JP2002312363A - Information distribution method and information distribution device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve a problem of a conventional information distribution method wherein a distribution request acquisition method previously giving a user's request by use of a keyword or a sentence takes a lot of trouble over a non- consideration call or request setting. SOLUTION: This method has a retrieval step for executing document retrieval of the periphery including distribution desire contents to a document set; a retrieval result preservation step for preserving information about a retrieval result; a distribution information setting step for registering information to be distributed; a comparison step for comparing a retrieval result document preserved in the retrieval result preservation step and information set in the distribution information setting step; and an information notifying step notifying the distribution information selected by the comparison means.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of distributing distribution information consisting of a large number of electronic documents including texts, for example, when receiving news articles or the like via the Internet or the like, document information matching the purpose of the user. And a method of distributing information.

[0002]

2. Description of the Related Art In a distribution service or the like on the Internet, there is a strong demand to receive only a document for a specific purpose of a user from an electronic document including a text created every day. For example, when priority is given to only really necessary information from a large number of news articles and the like by email or the like, the user wants to receive only news that meets his purpose.

In response to such a request, there is a service that uses a keyword given in advance as an information request and distributes only text containing the keyword. However, in this case, since text information is selected only by a specific keyword, it is not always possible to retrieve a document that matches the purpose of the user. In order to solve this problem, a method of setting a sentence instead of a keyword as a user's information request is disclosed in Prior Art 1 (Japanese Patent Laid-Open No. 2000-259655).
Is disclosed. This prior art 1 aims at generating a database reflecting a user's information request.

[0006] FIG. 6 shows a configuration for creating a database according to prior art 1. In the prior art 1, a request information collecting unit 601 collects a user's information request by designating a document, and determines a content information to be added to a database 604. An adding unit for adding content information (document information) to the database 604 by the adding unit 603.

Next, an information distribution method according to the prior art 1 will be described with reference to FIG. The distribution request setting unit 13 sets a sentence or a keyword as the distribution request 14. In addition, user information 12 such as a mail address of each user is set by the user information setting unit 11 in advance.

On the other hand, distribution information 5 such as news articles is set as needed by a distribution information setting unit 6, and a comparison unit 7 compares the distribution request 14 with the distribution information. At this time, distribution information having a value equal to or greater than a certain threshold is extracted, and the information
With this, distribution is performed according to the user information.

[0007]

As described above, in the information distribution method according to the prior art, a user is requested to give an information request by using a keyword or a sentence in advance. There is a problem that it takes time to set the information request because the acquisition is not considered.

[0008] The present invention has been made to solve the above-described problems.
A search unit for a large number of document sets is provided, and the search results are stored in a search result storage unit, and can be given as a user's information request, thereby facilitating the setting of the user's information request. The present invention provides an information distribution method and an information distribution device capable of performing the above.

[0009]

An information distribution method according to the present invention includes a retrieval step of retrieving a document containing contents desired to be distributed to a set of documents and a document surrounding the retrieval, and a retrieval result storing step of retaining information of the retrieval result. A distribution information setting step of registering information to be distributed, a comparison step of comparing the search result document stored in the search result storage step with the information set in the distribution information setting step, and the distribution information selected by the comparing means. Is provided.

Further, the information distribution method according to the present invention includes a user information setting step for specifying a distribution target, and the information notification step notifies the distribution information to the target person specified by the user information setting step. .

Further, in the information distribution method according to the present invention, the distribution information setting step and the comparison step include extracting a keyword from the document, registering and comparing the distribution information, and extracting the keyword from the document. , A dictionary that describes the classification of words is used.

Further, the information distribution method according to the present invention includes a classification step of preliminarily setting a classification of a document with respect to a set of documents to be searched in the search step.

[0013] Also, an information distribution apparatus according to the present invention includes a search means for searching for a document including contents desired to be delivered to a set of documents, and surrounding documents, a search result storage means for selecting and storing information of search results, Distribution information setting means for registering information to be distributed, comparison means for comparing the search result document stored by the search result storage means with the information set by the distribution information setting means, and distribution information selected by the comparison means. An information notification unit for notifying is provided.

[0014] The information distribution apparatus according to the present invention further comprises user information setting means for specifying an object to be distributed.
The information notification means is configured to notify the target person specified by the user information setting means of the distribution information.

Further, in the information distribution apparatus according to the present invention, the distribution information setting means and the comparison means extract a keyword from the document, register and compare the distribution information, and extract the keyword from the document. , A dictionary that describes the classification of words is used.

Further, the information distribution apparatus according to the present invention includes a classification unit for setting a classification of a document in advance for a set of documents to be searched by the search unit.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1. FIG. 1 shows an example of the configuration of an information distribution system according to Embodiment 1 of the present invention. The information distribution system shown in FIG. 1 includes a search unit 2 for searching for a document from a large set of documents 1, a search result storage unit 3 for storing a search result by the search unit 2, and distribution information 5 for news articles and the like. A delivery information setting unit 6, a search result document 4 for expressing a user information request stored by the search result storage unit 3, and a delivery information setting unit 6.
A comparing unit 7 that compares the index of the distribution information set in
An information notification unit 8 that distributes the information selected by the comparison unit 7 to the user.

FIG. 1 further includes a user information setting unit 11 for setting user information 12 such as an e-mail address of a user of a distribution destination. In the configuration shown in FIG. 1, the document set 1 and the distribution information 5 searched by the search unit 2 are information of different databases. However, even if the document set 1 and the distribution information 5 are information of the same database, good.

The operation of the first embodiment of the present invention will be described. First, the search unit 2 searches a set 1 of a large number of documents including texts, for example, a set of arbitrary documents on the Internet or a set of information on a specific field in an intranet. Performs a search. At this time, what is given as the search request may be a keyword or a sentence.

Next, from among the search results searched by the search unit 2, the user selects a document close to the content that he / she wants to be distributed, and the search result storage unit 3 converts the document into a search result document. Save as 4. At this time, search result document 4
Contains text data described in a natural language.

The distribution information setting section 6 analyzes text data in the distribution information 5 composed of documents such as news collected or created regularly or irregularly, extracts keywords, and extracts the keywords of each document 5 in the distribution information. Generate and maintain a vector as an index.

Next, the comparison unit 7 compares the distribution information set by the distribution information setting unit 6 with the search result document 4 stored by the search result storage unit 3 to calculate the similarity. Then, information having a value equal to or greater than a certain threshold is extracted as notification information (information to be actually distributed).

Next, the information notifying unit 8 distributes the distribution information selected by the comparing unit 7 to the user based on the user information 12 set by the user information setting unit 11 by means such as mail.

FIG. 2 is a flowchart of the distribution information setting unit 6 according to the first embodiment of the present invention. Hereinafter, the operation of the distribution information setting unit 6 will be described according to the flowchart of FIG. First,
In step S201 in FIG. 2, the text in the document of the distribution information 5 is analyzed.

At this time, in the analysis of the document, the text in the document is divided into words using morphological analysis. For each word, the part of speech included in the morphological analysis dictionary 205 used for analysis,
It shall include information such as word classification. Here, the morphological analysis is a widely known technique, and thus the details are omitted.

Next, in step S202, a keyword is extracted from the result of the morphological analysis of the document. In extracting the keyword, the keyword may be limited to an independent word or a part of speech specified arbitrarily may be removed. Alternatively, it is also possible to use the dictionary 205 that describes the classification of words, and to specify whether or not to remove the words by the classification of the words.

Next, in step S203, a vector including a keyword is generated. Here, the weight according to the appearance frequency of the word appearing in the document of the distribution information 5 in the entire document is stored as a vector for each document as an index. The processing of step S203 is to perform index generation processing of a search method using a conventionally known vector space model, and the tf · idf value ( The value obtained by multiplying the tf value, which is the frequency of each word in each document, by the idf value, which is the reciprocal of the frequency of the word in all documents. May be used), and a χ square value or the like is used.

Next, in step S204, it is assumed that the vector generated in step S203 is held as an index of each document in the distribution information 5.

FIG. 3 shows an example of a document search in the search unit 3, and “mobile phone”,
Specify the character string "Kana-Kanji conversion" and search button 30
An example in which a search result 303 is returned when a search is executed by designating 2 is shown. At this time, as shown in FIG. 3, it is assumed that a registration button 304 as to whether or not to save the search result document as a distribution request is presented for each search result.

The user refers to the contents of the document from the list of the search results 303, or, based on information such as the title and the outline displayed in the list, is a document including the contents that the user wants to distribute. If it is determined whether or not to store the distribution request, the registration button 304 is pressed, and the corresponding document is stored as the search result document 4.

FIG. 4 is a flowchart showing the operation of the comparing section 7. First, in step S401, the search result document 4 stored according to the instruction from the user described with reference to FIG. 3 is analyzed to generate a vector. This procedure is the same as the operation from step S201 to step S203 in the distribution information setting unit 6 shown in FIG.

Next, in step S402, a similarity calculation is performed by taking an inner product of an index of distribution information and a vector, and it is determined in step S403 whether the similarity is equal to or greater than a certain threshold. Information that is equal to or larger than the threshold is selected as the notification information, and is passed to the information notification unit 8. If the similarity is equal to or smaller than a certain threshold in step S403, the similarity with another distribution information is calculated.

Since the first embodiment of the present invention is configured as described above, there is an effect that the setting of the information request of the user can be facilitated.

Second Embodiment FIG. 5 shows a configuration example of an information distribution system according to a second embodiment of the present invention. The information distribution system shown in FIG. 5 includes, in addition to the retrieval unit 2, the retrieval result storage unit 3, the comparison unit 7, the distribution information setting unit 6, etc., which are the main components of the present invention described in the first embodiment. And a classification unit 9 that can provide classification items to documents to be searched in advance.

Next, the operation of the second embodiment of the present invention will be described. The classification unit 9 performs document classification processing based on a conventionally known vector space model when document data is given. The frequency of occurrence of words that appear in the document for each classification item in the entire document Is stored as an index with a vector for each document. The statistic of the word as a vector element is tf
-It may be an idf value, a chi-square value, or the like.

First, the classifying section 9 classifies documents to be searched in advance, and holds, for each classification item, a vector expressing the content of the classification item. For example, if the classification items include “mobile phone”, “digital camera”, “color printer”, etc., some sampling documents suitable for each classification item are given to create a vector for each classification item. The remaining search target documents other than the sample may be assigned to each classification item based on the similarity between the vector for each classification item and the vector created from each search target document, or may be manually assigned to each classification item. good.

Next, when the user selects a document that matches the distribution request from the search results in the search unit 2, the search result storage unit 3 corresponds to the classification of the document selected by the user. The search result vector 10 is stored as a document distribution request.

Next, the comparison unit 7 compares the vector 10 with the distribution information set by the distribution information setting unit 6.

The processes after the comparison unit 7 are the same as those in the first embodiment.

According to the present invention, as described in the first embodiment, a search portal site or the like includes a search unit 2 for a large set of documents, and the search result 4 is stored in the search result storage unit 3. In addition, since it is possible to easily set the information request of the user by enabling the information request to be given as the information request of the user, it is possible to perform the information distribution reflecting the information request of the user. . Also, in the second embodiment, the documents to be searched by the search unit 2 are classified in advance by the classifying unit 9 and the classification information is used as a distribution request, thereby distributing a document that more closely matches the information request of the user. can do.

Further, in the configuration of the present invention, instead of the distribution information, advertisement information on a portal site is used,
By using the information notification unit 8 for displaying search results, it is also possible to use it as a method for displaying advertisement information on search results.

[0042]

As described above, according to the present invention, it is possible to provide a search unit for a large amount of document sets, store the search results in a search result storage unit, and provide the search results as a user information request. By doing so, it is possible to easily set the information request of the user, and it is possible to perform information distribution reflecting the information request of the user.

Further, by preliminarily classifying documents to be searched in the search unit by the classifying unit and using the classification information as a distribution request, it is possible to distribute a document that more closely matches the information request of the user.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an information distribution system according to a first embodiment of the present invention.

FIG. 2 is a flowchart of a distribution information setting unit according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram of an example of a document search in a search unit.

FIG. 4 is a flowchart showing the operation of a comparison unit.

FIG. 5 is a configuration diagram of an information distribution system according to a second embodiment of the present invention.

FIG. 6 is a configuration diagram of a conventional database creation system.

FIG. 7 is a configuration diagram of a conventional information distribution system.

[Explanation of symbols]

1: Document set, 2: Search unit, 3: Search result storage unit, 4: Search result document, 5: Distribution information, 6: Distribution information setting unit, 7: Comparison unit, 8: Information notification unit.

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Katsushi Suzuki 2-3-2 Marunouchi, Chiyoda-ku, Tokyo Mitsubishi Electric Corporation F-term (reference) 5B075 NK32 PP22 PQ74 PR06 QM08

Claims (8)

[Claims] A search step of searching a set of documents for a peripheral document containing desired content to be distributed; a search result storing step of storing search result information; a distribution information setting step of registering information to be distributed; An information delivery step, comprising: a comparison step of comparing the search result document stored in the storage step with information set in the delivery information setting step; and an information notification step of notifying delivery information selected by the comparison unit. Method. 2. The method according to claim 1, further comprising a user information setting step of specifying an object to be distributed, wherein the information notifying step notifies the distribution information to a target person specified by the user information setting step. Information distribution method described. 3. A distribution information setting step and a comparison step, wherein a keyword is extracted from a document, a distribution information is registered and compared, and a keyword classification is described when extracting a keyword from the document. 3. The information distribution method according to claim 1, wherein the information distribution method comprises: 4. The information distribution method according to claim 1, wherein said search step includes a classification step of previously setting a document classification for a set of documents to be searched. 5. A search means for searching for a document set containing contents desired to be distributed and surrounding documents, a search result storing means for selecting and storing information of a search result, and a distribution for registering information to be distributed. An information setting unit, a comparison unit that compares the search result document stored by the search result storage unit with the information set by the distribution information setting unit, and an information notification unit that notifies the distribution information selected by the comparison unit. An information distribution device characterized by the above-mentioned. 6. A system according to claim 1, further comprising user information setting means for specifying a distribution target, wherein said information notifying means is configured to notify the distribution information to a target person specified by said user information setting means. The information distribution device according to claim 1, wherein: 7. A distribution information setting means and a comparison means for extracting a keyword from a document, registering and comparing distribution information, and, when extracting a keyword from the document, a dictionary describing a classification of words. The information distribution apparatus according to claim 1, wherein the information distribution apparatus is configured to use the information distribution apparatus. 8. The information distribution apparatus according to claim 5, further comprising a classification unit configured to set a classification of a document in advance for a set of documents to be searched by the search unit.