HDD), and a display device 16 or 24. The display control unit, the communication control unit for connecting to the network 18, and the like are connected via a bus. The operating system (OS) and the application program for performing the processing in the above embodiment can be stored in the HDD, and when executed by the CPU, are read from the HDD into the memory. If necessary, the CPU controls the display control unit, the communication control unit, and the like to perform necessary operations. Further, the data in the process of processing is stored in the memory, and if necessary, is stored in the HDD. The various functions described above are realized by organically linking hardware such as a CPU and memory with an OS and necessary application programs.

As described above, according to the information search system of this embodiment, the search candidate keyword reading unit 32 searches the keyword dictionary 30 for keywords related to the input search keyword extracted from the question text input by the searcher. In addition to the input search keyword, the search execution unit 40 performs a match search on the database 12 based on the read search candidate keyword. Therefore, even if an appropriate search keyword is not input, it is supplemented and accurate. Search can be performed, it is easy to use even for inexperienced searchers, work time can be shortened, and information can be used efficiently. At the same time, it also helps the searcher's learning by presenting related search candidate keywords. Then, based on the information searched and displayed in this way, the searcher can perform the failure recovery work of the equipment.

According to the information retrieval system of this embodiment, since keywords having synonyms with each other are registered in the keyword dictionary 30 in association with each other, synonyms of input search keywords can be automatically added to the search keywords. It will be possible.

Further, according to the information retrieval system of this embodiment, since keywords synonymous with each other are registered in the keyword dictionary 30 in association with each other, not only the input search keyword but also the synonyms thereof are searched. become able to.

Further, according to the information retrieval system of this embodiment, since the search keyword storage unit 34 that cumulatively stores the search keywords is provided, in a series of search operations for obtaining necessary (search target) information, in a series of search operations. If the question text cannot be searched with one question text, the input with a different question text will be repeated twice or three times, but at that time, each time a different question text is added. Search keywords are accumulated. In this way, by using a large number of accumulated search keywords and including all or some of the plurality of search keywords from the database, information close to the search keywords is extracted. As the search keywords are accumulated, you can search by many keywords using all the accumulated keywords or some of the keywords, so enter different question sentences multiple times until you get the information you need. Then, the necessary information is gradually approached, and a more accurate refined search becomes possible.

Further, according to the information retrieval system of this embodiment, the information type determination unit 36 that determines the information type that is the search intention from the question text received by the question text reception unit 26 is provided, and the database 12 contains a plurality of types of information. Separately, the information is stored separately in a plurality of groups A, B, C ..., And the search execution unit 40 is determined by the information type determination unit 36 among the plurality of groups A, B, C ... Of the database 12. Since it is configured to perform a text search for a group corresponding to an information type, it is possible to efficiently and accurately obtain information that matches the search intention of the searcher, that is, information of the type desired by the searcher.

Further, according to the information retrieval system of this embodiment, it is possible to arbitrarily select whether or not to include the search keyword stored in the search keyword storage unit 34 in the search target, so that unnecessary search keywords can be selected. By excluding it from the search target, it is possible to efficiently obtain only the search results desired by the searcher.

Further, according to the information retrieval system of this embodiment, the keyword dictionary 30 includes a search candidate keyword change request receiving unit 42 that receives a request for adding a keyword to the keyword dictionary 30 or deleting a keyword from the keyword dictionary 30. The number of registered words can be increased to enable more accurate and efficient search, and the check function by the searcher works for each use, so that the accuracy of the search can be improved. Further, since it is possible to prevent the searcher from changing the contents of the keyword dictionary 30, the accuracy of the search can be ensured.

In particular, in the manufacturing industry, similar equipment is often provided in a factory of another manufacturing process or manufacturing base. However, according to the information retrieval system of this embodiment, between those manufacturing processes or manufacturing bases. It is also possible to search for equipment maintenance work information.

Although the present invention has been described above based on the illustrated examples, the present invention is not limited to the above-described embodiment, and various improvements and modifications can be made without departing from the gist of the present invention.

According to the present invention, it is possible to provide an information retrieval system capable of obtaining good search results even by a person who has little knowledge and experience and who has difficulty in inputting an appropriate keyword.

10 Information retrieval device 12 Database 14 Input device 16 Display device 18 Network 20 Searcher side terminal 22 Input device 24 Display device 26 Question text reception unit 28 Natural language processing unit 30 Keyword dictionary 32 Search candidate keyword reader 34 Search keyword storage unit 36 Information type judgment unit 38 Information type judgment dictionary 40 Search execution department 42 Search candidate keyword change request reception department 44 Input search keyword extraction unit 46 Information type judgment keyword extraction unit

Claims (3)

A database for storing a plurality of information can be searched for work information in a plurality of manufacturing sites with similar facilities,
The question text reception department that accepts question texts in natural language format,
An input search keyword extraction unit that extracts input search keywords from the question sentences received by the question sentence reception unit, and an input search keyword extraction unit.
A keyword dictionary in which related keywords are associated with each other and registered,
A search candidate keyword reading unit that reads keywords related to the input search keyword as search candidate keywords from the keywords registered in the keyword dictionary.
A search execution unit that performs a search process on the database by adding the search candidate keyword read by the search candidate keyword reading unit to the input search keyword and using the search keyword as a search keyword.
A search keyword storage unit that cumulatively stores the search keywords,
The information type judgment unit that determines the information type that is the search intention from the question text received by the question text reception department,
With
The information type determined by the information type determination unit is configured to be changeable.
The database stores the plurality of pieces of information in a plurality of groups according to the information type.
When a question sentence is input a plurality of times in a series of search operations, the search candidate keyword read by the search candidate keyword reading unit is accumulated in the search keyword storage unit as a search keyword each time.
The search execution unit uses an arbitrarily selected part of the plurality of search keywords stored in the search keyword storage unit to determine by the information type determination unit among the plurality of groups in the database. An information retrieval system characterized in that a search process is performed on a group corresponding to a changed information type or a changed information type. The information retrieval system according to claim 1 , wherein keywords having the same meaning as each other are registered in the keyword dictionary in association with each other. The information retrieval system according to claim 1 or 2 , wherein keywords having synonyms with each other are registered in the keyword dictionary in association with each other.

Images