JP6056829B2 - Recommendation creation device - Google Patents

Description

  The present invention relates to a recommendation creation device, and more particularly to a recommendation creation device that creates an appropriate recommendation in comparison with a person or an organization.

  In recent years, tools for extracting a person and a keyword from document data and using information obtained by linking them are widely used. For example, a savvy person extraction device disclosed in Patent Document 1 (Japanese Patent Application Laid-Open No. 2002-56001) extracts a technical term from a document and stores a degree of familiarity in a field to which the technical term belongs in advance for each person. An appropriate person is extracted from.

  However, Patent Document 1 discloses a method for identifying a savvy person by combining technical terms and a person, but does not show a method for utilizing it for human resource development.

  An object of the present invention is to provide a recommendation creating apparatus that can “visualize” knowledge for each person in charge and support human resource development from the viewpoint of knowledge.

The recommendation creation device according to the first aspect of the present invention includes a keyword extraction unit, a creator extraction unit, a person keyword extraction unit, a co-occurrence information extraction unit, and a recommendation creation unit. The keyword extraction unit extracts keywords from a large number of document data created and accumulated by each person. The creator extraction unit extracts a creator from the document data. The person keyword extraction unit creates and extracts association information between the creator and the keyword. The co-occurrence information extraction unit assigns a co-occurrence keyword to a keyword associated with the creator, and creates and extracts co-occurrence information for each creator. The recommendation creation unit creates recommendation content for the creator based on “a keyword having a high appearance rate in a document of a comparison subject or a group of comparison subjects but a low appearance rate in a self document” for each creator. In addition, the recommendation creation unit sets a keyword that has a high appearance rate in its own document to a keyword that is low in its own document even though the degree of co-occurrence in the document of the comparison subject or the comparison target group is large. Display the recommended content based on the priority.

  This recommendation creation device makes it possible to analyze the gaps that exist between the comparison target and yourself (for example, the difference from the top, the difference from the average, or the difference from the ideal form), so that the shortage is revealed. And can be encouraged to make up for the shortage.

  Moreover, in this recommendation creation apparatus, the effort for preferentially supplementing what is not the original knowledge but is not required is promoted.

  The recommendation creation device according to the second aspect of the present invention is the recommendation creation device according to the first aspect, and the recommendation creation unit further selects a keyword having a low appearance rate as a comparison target document and a high appearance rate as a self document. Create recommended content based on it.

  This recommendation creation device enables analysis of one's strengths within an organization or team, thus encouraging efforts to increase strengths.

  The recommendation creating apparatus according to the third aspect of the present invention is a recommendation creating apparatus according to the first or second aspect, and presents an inquiry destination for supplementing information lacking in the recommendation content. Contains content.

  In this recommendation creation device, by presenting a contact (person or organization), even if the organization to which the comparison is made belongs, it can be widely used both inside and outside the organization to compensate for the lack of information. Is easy to use.

The recommendation creating apparatus according to the fourth aspect of the present invention is the recommendation creating apparatus according to any one of the first to third aspects , wherein the keyword extracting unit extracts the number of keywords extracted from one document data. Determine the upper limit of.

  In this recommendation creating apparatus, the quantity of keywords can be made independent of the size of document data.

The recommendation creating apparatus according to the fifth aspect of the present invention is the recommendation creating apparatus according to any one of the first to third aspects , wherein the keyword extraction unit extracts the quantity of keywords extracted from one document data. Is determined according to the size of the document data.

  In this recommendation creating apparatus, the number of keywords proportional to the size of document data can be secured.

  In the recommendation creating apparatus according to the first aspect of the present invention, it is possible to analyze a gap (for example, a difference from the top, a difference from the average, a difference from an ideal form) existing between the comparison target and oneself. Therefore, the shortage can be made obvious and efforts to make up for the shortage are encouraged.

  In addition, efforts are made to preferentially compensate for what is supposed to be, although it should be.

  In the recommendation creating apparatus according to the second aspect of the present invention, it becomes possible to analyze one's strengths within an organization or a team, so that efforts to increase the strengths are encouraged.

  In the recommendation creating apparatus according to the third aspect of the present invention, by presenting a contact (person or organization), even if the comparison target is the organization to which the user belongs, it belongs to make up for the lack of information. As information can be requested widely regardless of inside or outside the organization, it is easy to use.

In the recommendation creating apparatus according to the fourth aspect of the present invention, the quantity of keywords can be made independent of the size of document data.

In the recommendation creating apparatus according to the fifth aspect of the present invention, the number of keywords proportional to the size of the document data can be ensured.

The block diagram which shows the structure of the recommendation production apparatus which concerns on one Embodiment of this invention. The flowchart showing the 1st operation | movement of a recommendation production apparatus. “Person keyword table” for engineer A with creator ID “101” created by the person keyword extraction unit. “Person keyword table” for engineer B with creator ID “201” created by the person keyword extraction unit. “Person keyword table” for engineer C with creator ID “301” created by the person keyword extraction unit. A “co-occurrence information table” relating to the keyword “digital signal” in the document data of engineer A created by the co-occurrence information extraction unit. A “co-occurrence information table” relating to the keyword “digital signal” in the document data of the engineer B created by the co-occurrence information extraction unit. A “co-occurrence information table” regarding the keyword “digital signal” in the document data of the team to which the engineers A, B, and C belong, created by the co-occurrence information extraction unit. The flowchart showing the 2nd operation | movement of a recommendation production apparatus. The example of the recommendation content with respect to the engineer B in the comparison with the engineer A and the engineer B. Example of recommendation contents for engineer B in comparison between team and engineer B.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The following embodiments are specific examples of the present invention and do not limit the technical scope of the present invention.

(1) Configuration of Recommendation Creation Device 10 FIG. 1 is a block diagram showing the configuration of the recommendation creation device 10 according to an embodiment of the present invention. In FIG. 1, a recommendation creation apparatus 10 includes a document data storage unit 20, a keyword extraction unit 30, a creator extraction unit 40, a person keyword extraction unit 50, a co-occurrence information extraction unit 60, a recommendation creation unit 70, and an overall control unit. 100.

(2) Detailed description (2-1) Document data storage unit 20
The document data accumulation unit 20 has a mechanism for accumulating documents (including e-mails) created by each person as needed. At least a document name and a creator name or creator ID are attached to the document.

(2-2) Keyword extraction unit 30
The keyword extraction unit 30 extracts characteristic words / phrases in the document, that is, keywords. The number of words to be extracted can be specified for each document, preventing all from becoming keywords. Of course, the number of keywords can be determined according to the size of the document data.

(2-3) Creator extraction unit 40
The creator extraction unit extracts the creator of the document by using document property information, login information at the time of document registration, or from field in the case of mail.

(2-4) Person keyword extraction unit 50
The person keyword extraction unit 50 creates a person keyword table in which a person who is a document creator is associated with a keyword appearing in a document created by the person for each person, and stores the person keyword table.

  For example, as shown in FIG. 1, a table listing creator ID, document ID, document name, creation date / time, and keywords in order from the left is created. The creator ID is an ID for identifying the person who created the document. The document ID is an ID for specifying what kind of document the document created by the person is. As an example, if the person is a technical report created in 2001, an ID of 2001TR is given.

  Since the document name is usually determined by the creator of the document in accordance with its contents, there is a high possibility that the most important keyword is included. The creation date / time is the creation date / time of the document, but may be either the creation date / time or the update date / time.

  Further, in the person keyword table of FIG. 1, following the document “2001 TR”, information related to other documents also continues downward. It is assumed that such a person keyword table is created and stored for each creator.

(2-5) Co-occurrence information extraction unit 60
The co-occurrence information extraction unit 60 creates a “co-occurrence information table” for each creator based on the “person keyword table” created by the person keyword extraction unit 50.

  For example, as shown in FIG. 1, the co-occurrence information table displays a rate at which a keyword assigned to a row and a keyword assigned to a column simultaneously occur at the intersection of the row and the column. Thus, the “degree of association (degree of co-occurrence)” between keywords can be grasped by the number “the rate of simultaneous occurrence (simultaneous occurrence rate)”.

  As a result, it is possible not only to search for one or more persons who know the keyword, but also to know what information (co-occurrence information) each person has. Since a person who has more co-occurrence information about the keyword tends to have a higher level of proficiency with the keyword, an appropriate person can be searched for.

(2-6) Recommendation Creation Unit 70
The recommendation creation unit 70 creates recommendation content for the creator based on “a keyword having a high appearance rate in the document of the comparison subject or group of comparison subjects but a low appearance rate in the self document” for each creator.

  Compared with your co-occurrence information and team or organization co-occurrence information, or through comparison of your co-occurrence information with the co-occurrence information of the higher-ranking ability of the team to which you belong The keywords that are expected but rarely generated are extracted, and the recommendation content based on the keywords is created.

(2-7) General control unit 100
The overall control unit 100 is a so-called CPU, and executes necessary processing according to the OS and other programs.

  When the user inputs a keyword from the dedicated terminal, the overall control unit 100 sends appropriate commands to the keyword extraction unit 30, the creator extraction unit 40, the person keyword extraction unit 50, the co-occurrence information extraction unit 60, and the recommendation creation unit 70. Extract necessary information.

(2-8) Others The keyword extraction unit 30, the creator extraction unit 40, the person keyword extraction unit 50, the co-occurrence information extraction unit 60, and the recommendation creation unit 70 individually store storage areas for storing the created data. Or it shall have in a common memory | storage device.

(3) Operation of Recommendation Creation Device 10 Here, two operations will be described. The first operation is an operation when a document is input from each terminal to the recommendation creating apparatus, and the second operation is an operation when a recommendation request is input from the terminal to the recommendation creating apparatus.

(3-1) First Operation of Recommendation Creation Device FIG. 2 is a flowchart showing a first operation of the recommendation creation device. In FIG. 2, the overall control unit 100 determines whether or not there is a document input from any of the terminals connected in step S1, and proceeds to step S2 when there is a document input, and continues when there is no document input. Continue judgment.

  Next, the overall control unit 100 stores the document input in step S2 in the document data storage unit 20, and proceeds to step S3.

  Next, the overall control unit 100 extracts keywords from the document data stored in the document data storage unit 20 via the keyword extraction unit 30 in step S3, and proceeds to step S4.

  Next, the overall control unit 100 extracts the creator from the document data stored in the document data storage unit 20 via the creator extraction unit 40 in step S4, and proceeds to step S5.

  Next, the overall control unit 100 sends a “person keyword table” (see FIGS. 3A to 3C) from the keyword extracted in step S3 and the creator extracted in step S4 via the person keyword extraction unit 50 in step S5. Create and remember.

  Next, based on the “person keyword table” created in the previous step S5 via the co-occurrence information extraction unit 60 in step S6, the overall control unit 100 creates a “co-occurrence information table” for each creator (FIG. 4A). And FIG. 4B).

(3-1-1) Detailed Explanation of Person Keyword Table For convenience of explanation, there are three digital signal processing engineers A, B, and C belonging to a certain development team. The explanation will proceed under the assumption that the engineer is in the 3rd year.

  FIG. 3A is a person keyword table for the engineer A with the creator ID “101” created by the person keyword extraction unit 50. In FIG. 3A, it is found that there are 100 documents including “digital signal” from the person keyword table created for all 150 documents created by Mr. A. In the keyword column, keywords with a high simultaneous occurrence rate are displayed from the left.

  Here, “sine wave”, “discretization”, “sampling”, “frequency domain”, and “filter” are simultaneously generated under the keyword “digital signal”.

  FIG. 3B is a person keyword table for the engineer B with the creator ID “201” created by the person keyword extraction unit 50. In FIG. 3B, it is found that there are 50 documents including “digital signal” from the person keyword table created for all 100 documents created by Mr. B.

  In Mr. B's document, co-occurrence keywords are simultaneously generated in the order of “sine wave”, “quantization”, “frequency domain”, and “filter” under the keyword “digital signal”. Since the keyword “quantization” is a keyword that has not been extracted from Mr. A's document, the keyword 7 is added and displayed next to the keyword 2 “sine wave”.

  FIG. 3C is a person keyword table for the engineer C with the creator ID “301” created by the person keyword extraction unit 50. In FIG. 3C, it is found that there are 40 documents including “digital signal” from the person keyword table created for all 100 documents created by Mr. C.

  In Mr. C's document, co-occurrence keywords are simultaneously generated in the order of “sine wave”, “sampling”, “frequency domain”, and “filter” under the keyword “digital signal”. Since all the keywords extracted from Mr. C's document were included in the keyword range extracted from Mr. A's document, they are displayed with the same symbols as the keywords extracted from Mr. A's document.

(3-1-2) Detailed Description of Co-occurrence Information Table The co-occurrence information extraction unit 60 is based on the “person keyword table” described in FIGS. 3A to 3C created by the person keyword extraction unit 50. A person having expertise in the field of “digital signal” is selected, and a “co-occurrence information table” is created based on the “person keyword table” shown in FIG. 3A.

  As shown in FIG. 3A, the keywords “digital signal”, “sine wave”, “discretization”, “sampling”, “frequency domain”, and “filter” are repeatedly extracted from Mr. A's document data. The co-occurrence information extraction unit 60 creates a matrix-like co-occurrence information table from these keywords.

  This co-occurrence information table displays the rate at which keywords assigned to rows and keywords assigned to columns simultaneously occur at the intersections of rows and columns.

  FIG. 4A is a co-occurrence information table related to the keyword “digital signal” in the document data of the engineer A created by the co-occurrence information extraction unit 60. In FIG. 4A, “digital signal”, “sine wave”, “discretization”, “sampling”, “frequency domain”, and “filter” are listed in order from top to bottom in the left end column of the co-occurrence information table.

  Also, “digital signal”, “sine wave”, “discretization”, “sampling”, “frequency domain”, and “filter” are listed in order from the left to the right in the upper row of the co-occurrence information table.

  At the intersection of the row and the column, the ratio of the keyword assigned to the row and the keyword assigned to the column is displayed in%.

(3-1-2-1) First aspect represented by co-occurrence information table This co-occurrence information table represents two aspects. The first aspect is the degree of co-occurrence as viewed from the keyword (co-occurrence ratio), and the second aspect is the degree of co-occurrence as viewed from the co-occurrence keyword. Hereinafter, it demonstrates concretely, referring drawings.

  First, “the co-occurrence ratio of co-occurrence keywords viewed from keywords” as the first aspect will be described. In FIG. 4A, the area in which the background is shaded displays “co-occurrence ratio of co-occurrence keywords viewed from keywords”.

  A numerical value of 80 is displayed at the intersection of the “digital signal” row and the “sine wave” column. This means that the keyword “digital signal” exists in 100 documents, and the keyword “sine wave” appears together with “digital signal” in 80 documents among them. In this case, “sine wave” is a co-occurrence keyword with respect to the keyword “digital signal”, and both appear in 80 documents out of 100, so the coincidence ratio is 80%.

  Also, when the keyword “discretization” appears in 65 documents out of 100 documents in which the keyword “digital signal” appears, “discretization” is shared with the keyword “digital signal”. Since both keywords appeared in 65 out of 100 documents, the simultaneous occurrence rate is 65%.

  If the keyword “sampling” appears in 60 documents out of 100 documents in which the keyword “digital signal” appears, “sampling” is a co-occurrence keyword for the keyword “digital signal”. Since both of them appeared in 60 documents out of 100, the simultaneous occurrence rate is 60%.

  If the keyword “frequency domain” appears in 35 documents out of 100 documents in which the keyword “digital signal” appears, the “frequency domain” is shared with the keyword “digital signal”. Since both keywords appeared in 35 documents out of 100, the simultaneous occurrence rate is 35%.

  Further, when the keyword “filter” appears in 30 documents out of 100 documents in which the keyword “digital signal” appears, “filter” is a co-occurrence keyword for the keyword “digital signal”. Since both of them appeared in 30 documents out of 100, the simultaneous occurrence rate is 30%.

(3-1-2-2) Second Aspect Represented by Co-occurrence Information Table The following is a second aspect of the “simultaneous occurrence ratio of other keywords viewed from co-occurrence keywords”. In FIG. 4A, a region with a colorless background displays “the simultaneous occurrence ratio of other keywords viewed from co-occurrence keywords”.

  First, a numerical value of 61 is displayed at the intersection of the “sine wave” row and the “digital signal” column. This is because the keyword “sine wave” exists in 130 of 150 documents of Mr. A, and the keyword “digital signal” appears along with “sine wave” in 80 documents among them. It means that. In this case, “digital signal” is a co-occurrence keyword with respect to the keyword “sine wave”, and both appear in 80 documents out of 130, so the simultaneous occurrence ratio is 61%.

  In addition, when the keyword “digital signal” appears in 65 documents out of 118 documents in which the keyword “discretization” appears, “digital signal” is shared with the keyword “discretization”. Since both keywords appeared in 65 out of 118 documents, the simultaneous occurrence rate is 55%.

  Further, when the keyword “digital signal” appears in 50 documents out of 111 documents in which the keyword “sampling” appears, “digital signal” is a co-occurrence keyword with respect to the keyword “sampling”. Since both of them appeared in 50 documents out of 111 documents, the simultaneous occurrence rate is 45%.

  If the keyword “digital signal” appears in 35 of the 117 documents in which the keyword “frequency domain” appears, the “digital signal” is shared with the keyword “frequency domain”. Since both keywords appeared in 35 documents out of 117, the simultaneous occurrence rate is 30%.

  If the keyword “digital signal” appears in 30 documents out of 120 documents in which the keyword “filter” appears, “digital signal” is a co-occurrence keyword with respect to the keyword “filter”. Since both of them appeared in 30 out of 120 documents, the simultaneous occurrence rate is 25%.

  In the same manner, a co-occurrence information table related to the keyword “digital signal” in Mr. B's document data is created and stored. For reference, FIG. 4B shows a co-occurrence information table for the keyword “digital signal” in the document data of engineer B created by the co-occurrence information extraction unit 60, and FIG. A co-occurrence information table relating to the keyword “digital signal” in the document data of the team to which B and C belong is described.

(3-2) Second Operation of Recommendation Creation Device Here, in order for the engineer B to analyze the gap existing between the host engineer A and himself in the field to which the term “digital signal” belongs The description will be made under the assumption that a recommendation request is input from the terminal of the recommendation creating apparatus 10.

  FIG. 5 is a flowchart showing the second operation of the recommendation creating apparatus 10. In FIG. 5, the overall control unit 100 determines whether or not there is an input requesting a recommendation in step S <b> 11, and proceeds to step S <b> 12 when there is a recommendation request, and continues the determination when there is no recommendation request.

  Next, the overall control unit 100 specifies the “person” requested for recommendation in step S12 and the “special field” to be recommended. Here, it is assumed that when a recommendation request is made, an input of “subject ID” and a “keyword” representing the specialized field are obtained. Therefore, the engineer B inputs his / her ID “201” and the keyword “digital signal” to identify “person” and “special field”.

  Next, the overall control unit 100 scans all the person keyword tables stored in the person keyword extracting unit 50 in step S13. By this scanning, engineer A and engineer C are detected in addition to engineer B. Then, the overall control unit 100 compares the engineer B (see FIG. 3B) who has created 50 documents including “digital signals” with the engineer who has created 100 documents including “digital signals”. A (refer FIG. 3A) is specified as a comparison subject.

  Next, the overall control unit 100 compares the co-occurrence information tables of the engineers A and B in step S14 (see FIGS. 4A and 4B).

  Next, the overall control unit 100 compares the co-occurrence information tables of the engineers A and B in step S14 (see FIGS. 4A and 4B).

  Next, the overall control unit 100 creates a recommendation content for the engineer B via the recommendation creation unit 70 in step S15.

(3-2-1) Recommendation Contents Based on Comparison with Higher Technician FIG. 6 is an example of recommended contents for engineer B in a comparison between engineer A and engineer B. In FIG. 6, the recommendation content is displayed on the monitor of the terminal. It is displayed in the comparison target person column (A column) so that the engineer A (superior) and the engineer B (self) can be compared. The number of documents of Engineer A and Engineer B is displayed in the lower B column. The number of documents of engineer A is 100/150, which means that 100 documents including the keyword “digital signal” out of all 150 documents of engineer A are created.

  In the C column below, co-occurrence keywords are displayed in descending order of the simultaneous occurrence ratio. Furthermore, the insufficient technology is displayed in the D column below. The deficient technology is a keyword “discretization” that is in the co-occurrence information table of engineer A but not in the co-occurrence information table of engineer B. From this point of view, the engineer B becomes aware that knowledge about “discretization” is insufficient.

  On the other hand, the strength of the engineer B is displayed in the E column below. The strength is in the co-occurrence information table of the engineer B but not in the co-occurrence information table of the engineer A, and the keyword “quantization” is displayed. From this point of view, engineer B will realize that knowledge about “quantization” is a strength.

  In this way, the shortage can be made obvious, efforts to make up for the shortage are promoted, and it becomes possible to analyze one's strengths within the organization and team, so that efforts to increase the strength are also encouraged.

(3-2-2) Recommendation Content Based on Comparison with Team Also, the co-occurrence information table of engineer B in FIG. 4B and the co-occurrence information table of the team to which engineers A, B, and C belong in FIG. In comparison, the co-occurrence keywords having a large degree of co-occurrence with “digital signal” in the team document are “discretization” and “sampling”, whereas the appearance rate of the keyword “digital signal” is in the document of the engineer B. Despite being high, it can be seen that “discretization” and “sampling” do not appear at all in the co-occurrence keywords co-occurring in the “digital signal”. In such a case, a supplementary comment is added together with the recommendation content similar to that shown in FIG.

  FIG. 7 is an example of recommended content for the engineer B in the comparison between the team and the engineer B. In FIG. 7, apart from the contents of columns A to E similar to FIG. 6, a comment that encourages acquisition of knowledge of “discretization” and “sampling” that are essential knowledge as a team is added as special notes in column F Is done.

  This is because the content of recommendations based on keywords with a low appearance rate in the self document even though the degree of co-occurrence for the keyword with a high appearance rate in the self document is large in all the document data. The purpose is to encourage efforts to preferentially compensate for what is supposed to be, though it should be.

  As described above, the recommendation creation device 10 aggregates keywords associated with a person and co-occurrence information based on the keywords, or totals keywords associated with the person of the team and the co-occurrence information based on the keywords, By comparing the co-occurrence information with the keywords of other persons or teams and the co-occurrence information, it is possible to extract a keyword to be recommended to the specific person and present the recommended content.

(4) Features (4-1)
In the recommendation creating apparatus 10, the recommendation creating unit 70 recommends content for creators based on “a keyword having a high appearance rate in a document of a comparison subject or a group of comparison subjects but a low appearance rate in its own document” for each creator. Create As a result, it becomes possible to analyze the gap that exists between the comparison target and oneself, to reveal the shortage, and to promote an effort to make up for the shortage.

(4-2)
In the recommendation creating apparatus 10, the recommendation creating unit 70 creates a recommendation content based on “a keyword having a low appearance rate in a comparison target document and a high appearance rate in a self document”. As a result, you can analyze your strengths within an organization or team, which encourages efforts to increase your strengths.

(4-3)
In the recommendation creating apparatus 10, the recommendation content includes a content for presenting an inquiry destination (person or organization) for supplementing information that is lacking in the recommendation content. As a result, even if the organization to which the subject belongs is compared, information can be widely obtained regardless of the inside or outside of the affiliated organization in order to compensate for the lack of information.

(4-4)
In the recommendation creating apparatus 10, the recommendation creating unit 70 sets a keyword having a low appearance rate in its own document even though the degree of co-occurrence with respect to a keyword having a high appearance rate in its own document is large in all document data. Display the recommended content based on the priority. As a result, efforts are made to preferentially supplement what is not the knowledge that should be.

  In the present embodiment, the “appearance rate” of the keyword is described. However, the present invention is not limited to this. For example, a threshold value of the number of appearances (absolute number) is set as a recommendation target keyword. Only keywords that exceed can be compared.

(5) Modification (5-1)
In the above embodiment, the co-occurrence information table based on the document data for 7 years of the engineer A who has joined the company for 7 years is compared with the co-occurrence information table based on the document data for 5 years of the engineer B who has joined the company for 5 years. However, for example, by comparing the co-occurrence information table based on the document data for five years from the entry of the engineer A and the co-occurrence information table based on the document data for the engineer B for five years, Comparison with the growth history of the upper engineer A can be performed.

  Thereby, it is possible to make a recommendation based on the comparison result with the “should be state” in which the influence of the year difference is excluded.

(5-2)
In addition, the comparison target is not limited to the engineer A, and a co-occurrence information table based on document data for five years after joining the company is averaged for all teams and organizations, and recommendations are made based on the comparison result. You can also

(5-3)
In the above embodiment, the co-occurrence information table created from the keyword or co-occurrence information of the team or organization and the co-occurrence information table created from the keyword and co-occurrence information of the upper engineer are used as comparative materials. By adding a co-occurrence information table created from keywords and co-occurrence information based on information, it becomes possible to compare with a target field in the industry.

(5-4)
In the co-occurrence information table created from keywords and co-occurrence information in a team or organization, there is a possibility that many co-occurrence keywords having the same degree of co-occurrence or low relevance are arranged because the population is wide. In such a case, for example, in consideration of the current phase such as the development phase, the co-occurrence keywords with high priority may be color-coded so as to stand out from the category to which the keywords belong. Or you may color-code into a co-occurrence keyword in order with high importance in an organization.

(6) Others (6-1)
As a method of displaying the recommendation content, it is also possible to categorize the keywords and show the relationship between each category and the point with a radar chart so that the strength and strength can be seen. A radar chart is created for each person and for a team or organization.

  Further, not only the display but also the recommendation information creation pre-processing, the keywords may be categorized, and each category unit may be the original data for creating the recommendation information.

(6-2)
The creator may be designated from the outside. Examples of the external include “UI input”, “corresponding file between file and creator”, and “document specifying the owner of the file”.

(6-3)
“Creator” and “person” are not limited to humans, but also include “department” and “sales”.

(6-4)
The “person keyword table” and the “co-occurrence information table” are created in the CPU regardless of whether or not they are displayed on the monitor, and are stored in a predetermined storage area for a short period or a long period.

DESCRIPTION OF SYMBOLS 10 Recommendation production apparatus 30 Keyword extraction part 40 Creator extraction part 50 Person keyword extraction part 60 Co-occurrence information extraction part 70 Recommendation creation part

JP 2002-560001 A

Claims (5)

A keyword extraction unit (30) for extracting keywords from a large number of document data created and accumulated by each person;
A creator extracting unit (40) for extracting a creator of the document data;
A person keyword extraction unit (50) for creating and extracting association information between the creator and the keyword;
A co-occurrence information extracting unit (60) for creating and extracting co-occurrence information for each creator by giving a co-occurrence keyword to the keyword associated with the creator;
A recommendation creating unit (70) for creating a recommendation content for the creator based on a keyword having a high appearance rate in a document of a comparison subject or a group of comparison subjects for each creator but having a low appearance rate in a self document; ,
With
The recommendation creating unit sets a keyword having a low occurrence rate in a self document even though a degree of co-occurrence with respect to a keyword having a high appearance rate in the self document is large in the document of the comparison target or the comparison target group. Display the recommendations based on the priority,
Recommendation creation device (10). Further, the recommendation creating unit (70) creates a recommendation content based on a keyword having a low appearance rate in the comparison target document and a high appearance rate in the self document.
The recommendation creating apparatus (10) according to claim 1. The content of the recommendation includes a content for presenting a contact for making up for information that is lacking in itself.
The recommendation creating apparatus (10) according to claim 1 or 2. The keyword extraction unit (30) determines an upper limit of the number of keywords extracted from one document data.
The recommendation creation apparatus (10) according to any one of claims 1 to 3 . The keyword extraction unit (30) determines an upper limit of the quantity of the keywords extracted from one document data according to the size of the document data.
The recommendation creation apparatus (10) according to any one of claims 1 to 3 .

Images