JP5142643B2 - Tag assignment support system and program - Google Patents

Description

  The present invention relates to a tag addition support system and a program for adding a tag to make a message with a tag in order to facilitate content search of an electronic file.

  In recent years, with the spread of the Internet, a large amount of information has been distributed, and the need for a technique for accurately extracting and using necessary information from the information has increased.

Conventionally, place names, person names, organization names, product names, etc. are automatically extracted from the body of a document using a so-called morphological analysis method, and tags are automatically assigned, making machine processing such as automatic search easy using this as a clue. A document processing apparatus has been proposed (see, for example, Patent Document 1).
JP 2007-94855 A

  The document processing apparatus described above is suitable for fully automatic machine processing, but immediately after creating a document, a message is sent and broadcast to a plurality of recipients, or sent in the past, or When searching for a required message in a received document, the tag that was automatically assigned is inappropriate and filtering of the message is inappropriate and the required message is overlooked, or automatically Since there are too many tags attached to, there are disadvantages that many unnecessary messages are mixed and the search operation is complicated.

  We hate the disadvantages of automatic tag assignment as described above, and it is conceivable that the user assigns a tag suitable for a message created by manually performing the tag assignment operation in consideration of the tag name. This imposes an excessive load on the user and is not realistic.

  In view of the above problems, the present invention automatically analyzes a document and selects tag candidates, and examines and designates a tag to be given from the tag candidates each time, thereby overloading the user. It is not compulsory, and because the tag to be automatically assigned is inappropriate, the message filtering is inappropriate and the necessary message is overlooked, or it is unnecessary because there are too many tags to be automatically added. It is an object of the present invention to provide a tagging support system and a program with high accuracy that do not cause inconvenience that a large number of messages are mixed and search work is complicated.

  The tag assignment support system of the present invention is extracted by the keyword extracting means according to the relation table, the relation table storing the keyword candidates associated with the keyword in association with the keyword, the keyword extracting means for extracting the keyword from the message, and the relation table. Tag conversion means for converting the keyword into tag candidates, a frequency table for storing the tag candidates in association with the frequency of the tag candidates previously assigned as tags, and the tag candidates converted by the tag conversion means , Tag candidate display means for displaying N (N is a natural number) tag candidates with high frequency in the frequency table, and a tag selected by the user from the N tag candidates displayed by the tag candidate display means in the message Tag assigning means to be assigned to the tag and the frequency in the frequency table of the tags assigned by the tag assigning means. Characterized in that it comprises a count-up means for Ntoappu.

  In addition, the frequency table includes a plurality of frequency tables, and one frequency table selected by the user is used, so that the frequency table can be properly used, for example, for business, hobby, or organization.

  Further, the tag candidate display means displays the corresponding conversion source keyword together with the tag candidate to be displayed, so that the user can know the conversion source keyword when selecting the tag candidate. can do.

  In addition, the tag assigning means, regardless of the tag candidate displayed by the tag candidate display means, a user manually assigns a tag, and registers the assigned tag in the frequency table and the relation table, or By updating, the user can decide the tag uniquely, so that a more accurate tag can be given.

  In addition, the tag assigning means can assign the tag attribute together with the tag to be attached depending on the tag, and can specify according to the tag attribute together with the tag. can do.

  According to the present invention, a computer is extracted by the keyword extracting means according to the relation table, a relation table storing a keyword and a tag candidate corresponding to the keyword in association with the keyword, a keyword extracting means for extracting the keyword from the message, and the relation table. Tag conversion means for converting the keyword into tag candidates, a frequency table for storing the tag candidates in association with the frequency of the tag candidates previously assigned as tags, and the tag candidates converted by the tag conversion means , Tag candidate display means for displaying N (N is a natural number) tag candidates with high frequency in the frequency table, and a tag selected by the user from the N tag candidates displayed by the tag candidate display means in the message Tag assigning means to be assigned to the tag and the frequency in the frequency table of the tags assigned by the tag assigning means. Is a program for functioning as a tagging support system and a count-up means for Ntoappu.

  According to the present invention, a tag to be assigned to a message is examined and selected from tag candidates each time, and a designation operation is performed, so that an excessive load is not imposed on the user and a tag to be automatically assigned is not required. Tag addition with high accuracy without the disadvantage of overlooking necessary messages for appropriateness or having too many tags to be automatically added, so that many unnecessary messages are mixed and search work is complicated A support system can be provided.

  The best mode for carrying out the present invention will be described below in detail with reference to the accompanying drawings.

  FIG. 1 is a diagram showing a configuration of a tag addition support system according to an embodiment of the present invention. In FIG. 1, the transmitting device 1 includes a communication function 11 for communicating with the server 2 via a communication line such as an Internet line or a mobile phone line, and a display screen 12. A message M created by a user (not shown) is displayed on the display screen 12 and sent to the server 2 by the communication function 11. The tag designation function 13 is a function for designating a tag to be added to the message M, and details of the operation will be described later.

  The server 2 receives the message M by the communication function 21 communicating with the communication function 11 and inputs it to the tagging processing function 22. The tagging processing function 22 refers to the message M and data stored in a tag database 23 prepared in advance (details will be described later), and analyzes the message M using, for example, a morphological analysis method to determine a tag to be assigned. . In the above tag determination process, n (n is a natural number) tags t1 to tn are determined (hereinafter, the determined set of n tags is referred to as “tag T”).

  The tagged message generation function 24 generates the tagged message A by combining the determined tag T to be added and the message M. The tagged message A is stored in the tagged message database 26 and guided to the screen functions 27a and 27b.

  The screen functions 27a and 27b pass the tagged message A if the tagged message A includes the designated tags a and b, respectively. If the tagged functions A and b are not included, the screen functions 27a and 27b This is a function to prevent passage.

  The receiving devices 3a and 3b include communication functions 31a and 31b having a function of communicating with the communication functions 28a and 28b of the server 2, message memories 32a and 32b for storing the tagged message A, and screen specifying functions 33a and 33b, respectively. doing.

  The screen designation functions 33a and 33b cause the screen functions 27a and 27b corresponding to the screen functions 27a and 27b, respectively, to determine whether to pass the tagged message A through the communication function 31a and the communication function 28a and the communication function 31b and the communication function 28b, respectively. Designate tags a and b.

  Therefore, only when the tagged message A includes the tag a, the tagged message A passes through the screen function 27a and is stored in the message memory 32a via the communication function 28a and the communication function 31a. Similarly, only when the tagged message A includes the tag b, the tagged message A passes through the screen function 27b and is stored in the message memory 32b via the communication function 28b and the communication function 31b.

  As described above, even though the transmitting device 1 broadcasts the tagged message A including the same message M to both the receiving devices 3a and 3b, the tag a and the reception indicated by the receiving device 3a are received. Depending on the contents of the tag b instructed by the device 3b, this may be received by both the receiving devices 3a and 3b, may be received by only one, or neither may be received. In other words, the receiving devices 3a and 3b can receive only the message M suitable for each by instructing the tag a and tag b, respectively, and can prevent the reception of the inappropriate message M.

  Note that the operations of the screen designating functions 33a and 33b described above perform the same function as described above when the message M stored in the tagged message database 26 is called, so that each of the receiving devices 3a and 3b has a tag a. By indicating the tag b, it is possible to receive only the appropriate message M from among the messages M stored in the tagged message database 26 and prevent the reception of the inappropriate message M.

  FIG. 2 is a diagram illustrating a configuration of the tagging processing function 22. In the figure, a message M is input from a terminal 2200 connected to the communication function 21 and stored in the text memory 221. The relation table and frequency table data corresponding to the transmitting device 1 are read from the data stored in the tag database 23 from the terminal 2203 connected to the tag database 23, the relation table is stored in the relation table memory 223, and the frequency table memory. The frequency table is stored in H.226.

  FIG. 3 is a diagram illustrating an example of a relationship table. The relationship table is a table in which tag examples for each keyword are registered in advance as data. In the figure, the top four tag examples are shown as tag examples 1 to 4 for each of the five keywords. Tag examples of about 0 to 10 are registered.

  FIG. 4 is a diagram illustrating an example of a frequency table. The frequency table is a table in which tag names assigned in the past for each user are registered as data in the order of frequency assigned. In the figure, the top five tag names are shown from frequency 1 to frequency in order of frequency. Normally, tag names of about 100 to 10,000 are registered.

  The keyword extracting means 222 sequentially inputs the storage contents of the body memory 221, refers to the storage contents of the relational table memory 223, and extracts keywords from the contents of the message M using a document analysis technique such as a morphological analysis technique, for example. Input to the tag conversion means 224.

  The tag conversion means 224 converts a plurality of keywords normally inputted with reference to the stored contents of the relation table memory 223 into a tag example. For example, wildflowers, spring flowers, yellow flowers, plants, and the like are registered as tag examples 1 to 4 in the keyword “cattle” in FIG. 3, and the tag conversion means 224 uses these tag examples 1 to tag examples. 4 and the like are output to the tag candidate selection means 225 together with the conversion source keyword.

  The tag candidate selection means 225 selects m as the natural number and tag examples 1 through m as the tag candidates from among the tag examples 1 through 4 inputted as described above. Usually, m is about 10-30. Hereinafter, for convenience of explanation, an example will be described in which one keyword is used and three tag names are displayed.

  The tag candidate selection unit 225 outputs the tag example 1 to the tag example 3 (wildflower, spring flower, yellow flower) selected above to the tag candidate display unit 227 and the tag determination unit 228 together with the keyword (cattle).

  The tag candidate display means 227 outputs the keyword and the selected three tag names from the terminal 2201, and displays the keyword name on the display screen 12 via the communication function 21 and the communication function 11 of the transmitting device 1. A certain birch and the three selected tag names, wildflowers, spring flowers, and yellow flowers, are displayed as tag candidates.

  The user of the transmission device 1 looks at the display content on the display screen 12 and knows that the tag candidates are wildflowers, spring flowers, and yellow flowers, and that each keyword is a kabatami, and operates the tag designation function 13. For example, tag examples 1 (wildflower) and tag example 3 (yellow flowers) are designated.

  The operation result of the tag specifying function 13 is guided to the tag determining means 228 via the communication function 11, the communication function 21 of the server 2, and the terminal 2202 of the tagging processing function 22, and the tag determining means 228 is the tag example 1 ( Wild grass) and tag example 3 (yellow flower) are determined as the tags t1 and t2 to be attached to the message M.

  Usually, since a plurality of keywords are extracted, the keywords and tag candidates are displayed on the display screen 12 for each keyword, and the tags T1 to tm are sequentially determined. It may be determined, or a plurality of keywords extracted from the message M and tag candidates corresponding to each keyword are simultaneously selected and displayed, and given to the message M by the user performing a designation operation of the tag designation function 13 at once. The tag T to be used may be determined by a single operation.

  When tag T is determined by a single operation by simultaneously selecting and displaying a plurality of keywords and corresponding tag candidates, for example, a list of tag candidates is displayed on the vertical axis, and tag candidates are displayed on the horizontal axis. Display the corresponding keyword.

  As described above, when the tag candidate selecting unit 225, the tag candidate display unit 227, and the tag determining unit 228 are operated, the user views the display screen 12 and operates the tag specifying function 13 to determine a tag T suitable for the message M. In addition, there are too many tag candidates to be displayed, so that the tag determination operation becomes complicated, and conversely, there are too few tag candidate tables to be displayed, and there is no inconvenience such as an accurate tag candidate not being displayed. It can be performed.

  The output of the tag determining means 228 is input to the tagged message generation function 24 via the terminal 2204. The output of the tag determining unit 228 is guided to the count-up unit 229, and the frequency table stored in the frequency table memory 226 is counted up. In the example of the frequency table shown in FIG. 4, tag example 1 (wildflower) is determined as the tag T. Therefore, the numerical value 800 in the frequency column of the first frequency (wildflower) in the frequency table memory 226 is a numerical value 801 obtained by adding 1. Updated. The update result of the frequency table is guided to the tag database 23 via the terminal 2203, and the accumulated data in the tag database 23 is updated.

  As explained in detail above, according to the tag assignment support system of the present invention, a tag is not automatically assigned, but a tag to be assigned is examined and selected from tag candidates each time and designated operation is performed. Thus, without imposing an excessive load on the user, the tag to be automatically given is inappropriate, and the filtering of the message M is inappropriate and the necessary message M is overlooked, or the tag to be automatically given Therefore, it is possible to provide a tagging support system with a high degree of accuracy without the disadvantage that a lot of unnecessary messages M are mixed and the search operation is complicated.

  In addition, this invention is not limited to the said Example.

  In the above-described embodiment, the tag data storage form has a structure in which a tag is simply added to a message. However, a tag attribute such as a conversion source keyword when a tag is assigned is associated with the tag, and the tag and tag You may make it the structure which gives the combination of an attribute to a message. At that time, it is desirable to have a structure in which there are a tag that gives only a tag and a tag that gives a combination of a tag and a tag attribute.

  For example, if “Group”, which is an organizational unit, is assigned as a tag, and “X”, “Y”, or “Z” is assigned to a message in combination with the tag as a keyword, it is related to all groups. If the user wants to receive the message, the tag “group” may be selected. If the user wants to receive the message related to the group X, the tag “group” is selected and the tag attribute for the tag is selected. By selecting the keyword “X”, a message particularly related to the group X can be specified and received.

  As another example, a message "creation date" is given as a tag, and a specific date, for example, the extracted keyword "August 31, 2007" is combined with the tag in a format that can be recognized as a date. If a message created on or after August 1, 2007 is received, the tag “creation date” is selected, and the tag attribute comparison operation “creation date ≧ 2007/8” is selected for the tag. By performing “Month 1st”, a desired message can be received.

  In addition, by setting a plurality of genres according to the message content such as plants, sports, work, etc., and using a frequency table suitable for each genre, the accuracy of tags to be added can be further improved. it can. In this case, for example, a plurality of frequency tables corresponding to plants, sports, and work are provided, and genre names J1 to J3 are assigned to the frequency tables, and the tagging processing function 22 is displayed on the display screen 12 of the transmission device 1 by the user. A question screen as to whether to use the genre name frequency table is displayed, and a user who views this question screen operates the tag designation function 13 to instruct a genre name to be used from among the genre names J1 to J3. The tagging processing function 22 may read the frequency table according to this instruction from the tag database 23 and store it in the frequency table memory 226, and then perform the functional operation described with reference to FIGS.

  Further, instead of the user specifying the genre by the operation of the tag specifying function 13 above, a word or sentence indicating the genre is inserted into the message M, and the tagging processing function 22 analyzes the message into the corresponding genre. A corresponding frequency table may be read from the tag database 23.

  In the description of the tagging function 22 using FIGS. 2 to 4, the user of the transmission device 1 knows the tag candidate by looking at the display content of the display screen 12, and operates the tag designation function 13 to identify the tag candidate. If a suitable tag name is specified from among them, but a suitable tag name cannot be found among the tag candidates displayed on the display screen 12, or if it is desired to designate a tag name not included in the tag candidates, the tag is manually You can enter the name directly. The directly input tag name is given to the message M by the tag determining unit 228, and the tag determining unit 228 registers the tag name in the frequency table and the relationship table via the count-up unit 229 (when there is no tag) ) Or update (if the tag is present but not so frequently displayed).

  The tag assignment support system of the present invention is also realized by a program for causing a computer to function as the tag assignment support system. This program may be stored in a computer-readable recording medium.

  The recording medium on which this program is recorded may be the ROM itself of the server 2 shown in FIG. 1, or a program reading device such as a CD-ROM drive is provided as an external storage device. It may be a CD-ROM or the like that can be read by insertion.

  The recording medium may be a magnetic tape, a cassette tape, a flexible disk, a hard disk, an MO / MD / DVD, or a semiconductor memory.

It is a figure which shows the structure of the tag provision assistance system by one Example of this invention. It is a figure which shows the structure of the tagging process function of this invention. It is a figure which shows the example of the relationship table of this invention. It is a figure which shows the example of the frequency table of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Transmission apparatus 2 Server 3 Reception apparatus 11 Communication function 12 Display screen 13 Tag designation function 21 Communication function 22 Tag processing function 23 Tag database 24 Message generation function with tag 26 Message database with tag 27 Screen function 28 Communication function 31 Communication function 32 Message memory 33 Screen designation function 221 Text memory 222 Keyword extraction means 223 Relation table memory 224 Tag conversion means 225 Tag candidate selection means 226 Frequency table memory 227 Tag candidate display means 228 Tag determination means 229 Count up means

Claims (6)

A relationship table for storing keyword candidates associated with the keyword in association with the keyword;
Keyword extraction means for extracting keywords from the message;
Tag conversion means for converting the keywords extracted by the keyword extraction means according to the relationship table into tag candidates;
A frequency table for storing the tag candidates in association with the frequency of the tag candidates previously assigned as tags;
Tag candidate display means for displaying N (N is a natural number) tag candidates having a high frequency in the frequency table among the tag candidates converted by the tag conversion means;
Tag attaching means for attaching a tag selected by the user from the N tag candidates displayed by the tag candidate display means to the message;
Counting up means for counting up the frequency in the frequency table of the tag attached by the tag attaching means ,
The frequency table is composed of a plurality of frequency tables, and one frequency table selected by the user is used . 2. The tagging support system according to claim 1, wherein the plurality of frequency tables are frequency tables suitable for a plurality of genres set by a user according to message contents . The tag addition support system according to claim 1 or 2, wherein the tag candidate display means displays a corresponding conversion source keyword together with a tag candidate to be displayed. Regardless of the tag candidates displayed by the tag candidate display means, the tag assigning means manually adds a tag separately, and registers or updates the assigned tag in the frequency table and the relation table. The tag assignment support system according to claim 1, wherein The tag assignment support system according to any one of claims 1 to 4, wherein the tag assigning unit can assign an attribute of the tag in association with the tag to be attached depending on the tag. Computer
A relation table for storing keyword candidates associated with the keyword in association with the keyword, keyword extraction means for extracting a keyword from a message, and a tag for converting the keyword extracted by the keyword extraction means according to the relation table into a tag candidate A frequency table that associates and stores the frequency of the tag candidate previously assigned as a tag to the tag candidate and a tag candidate converted by the tag conversion unit; Tag candidate display means for displaying tag candidates (N is a natural number), tag attaching means for attaching a tag selected by the user from the N tag candidates displayed by the tag candidate display means to the message, Count-up hand that counts up the frequency in the frequency table of the tag given by the tag granting means It equipped with a door,
The frequency table is composed of a plurality of frequency tables, and a program for functioning as a tagging support system in which one frequency table selected by a user is used .

Images