Information retrieving device and information management device

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Technical field to which the invention belongs]

The present invention relates to, for example, an information retrieval device capable of efficiently searching various accumulated information. The present invention also relates to an information management device capable of accumulating and managing various types of information.

【0002】

[Conventional technology]

In recent years, with the development of various information devices centered on computers and communication networks, various digital information has been distributed via networks. For example, digitized e-mail (e-mail) or the like is used for normal communication or the like. In addition, various information such as product information and technical information is provided on the WWW (World-Wide Web) server, and the user can browse such information via a web browser.

【0003】

Further, digital information is distributed and sold not only through the above-mentioned communication media but also through tangible recording media. For example, various documents and image data may be stored and distributed on a medium such as a CD-ROM (Compact Disc-ROM) or a DVD-ROM (Digital Video Disc-ROM).

【0004】

Since these various types of digital information are easier to store, process, edit, etc. than conventional analog information, business efficiency can be improved by using the digital information.

【0005】

[Problems to be Solved by the Invention]

As described above, as communication networks such as the Internet and intranets become widespread, information is exchanged between companies, individuals, groups, and the like via digital information. However, in the conventional system, various digital information is not centrally managed, and each user often manages the information individually. E-mail can be mentioned as a typical example. Aside from private e-mail, even business-related e-mail is managed individually by the sender and recipient, and is not a shared property of the company or organization. If valuable knowledge information contained in e-mail is managed individually and in a disorderly manner within the company or organization, the accumulation and use of knowledge information within the company will not progress, and it will be difficult to provide a unified response. Business efficiency cannot be improved. In addition, when each person manages the e-mail by himself / herself, the e-mail must be appropriately classified and organized according to the content, the name of the sender, etc., which is troublesome.

【0006】

Not limited to e-mail, for example, technical information, product information, etc. are provided on a WWW server or as a digital recording medium, but knowledge information with a high degree of interest is provided by multiple people within the organization. In many cases, each is managed. Even if multiple people download and manage the same type of information, it is only a memorandum unique to each person and does not contribute to the improvement of knowledge and responsiveness of the entire organization. Also, doing duplicate work on the same type of information is a waste of resources.

【0007】

Simply storing information such as e-mails and web pages in one place in order to accumulate knowledge information does not solve the problem. It cannot be used as a shared property simply by accumulating it without considering the use of knowledge information. If the knowledge information is classified and organized in detail and frequently, the usability will be improved, but it will take a lot of time and effort to classify and organize the knowledge information. Further, in a rapidly progressing technical field, it may be difficult to properly classify and organize the latest knowledge information, and it may be difficult to use the information.

【0008】

On the other hand, a system in which information can be freely retrieved by a full-text search engine with little or no classification of knowledge information can be considered. In this case, it is possible to search using the latest keywords according to technological changes without the trouble of sorting and organizing. However, since the setting of search conditions (keywords, etc.) is left to each person, the skill of the searcher is directly reflected in the search time and search accuracy. Experienced searchers and searchers with specialized knowledge can retrieve the desired information in a relatively short time. Inexperienced searchers and non-specialized searchers cannot set appropriate search conditions, so it takes a long time to extract useless information. For example, even in the case of a search service provided on the Internet, if an appropriate keyword can be set, information can be narrowed down in the range of about ten to several tens. If the appropriate keywords cannot be set, even one piece of information will not be hit, or thousands to tens of thousands of pieces of information will be hit. More specifically, for example, when searching for "α plane" in the technical field of image processing, a huge amount of information is hit only by the keyword "α". On the other hand, in the case of the keyword "α plane", the number of hits is an appropriate value, but the search accuracy is not sufficiently high. This is because related terms such as "α channel" and "α synthesis" are not taken into consideration.

【0009】

In fields with remarkable progress such as communication technology and computer technology, the meaning and notation of a word may change to a new word before it is officially decided. Therefore, it is not possible to obtain the latest information with keywords that are already in the past. On the contrary, the latest keywords alone do not always provide sufficient correct information. It is necessary to enter appropriate search conditions in consideration of the usage of technical terms and changes in notation, but it is difficult for inexperienced searchers to set appropriate search conditions.

【0010】

The present invention has been made in view of the above-mentioned various problems, and an object of the present invention is to provide an information retrieval device capable of improving information retrieval efficiency and search accuracy. Another object of the present invention is to provide an information management device capable of centrally managing information.

【0011】

[Means for solving problems]

In the invention according to claim 1, a storage means for storing information, a search condition input means for inputting search conditions, and information stored in the storage means are used as search conditions input in the search condition input means. A search means for searching based on, a history saving means for storing the search history by the search means, and a search support means for outputting the search history in response to a request are provided, and search conditions are input using the search history. It is characterized by making it possible.

【0012】

Here, examples of the "information" include various digital data such as e-mails and web pages. Not limited to this, even voice and information on paper media can be handled by digitization. Examples of the "storage means" include a large-capacity storage device such as a hard disk device.

【0013】

When the user inputs a search condition via the search condition input means, the search means searches for information according to the entered search condition. The history of the search by the search means is saved by the history saving means. When an inexperienced searcher performs a search, the search support means outputs a search history if the searcher desires. Here, "output of search history" typically means a case where the search history is displayed on the screen. However, the present invention is not limited to this, and the search history can be output by voice using synthetic voice, or the search history can be printed out. By inputting search conditions using the past search history, search efficiency and search accuracy can be improved, and search time can be shortened.

【0014】

As in the invention of claim 2, the search history can include at least search conditions and searcher attribute information.

【0015】

Not only search conditions such as keywords used in the past, but also attribute information of the searcher is saved as a search history. Examples of the "searcher attribute information" include the searcher's name, affiliation, job title, occupation, specialized technical field, related technical field, years of enrollment, work history, hobbies, family structure, and the like. By referring to the attribute information, it is possible to judge the usefulness of the past search conditions and the like.

【0016】

As in the invention of claim 3, the extraction condition input means for inputting the condition for extracting the search history stored in the history storage means is further provided, and the search support means is extracted based on the extraction condition. You can also output the search history.

【0017】

For example, it is possible to display the search history as a history list without processing it, but as the amount of the search history increases, it takes time to refer to it, and the usability decreases. Therefore, using the attribute information of the searcher as a key, only a predetermined search history is extracted from the search history and output. As a result, useful search history can be used promptly.

【0018】

As in the invention of claim 4, the search support means can also output the search history related to the search condition input to the search condition input means among the search history stored in the history storage means.

【0019】

As a method of outputting the related search history, for example, a method of extracting and outputting the search history including the input keyword, or a method of extracting and outputting the search history including words and phrases (synonyms, etc.) related to the input keyword is extracted and output. Can be adopted. In the former case, for example, when the input keyword is "α plane", the past search history "α plane + α channel" including "α plane", "image composition + α plane", etc. are output. In the latter case, when the input keyword is "α plane", the search history "RGB + image composition", "α blending", etc. including the keyword related to the keyword are output. As a result, the searcher can quickly detect an appropriate keyword by referring to the past search history.

【0020】

As in the invention of claim 5, the history saving means can also save the search history satisfying a predetermined condition set in advance.

【0021】

As the "predetermined condition", for example, whether or not the occupation type, the specialized field, or the like has reached a predetermined level can be mentioned. That is, the search history of an inexperienced searcher has little utility value and may be a waste of storage resources. Therefore, the conditions for saving the search history are set in advance, and only the search history with high utility value is saved. The setting of whether or not to save the search history can be left to the searcher's self-report.

【0022】

When the information searched by the search means is set to be private as in the invention of claim 6, the management means for transmitting the disclosure request information to the manager of the information related to the non-disclosure is provided. It can also be prepared further.

【0023】

There are various levels of confidentiality of information, from the level at which public disclosure is permitted to the level at which only specific persons can access it. In addition, some e-mails and the like contain private contents. Therefore, when information is centrally managed, access control is provided for who is allowed to use it to what extent. Here, for a search by a searcher who does not have access authority, it is possible to control so that highly confidential information or personal information is not included in the search target from the beginning. However, as in the case of being suddenly selected as a member of a special project, they may have practical access authority even if they do not formally have access authority now. In addition, even if the information includes personal content, it may be permitted to be disclosed to a specific person or the whole at the discretion of the manager of the information. Therefore, even if it is non-public information, it can be searched by including it in the search target, and if the searcher wants to access the non-public information, the disclosure request information is sent to the manager of the non-public information by e-mail or the like. .. Here, it should be noted that since the disclosure request information is issued by the management means, the searcher does not know who is the administrator. If the administrator refuses to publish, the searcher cannot access the private information. As a result, flexible search can be performed while performing access control, and information retrieval omission can be prevented.

【0024】

The information retrieval device can also be realized by having a computer read a recording medium on which a predetermined program is recorded. The recording medium is not limited to various tangible recording media such as hard disks, floppy disks, CD-ROMs, DVD-ROMs, magneto-optical disks, and memories, and programs can be remotely downloaded via a network. Communication media can also be included. Furthermore, the present invention can also be realized as an information retrieval method. Further, the information retrieval device according to the present invention can be realized on a single computer, but the present invention is not limited to this. For example, the search condition input means is realized on the client computer, the storage means is realized on the data management server, the search means is realized on the application server, and so on. You can also do it. Further, for example, when the search condition input means is realized on the client computer, the input means can be implemented on the client computer from the beginning, or the program from the server can be transferred to the client computer in whole or in stages. It can also be realized by downloading.

【0025】

On the other hand, in the invention according to claim 9, in the information management device that manages the input information, a storage means for storing the input information, a management means for managing the information stored in the storage means, and a search condition. The search condition input means for inputting the information, the search means for searching the information stored in the storage means based on the search conditions input in the search condition input means, and the search means for outputting the search result, and the search. It is provided with a determination means for comparing the public level set in the information with the acquisition authority level set in the searcher and determining whether or not the searcher has the acquisition authority for the searched information. The management means is characterized in that, when it is determined that the searcher does not have the authority to acquire the searched information, the disclosure request information is transmitted to the manager of the information.

【0026】

By targeting all information as a search target and transmitting disclosure request information to the administrator of non-public information, it is possible to prevent omission of search while maintaining appropriate access control.

【0027】

In the invention according to claim 10, in the information management device that manages the input information, the storage means for storing the input information is based on the degree of association between the input information and a predetermined predetermined department. The information is provided with a management means for selecting a delivery destination and delivering the information to the selected delivery destination, and the management means analyzes the input information and extracts predetermined information from the input information. The relevance is calculated by comparing the information with the attribute information preset in the predetermined department, and the department having the predetermined relevance is selected as the delivery destination.

【0028】

When certain information is input, this information is analyzed, and the predetermined information extracted from the information is compared with the attribute information preset in the predetermined department. As a result of the comparison, the input information is distributed to the departments having a predetermined degree of relevance. A simple example will be given for explanation. First, the input information is morphologically analyzed to extract the wording. Next, if any of the extracted words that are frequently used in the sales department such as "estimate" or "delivery date" are found, this information can be distributed to the sales department. Information can be distributed to appropriate departments by managing words that are closely related to each department in association with the depth of relevance.

【0029】

On the other hand, in the invention according to claim 11, in the information management device that manages the input information, a storage means for storing the input information and a distribution condition storage for storing the distribution conditions preset for each distribution destination. Means, a delivery management means for delivering all or part of the input information to a predetermined delivery destination based on the stored delivery conditions, and a monitoring means for monitoring the response status of the delivered information. The response history saving means for storing the response history detected by the monitoring means and the delivery condition updating means for updating the delivery condition based on the saved response history are provided.

【0030】

As the delivery condition, for example, a keyword can be mentioned. When information including a keyword registered in advance is input, the information is delivered to the delivery destination associated with the keyword. However, there is no guarantee that users have set appropriate delivery conditions. In addition, the type of information requested by the user changes according to changes in tastes, hobbies, living environment, work environment, and the like. Therefore, under fixed distribution conditions, unnecessary information may be distributed and necessary information may not be delivered. Therefore, the history of responses to the distributed information is saved, and the distribution conditions are dynamically changed based on the response history. As a result, the user can optimize the delivery conditions only by accessing the desired information. On the contrary, from the viewpoint of the information distribution source, the information can be distributed only to the interested users, and the response rate can be increased.

【0031】

As in the invention of claim 12, the delivery condition updating means can generate a new delivery condition based on the response history.

【0032】

The "new delivery condition" means a condition different from the original delivery condition, and includes, for example, a condition related to the original delivery condition, a condition corresponding to a superordinate concept or a subordinate concept of the original delivery condition, and the like. Can be done. Specifically, for example, it is assumed that the keyword "car" is set as the original distribution condition. In this case, all information including the keyword "car" is delivered to the user. Here, as a result of referring to the response history, when only the response rate to the information about the four-wheel drive vehicle is high, "rectional beagle" can be set as a new keyword. Alternatively, "outdoor sports" can be set as a related new keyword.

【0033】

In the invention according to claim 13, the delivery condition updating means can also delete the delivery condition whose response rate is equal to or less than a predetermined value based on the response history.

【0034】

By deleting the delivery condition with a low response rate, it is possible to prevent unnecessary information from being delivered to the user.

【0035】

In the invention of claim 14, the delivery condition updating means can also update the delivery condition after waiting for the approval of the delivery destination.

【0036】

Even if the response rate is low, the user may wish to distribute the information, and if the distribution conditions are changed without prior notice, the usability may be reduced. Therefore, the delivery conditions are updated after obtaining the user's approval.

【0037】

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, embodiments of the present invention will be described in detail with reference to FIGS. 1 to 12.

【0038】

1. First embodiment FIG. 1 is a block diagram showing a functional configuration of an information retrieval device according to the first embodiment of the present invention.

【0039】

As will be described later, the information retrieval device 1 can be configured to include an input unit 2, a management unit 3, a storage unit 4, a browsing unit 5, a search unit 6, a search support unit 7, and a search history storage unit 8. ..

【0040】

The input unit 2 is for taking various kinds of information from the outside into the information retrieval device 1. Examples of the information to be captured include an e-mail, a document file, a digitally converted audio signal, an image file read by a scanner device, information read from a recording medium such as a CD-ROM, and the like. For example, the system can be simplified by capturing all information by e-mail such as MIME (Multipurpose Internet Mail Extensions) format. However, the present invention is not limited to this, and as described above, data from a fax machine or a scanner device can also be taken in by performing a predetermined conversion or the like.

【0041】

The management unit 3 is for managing the input / output of information and the like. The management unit 3 stores various information input via the input unit 2 in a storage unit 4 including a hard disk device or the like. In addition, the management unit 3 retrieves the information stored in the storage unit 4 in response to the request from the search unit 6. The storage unit 4 stores the captured information, management information such as the sender name and receiver name of the information, and attribute information of each user (searcher).

【0042】

The reading unit 5 realizes, for example, a web server function, and performs two-way communication with a browser on a client computer. The search unit 6 searches for the stored information in response to the search request input via the browsing unit 5, and provides the search results to the client computer via the browsing unit 5.

【0043】

The search support unit 7 provides search support at the request of the searcher. The search history storage unit 8 stores history information related to past searches. Examples of the saved search history include search conditions, search date and time, searcher name, affiliation, occupation, job title, and specialized field. The search support unit 7 supports a new search by providing the search history to the searcher.

【0044】

FIG. 2 is an explanatory diagram schematically showing an example of a search by a browser on the client computer side. As shown in Fig. 2 (a), various search conditions can be entered on the data search screen. For example, one or a plurality of keywords can be input to the keyword input unit G1. Keywords arranged in the horizontal direction are processed by the AND condition, and keywords arranged in the vertical direction are processed by the OR condition. The sender name and receiver name of information can be input to the sender name input section G2 and the receiver name input section G3, respectively. As a result, only the information belonging to a specific person can be searched. The search target period can be input to the period input unit G4. In addition, on the data search screen, a search start button B1 for instructing the start of a search, a cancel button B2 for instructing the cancellation of the search or deletion of the search conditions, and a search support button B3 for calling the search support screen are displayed. And can be provided.

【0045】

FIG. 2B is an explanatory diagram showing a list of search results. The list of search results can be configured to include, for example, a serial number, a public classification, a title of information, a date of information acquisition, a name of a sender of information, and the like. The disclosure category indicates the disclosure level of the information registered in the storage unit 4. Here, for convenience of explanation, it is divided into two types, public and private, but it is not limited to this, and there are multiple types such as "public", "confidential", "confidential", "confidential", and "private". It may be set by type. In addition, on the search result list screen, there is a display button B4 for instructing the screen display of the searched information, a search support button B5 for calling the search support screen, and a keyword for reading the keyword list, which is a type of search support. A list button B6 can be provided.

【0046】

FIG. 3 is an explanatory diagram schematically showing an example of the search support screen, and FIG. 3A is a screen for extracting the search history. It is difficult for an inexperienced searcher to retrieve necessary information even over a long period of time because appropriate search conditions cannot be set. Therefore, an inexperienced searcher requests search support from the search support unit 7.

【0047】

The simplest search support is to disclose all past search history (search log) as it is without passing through any filter. This is an effective method when the amount of search history is small, but as the amount of search history stored increases, it becomes more difficult to obtain the required search history. Therefore, in order to extract useful search history, various extraction conditions are set on the search history reference screen shown in FIG. 3 (a). In the searcher name input unit G5, by inputting the searcher name, only the search history by a specific searcher with abundant experience can be extracted. In the affiliation input unit G6 and the job type input unit G7, the search history by a searcher of a specific department or a specific job type can be extracted. This is effective when you do not know the name of an in-house expert in the field. In the keyword input unit G8, a desired keyword can be input. As a result, the history searched by the search conditions including the input keyword is extracted. In the period input unit G9, the extraction target of the search history can be divided in time.

【0048】

Figure 3 (b) is a screen showing a list of extracted search histories. The search history list screen can be composed of, for example, a serial number, a search date, a search time, a searcher's name, an affiliation, a job title, a job title, and the like. When you select a specific search in the search list, the search conditions (typically keywords) at that time are displayed. Therefore, an inexperienced searcher can efficiently learn the search know-how by referring to the keywords used by the experienced searcher, and can set more appropriate search conditions.

【0049】

FIG. 4A is a keyword list display screen. On the keyword list screen, keywords used in the past are displayed in association with the frequency of use. The keywords of all search histories may be displayed in a list, or only the keywords of the search history by a specific searcher, a specific department / occupation, etc. may be displayed in a list. By narrowing down the target to some extent by the searcher name and displaying the list, the search know-how can be intensively learned. In addition, it is possible to easily grasp what kind of information and how much interest a predetermined searcher or a predetermined department is interested in.

【0050】

FIG. 4B is an explanatory diagram showing another example of the search support screen. The search support screen may be provided with, for example, a keyword input unit G10 and a free input unit G11 for freely inputting irregular sentences. The text input to the free input unit G11 is analyzed by, for example, morphological analysis, to estimate the search target desired by the searcher. Then, the search history is searched based on the estimated search target, and related keywords are displayed. In addition, keywords related to the keywords input to the keyword input unit G10 are also displayed. The searcher can freely select the keyword input unit G10 and the free input unit G11. For example, when the searcher inputs the keyword "four-wheel drive vehicle", the search history including this "four-wheel drive vehicle" includes related keywords such as "recreational beagle" and "multipurpose passenger car". The search history is extracted.

【0051】

When the search is completed, the search history is saved as the case may be. In the simplest configuration, all search history can be saved. In a more advanced configuration, only search histories that meet certain conditions are saved. For example, it is possible to save only the search history estimated to be worthy of saving based on information such as occupation and specialized field. Alternatively, a "history save button" can be provided on the screen to save the search history by the searcher's self-report. Furthermore, searchers of a predetermined level or higher are registered in advance on the information retrieval device 1 side, all search histories by this high-level searcher are saved, and search histories of other persons are saved by his / her choice. You can also do it.

【0052】

When all the search history is saved, it is possible to prevent omission of saving of the search history, but since the amount of data increases, the extraction time of the search history may become long. On the other hand, if only the search history by a high-level person is saved, useful search history can be quickly extracted, but the search history may be missed. Which search history is saved and how much can be set depending on the case.

【0053】

According to the present embodiment configured as described above, the following effects are obtained.

【0054】

First, since the search history is displayed according to the searcher's wishes, an inexperienced searcher can refer to the search history and set new search conditions. Therefore, the search can be performed efficiently and the search time can be shortened. In addition, search know-how can be efficiently learned in the so-called OJT (On the Job Training) format.

【0055】

Secondly, since the search history includes not only the search conditions but also the attribute information of the searcher, it becomes easier to determine which is a useful search history based on the attribute information, and the usability is improved. To do.

【0056】

Thirdly, since only the extracted search history can be displayed, the searcher can extract and refer to only the necessary search history from the huge search history.

【0057】

Fourth, not only the search history directly related to the input keyword but also the search history including the related keyword can be displayed, so that a more appropriate keyword can be detected quickly.

【0058】

2. Second embodiment Next, a second embodiment of the present invention will be described with reference to FIGS. 5 and 6. In each of the embodiments described below, the same components as those described above are designated by the same reference numerals, and the description thereof will be omitted. The feature of this embodiment is that when the protected information is searched, the administrator of the protected information is requested to disclose the information from the system side.

【0059】

FIG. 5 is a block diagram of the information retrieval device 11. The apparatus 11 inspects whether or not non-public information is included in the search result list by the search unit 6. When the viewing of the non-public information is requested, the management unit 12 creates an e-mail requesting the disclosure of the information from the manager of the non-public information (for example, the recipient of the e-mail). This information disclosure request mail is sent to a predetermined administrator via the output unit 13. The information disclosure request email includes the name and attributes (occupation type, affiliation, etc.) of the person who wishes to view the information. The administrator can change the information disclosure category to "public" in response to the information disclosure request email. When the disclosure category is changed to "public", the viewer can use the information via the information retrieval device 11. Alternatively, the administrator can directly send the information to the person who wants to browse by e-mail or the like. The administrator does not need to respond to the information disclosure request. Whether or not to respond to information disclosure is left to the discretion of the administrator.

【0060】

FIG. 6 is a flowchart showing a main part of the processing flow by the information retrieval device 11. First, when the searcher inputs a search condition (S1), the search unit 6 searches according to the search condition (S2). The search result by the search unit 6 is displayed, for example, as shown on the right side in FIG. 6 (S3). Here, in the case of information in which the disclosure category is set to "private", the information necessary for identifying the information such as the sender name, the recipient name, and the title is not displayed.

【0061】

For example, if the searcher wants to display the private information by clicking on the private information (S4: YES), the access authority of the searcher and the disclosure level of the information are compared (S5). The access authority of the searcher can be set in advance when logging in to the information retrieval device 11. In that case, it is not necessary to check the existence of access authority with S5. However, when logging in, if you do not set the access authority or if you want to thoroughly protect the information, you can authenticate the searcher again with S5 and check the access authority.

【0062】

If the searcher has access privileges (S5: YES), the information hidden as "***" is displayed (S6). On the other hand, if the searcher does not have access authority (S5: NO), an information disclosure request email is sent from the management unit 12 to the administrator of the non-public information (S7). The administrator of non-public information can change the disclosure category according to the information disclosure request email, allow the requesting searcher to browse in particular, and search directly without going through the information retrieval device 11. Information can also be sent to a person.

【0063】

According to the present embodiment configured as described above, since it is possible to search including non-public information, it is possible to prevent omission of search. Further, in the case of non-public information, the information disclosure request mail is sent to the administrator so that the information cannot be seen by the searcher, so that the information can be used while appropriately protecting the information.

【0064】

3. Third embodiment Next, a third embodiment of the present invention will be described with reference to FIGS. 7 to 9. A feature of the present invention is that information is distributed based on preset distribution conditions.

【0065】

FIG. 7 is a block diagram of the information management device 21. The information management device 21 includes a distribution management unit 22, a distribution condition storage unit 23, and a summary generation unit 24. For example, distribution conditions such as keywords are registered in advance in the distribution condition storage unit 23. The distribution management unit 22 acquires the input information via the management unit 3 and determines whether or not this information conforms to the registered distribution conditions. In the case of information that meets the distribution conditions, the information is distributed to one or more users associated with the distribution conditions.

【0066】

The summary generation unit 24 summarizes information that meets the distribution conditions and reduces the amount of distribution data. As a method of summarizing, for example, simply extract a predetermined number of characters from a paragraph containing a keyword, extract a highlighted part, extract only a predetermined number of characters from the beginning, or only a text part excluding image data. Various methods such as extracting the above can be adopted. If the original amount of information is small, it is not necessary to generate a summary. By reducing the amount of data and delivering it, the time for sending and receiving e-mail can be shortened. Therefore, it is effective when it is required to shorten the connection time with the network, such as mobile communication. Specifically, when moving at high speed, it is difficult to maintain a stable session for a long time, so that communication may be interrupted while receiving an e-mail. In addition, since it is almost impossible to take an eye off while driving, it is necessary to receive an e-mail in a short time. Therefore, in such a case, the image data can be removed or the text can be extracted and distributed.

【0067】

FIG. 8 (a) shows an example of information distribution to registered users. When information matching the pre-registered distribution conditions is input to the information management device 21, this information is distributed to the user associated with the distribution conditions. In this case, for example, an e-mail containing the title of the information, a URL (Uniform Resource Locator) or an e-mail address, the body or summary of the information (hereinafter, also referred to as "e-mail") is sent to the user. When the summary is delivered or when the original information is to be confirmed directly, the user can browse with a browser based on the delivered URL or the like. The distributed information can be read and discarded. This is because the original information is stored in the storage unit 4.

【0068】

FIG. 8B is an explanatory diagram showing the flow of information distribution. First, each user registers desired distribution conditions in the device 21 (b1). When the information conforming to the delivery conditions is taken into the device 21, the device 21 delivers the information to each user by e-mail (b2). Each user can browse information from his / her own terminal or the like via a browser (b3).

【0069】

The information management device 21 can be applied to an automatic response system, an email distribution system, a questionnaire collection system, and the like. An example is shown in FIG.

【0070】

As shown in FIG. 9A, various e-mails such as an e-mail requesting a quotation, an e-mail requesting technical support, and a market complaint arrive at the apparatus 21. The device 21 analyzes the content of each email and delivers it to the appropriate department. For example, a quote request email is delivered to a sales representative, a technical support request email is delivered to a technical support representative, and a market complaint is delivered to a grievance handler.

【0071】

Figure 9 (b) shows an example of delivery processing by analyzing the contents of an email. First, the contents of the email are morphologically analyzed, and each word is extracted, such as "your company, product, price ... estimate" (S11). Next, the degree of relevance between each extracted word and the registered department is calculated (S12). Specifically, the degree of relevance between each representative word and each department is registered in a database in advance, and the degree of relevance is obtained by referring to this database. Then, for each department, the degree of relevance is aggregated and comprehensively analyzed (S13), and the mail is delivered to the department having a degree of relevance equal to or higher than a predetermined value (S14).

【0072】

Here, the delivery destination of the mail is not limited to one place. If there are multiple departments with a degree of relevance equal to or higher than a predetermined value, they are distributed to each department. As a result, it is possible to prevent a situation in which the mail is not delivered to the department to be delivered, and it is possible to prevent a delay in the response. In addition, the mail can be delivered even if the degree of relevance is not higher than the predetermined value. For example, market complaints are important feedback information from the market and can be distributed not only to grievance personnel but also to sales personnel and technical personnel.

【0073】

Although the automatic mail sorting system has been illustrated, those skilled in the art can easily apply it to an automatic response system and a questionnaire collection system. In the case of an automatic response system, the contents of digitized voice, e-mail, etc. are analyzed according to FIG. 9 (b) to determine which of the pre-registered response patterns corresponds to. Then, when the corresponding response pattern is detected, a pre-registered message is returned to the source of the information. In the case of a questionnaire collection system, the content of the questionnaire is often standardized, so it is possible to further simplify it. The contents of the questionnaire can be analyzed and classified into "prospect files", "important customer files", etc. and saved.

【0074】

According to the present embodiment configured in this way, information conforming to the distribution conditions can be distributed to the user associated with the distribution conditions, so that the user can obtain the desired information simply by registering a keyword or the like. You can get it. In addition, since the input information is analyzed to determine the distribution destination of the information, the information can be distributed to an appropriate department.

【0075】

4. Fourth embodiment Next, a fourth embodiment of the present invention will be described with reference to FIGS. 10 and 11. The feature of this embodiment is that the above-described embodiments are integrated and the distribution conditions can be changed according to the response history of the distributed information.

【0076】

FIG. 10 is a block diagram of the information management device 31 according to the present embodiment. The apparatus 31 includes the search support system described in the first embodiment, the disclosure request system for non-public information described in the second embodiment, and the information distribution system described in the third embodiment. It has all of. In addition to this, the present device 31 further includes a response monitoring unit 32, a response history storage unit 33, and a distribution condition update unit 34, respectively, as will be described later.

【0077】

The response monitoring unit 32 monitors the information browsing request from the user. That is, the response monitoring unit 32 monitors the response status such as which user accessed what information at what time. The response status detected by the response monitoring unit 32 is stored in the response history storage unit 33 as a response history. The delivery condition update unit 34 determines whether or not the already set delivery conditions are appropriate based on the saved response history, deletes unnecessary delivery conditions, and sets new delivery conditions. Is.

【0078】

FIG. 11 is a flow chart showing an example of the process of updating the delivery condition. First, when an information browsing request from a user is detected (S21), a response rate is calculated for each registered delivery condition (for example, a keyword) (S22). For example, if a user who has delivered a total of 157 pieces of information including the keyword "car" accesses the information including the keyword "car" 110 times in total, the response rate is about 70%.

【0079】

Next, the response history is used to predict what the user is interested in (S23). Specifically, for example, information that the user has accessed in the past is morphologically analyzed and then a full-text search is performed to detect words that are common to information having a high response rate. For example, even if the response rate to the keyword "economy" is low, the response rate to information containing a specific company name or word may be high if the information group including "economy" is analyzed. Therefore, in S23, a common word is detected, and a new keyword related to this word is detected (S24). For this new keyword, a common word can be used as it is, or a superordinate concept, a subordinate concept, a related word, or the like of the common word can be used.

【0080】

Next, among the pre-registered keywords, a keyword having a response rate of a predetermined value or more is detected (S25), and a new keyword related to the keyword having a high response rate is detected (S26). The case where the response rate to information including the keyword "Western music" is high will be described as an example. A user who is interested in "Western music" can add "dance" as a new keyword when it is known by analysis of the response history of other users that he / she is also highly interested in "dance".

【0081】

Further, among the keywords registered in advance, a keyword having a response rate of a predetermined value or less is detected (S27). Then, the new keywords detected in S24 and S26 are added to the existing keyword group, and the user is notified that the keywords with a low response rate detected in S27 are deleted (S28). If the user approves this notification, it will be updated to a new keyword group (delivery condition) (S29). If the user's approval is not obtained, the registered keyword group will not be changed.

【0082】

According to the present embodiment configured in this way, all the effects of each of the above-described embodiments can be obtained. In addition to this, in the present embodiment, since the distribution conditions are changed according to the response status from the user, appropriate information can be distributed to the user according to the change in the situation, and the distribution conditions are manually set each time. There is no need to change it with, and usability is improved. In addition, since keywords with a low response rate are deleted, unnecessary information can be prevented from being delivered, communication traffic can be reduced, and storage capacity can be prevented from being wasted. Furthermore, since the delivery conditions are changed after waiting for the user's approval, usability is improved. Even when the new distribution conditions match the user's taste, it may not be necessary for others to know that the information is being distributed. Alternatively, even if the delivery condition has a low response rate, the collection of information may be preceded and the access may be delayed. Therefore, usability is improved by changing the distribution conditions after obtaining the approval of the user.

【0083】

The above-mentioned dynamic change system of distribution conditions can be easily applied to, for example, a so-called push-type advertising service. That is, at first, a relatively wide range of product information can be distributed, and the next product information to be distributed to each user can be individually adjusted according to the response rate of each user.

【0084】

5. Specific example FIG. 12 is a block diagram showing a specific example of the present invention. The information management system 100 is connected to a network 102 such as the Internet via, for example, a router 101.

【0085】

The mail server 103 receives an e-mail from the internal network or the external network. Further, the conversion unit 104 converts the information input in a form other than the e-mail into an e-mail and inputs the information to the mail server 103. In this way, various digital information captured in the form of e-mail is managed by the information management unit 105.

【0086】

The information management unit 105 stores the body and attached files of the e-mail in the body spool 106. On the other hand, the header information of the e-mail is analyzed into the sender name, the recipient name, the title, the amount of data, and the like, and stored in the management database 107. The data of the mail body stored in the body spool 106 is analyzed by the morphological analysis engine 107. The result of the morphological analysis is saved in the index file 109 for full-text search. Therefore, the search engine 110 can detect the data of the target mail body by performing a search on the full-text search index file. An image file, a document file, or the like may be attached to the e-mail, but it is optional whether or not the attached file is also subjected to morphological analysis to create a full-text search index.

【0087】

The mail that arrives at the mail server 103 is also delivered to each user described in the destination of the header. When the external user is the recipient of the mail, the mail is transferred from the mail server 103 to the mail server to which the destination user belongs via the router 101 or the like. When forwarding to an internal user, it is saved in the spool 112 via the mail delivery unit 111. Each user receives mail addressed to him / her by accessing the spool 112 individually.

【0088】

The WWW server 113 provides a so-called web server function. The information providing unit 114 performs two-way communication with the operation unit 122, which will be described later, to provide a user's search and database operation. The information providing unit 114 is always running as a UNIX daemon, for example. The database management unit 115 is for operating the database 107. In addition to the above-mentioned email header information, the database 107 stores search history, response history, keywords registered for delivering information, related keywords, and the like.

【0089】

On the other hand, the user's client computer 120 includes a browser 121 for viewing information provided by the WWW server 113, an operation unit 122 sent from the information management system 100 side and activated, a mail program (not shown), and the like. It is provided. The user can input search conditions, distribution conditions, and the like via the operation unit 122, and can request display of search results. From the viewpoint of environmental management and maintenance on the client side, it is preferable to collectively or stepwise transmit the program for realizing the operation unit 122 from the system 100 to the client computer 120. However, the present invention is not limited to this, and a program for realizing the operation unit 122 can be implemented in advance on each client computer 122.

【0090】

According to the system 100 configured in this way, the above-mentioned search support service, information distribution service, and the like can be realized.

【0091】

The present invention is not limited to the above-described embodiments. A person skilled in the art can make various modifications by adding, deleting, changing, or the like a component to each embodiment. For example, each function of the system can be distributed to a plurality of computers.

【0092】

Further, in the first and second embodiments, although it is described as an "information retrieval device", it can also be grasped as an "information management device".

【0093】

[Effect of the invention]

As described above, according to the information retrieval device according to the present invention, a new search can be performed using the search history, so that the search efficiency can be improved. Further, according to the information management device according to the present invention, the input information can be distributed to an appropriate department. Furthermore, the delivery conditions can be dynamically changed to optimize the delivery conditions.

[Simple explanation of drawings]

[Figure 1]

It is a block diagram which shows the functional structure of the information retrieval apparatus which concerns on 1st Embodiment of this invention.

[Figure 2]

Fig. 2 (a) shows the search screen, and Fig. 2 (b) shows the search result list screen.

[Fig. 3]

Fig. 3 (a) shows the search support screen, and Fig. 3 (b) shows the search history list screen.

[Fig. 4]

Fig. 4 (a) shows the keyword list screen, and Fig. 4 (b) shows another example of the search support screen.

[Fig. 5]

It is a block diagram which shows the functional structure of the information retrieval apparatus which concerns on 2nd Embodiment of this invention.

[Fig. 6]

It is a flow chart which shows the disclosure request processing for non-public information.

[Fig. 7]

It is a block diagram which shows the functional structure of the information management apparatus which concerns on 3rd Embodiment of this invention.

[Fig. 8]

FIG. 8 (a) is an explanatory diagram showing an example of information distribution to users, and FIG. 8 (b) is an explanatory diagram showing a flow of a distribution service.

[Fig. 9]

FIG. 9 (a) is an explanatory diagram of a state in which the contents of the mail are analyzed and distributed to each department in charge, and FIG. 9 (b) is a flow diagram of the process of determining the delivery destination of the mail.

[Fig. 10]

It is a block diagram which shows the functional structure of the information management apparatus which concerns on 4th Embodiment of this invention.

[Fig. 11]

It is a flow chart of the process which changes a delivery condition according to the response state from a user.

[Fig. 12]

It is a block diagram which shows the specific example of this invention.

[Explanation of symbols]

1 Information retrieval device 2 Input section 3 Management Department 4 Preservation section 5 Reading section 6 Search section 7 Search Support Department 8 Search history storage section 11 Information retrieval device 12 Management Department 13 Output section 21 Information management device 22 Distribution Management Department 23 Delivery condition storage unit 24 Summary generator 31 Information management device 32 Response monitoring unit 33 Response history storage unit 34 Delivery condition update department