JP5677262B2 - Article distribution setting support apparatus, method, and program - Google Patents

Description

Embodiments described herein relate generally to an article distribution setting support apparatus, method, and program for supporting distribution settings for e-mail and the like in an information system.

In a conventional communication system, there are community tools such as groupware, and an information providing site by e-mail. In such a system, there is one in which a user can designate information received as an e-mail from posted information by specifying a condition. In such a mail distribution setting related to article distribution, it is possible to control whether or not to perform mail distribution of a newly posted article by specifying an article condition.

FIG. 6 is a diagram illustrating an example of a conventional mail delivery setting.

The user directly specifies the conditions for receiving the mail of the posted article by the label (“contact” or “important”) attached to the article, the action executed when the article is posted, or the like. The article action here refers to the type of article and what it is intended to do or for what purpose ("approval request" or "confirmation request") It is.

With such a delivery setting, the user can select and receive only information related to him / her by mail.

However, in the above-described conventional method, it is necessary to explicitly specify an article label, an action, an included keyword, and the like as mail reception conditions. Since the user needs to know in advance the judgment materials for specifying conditions such as the keywords included in the article, the meaning of labels and actions, and the frequency of use, there is an aspect in which setting conditions is difficult. Therefore,
On the user side, there is a risk of error in setting conditions and leakage. On the other hand, there may be a case where an article submitter forgets to attach a label to an article, so that an article to be delivered is not delivered even if the condition setting is correct. As described above, in the conventional technique, there is a possibility that the distribution article is leaked.

Japanese Patent No. 4393482 Japanese Patent No. 4745419

The problem to be solved by the present invention is to provide a technique capable of reducing the mail delivery omission of newly arrived articles to the user.

According to an embodiment of the present invention, a past article that has been posted in an information system that has a function of posting an article from a user and a function of distributing an article to a user and having a storage unit that stores the posted article. An article, its posting user, and related information between articles are stored. The past articles stored in the storage means are clustered to obtain a classification category and articles belonging to this classification category. With reference to the storage means, a category relevance level indicating a relevance level between the user and the category is obtained. When setting conditions for delivering posted articles to users, the classification category and the category relevance are presented.

Conceptual diagram of mail delivery setting of the information system of this embodiment The figure which shows an example of calculation of a category relevance The block diagram which shows the whole structure of the information system of this embodiment Flow chart of category relevance calculation processing Flow chart of processing for presenting classification category list and category relevance to user The figure which shows an example of the conventional mail delivery setting

FIG. 1 is a conceptual diagram of mail delivery setting of the information system of this embodiment.

In the present embodiment, for example, an information system such as groupware having a function of posting and distributing articles by e-mail (collaboration function) used by a plurality of users is assumed. Clustering (hereinafter referred to as “automatic classification”) is performed on the past article DB 101 that stores the group of articles posted to the groupware, etc., and the degree of association with the user of the generated category is calculated. Present as a candidate to the user. The user designates the condition of an article to be received by selecting a category from the presented items.

Specifically, a classification category that is a classification result obtained by automatic classification is set as one of the mail reception conditions. When the user performs mail delivery setting, the obtained category category list 102 is presented as an option.

The mail delivery setting screen 103 in FIG. 1 is an example. When presenting the classification category, a “category relevance” indicating the relevance between the user and the classification category is presented to the user for each classification category. The category relevance is a value calculated based on how much “the original article of the reply article posted by the setter in the past” and “the article bookmarked by the setter” are included.

By supporting email delivery settings, users can specify conditions such as what category a new post belongs to, rather than explicitly or fixedly specifying a specific label, etc. E-mail delivery can be controlled. Therefore, omission of article mail delivery can be reduced.

(1) Automatic classification of past articles All articles posted in the past to the groupware or the like in advance (including each group of posted articles for tasks 1, 2, and 3) are used as a group for automatic classification. Let's say articles.

Automatic classification is performed on the article group as the population, and among the obtained classification categories, some of the higher ranks are held as a classification category list 102. Some of the higher ranks are obtained under conditions that are equal to or higher than a preset threshold. For example, it is assumed that “approval request”, “internal regulations”, “revision of regulations”, “operation rules”, and the like are obtained. In addition, for each category of the classification category list 102,
Keep a list of articles belonging to a category.

Note that the automatic classification performed on past articles uses, for example, the existing technology disclosed in Patent Document 2.

(2) Method for calculating category relevance Next, for each user of this system, the “category relevance” indicating the relevance between the user and the classification category is obtained by identifying the user's identification information and the past information belonging to the classification category. Based on the posted article, it is calculated for each classification category. Specifically, the category relevance is calculated according to the following equation.

Category relevance = User category related articles / Total number of articles belonging to category

Number of articles related to the category of the user = Number of articles that belong to the category and that meet either of the following conditions

(A) When a user replies to an article belonging to a category (regardless of mail or web), the article is a category-related article of the user.

(B) When the user bookmarks an article belonging to a category, the article is an article related to the user's category.

The calculated “category relevance” is totaled to obtain a “category relevance” list for each category / user.

FIG. 2 is a diagram illustrating an example of calculating the category relevance.

In this figure, from the past article information and its automatic classification results, the “category relevance” for each user
Is calculated in advance.

Past article information (articles A to E) is accumulated in the past article DB 101.
FIG. 2 illustrates a task 1 including articles A, B, and C and a task 2 including articles D and E.

The poster of the article A is the user A. The author of article B, which is a reply to this article A, is user B
It is. The contributor of the article C that is a reply to the article B is the user C.

The poster of the article D is the user B. The contributor of the article E which is a reply to the article D is the user C
It is. Note that the user C bookmarks the article E.

The result of automatically classifying the contents of the past article DB 101 is a classification category list 102. The classification category list 102 in FIG. 2 shows a part of the classification category list 102 in FIG. 1 in more detail, and a list of articles belonging to the category is held for each category.

Articles A, B, and D are automatically classified in the category “approval request”.
Articles C and E are automatically classified in the category “internal regulations”.
In addition, the contributor of each article remains as shown in the past article DB 101.

A list 201 in FIG. 2 is a list of reply source articles, reply articles, and posters that match the above condition (a) based on the contents of the past article DB 101 and the category category list 102.

“Request for approval” is collected in the item “category category of reply source article”, and articles A, B, and D are classified in the item of “reply source article”. These two items are classified category list 102.
The result is the same.

The item “reply article” indicates each reply article corresponding to each “reply source article”. Past article D
Based on the content of B101, article B for article A, article C for article B, article D
Article E is a reply article.

The item “poster of reply article” indicates the poster of each “reply article”. Past article DB1
Based on the content of 01, user B for article B, user C for article C, article E
User C is a contributor.

Further, based on the contents of the past article DB 101 and the classification category list 102, the above condition (
A list 202 is a list of bookmark articles that match b).

Referring to the past article DB 101, only the user “bookmark” is “bookmark C”, and therefore the user C is shown in the “bookmark” item of the list 202. The user C is bookmarking the article A, which is shown in the “article” item. In addition, the “category category” of this article A is based on the result of the category list 102.
"Approval request".

By merging the information of the list 201 and the list 202, the category relevance 203 for each user is obtained.

User B and user C are listed as “users”. Note that the user A is excluded (not shown) from this list, but the relevance of the user A depends on the above conditions (a) and (b).
This is because “0%” was calculated.

In the list 201 and the list 202 in FIG. 2, “request for approval” is listed as the “classification category” regarding the user B. When the “category relevance” is applied to the above formula, the category relevance = the number of articles related to the category of the user / the total number of articles belonging to the category = (user B with respect to the articles belonging to “request for approval” in the list 201) The reply is article A, that is, 1) / (the number of articles belonging to “approval request” in the category category list 102 is articles A, B, and D, that is, 3) = 1/3 = 33%.

In the list 201 and the list 202 in FIG. 2, “classification category” regarding the user C includes “approval request” and “internal regulations”. User C for classification category "Request for approval"
“Category relevance” is applied to the above formula, category relevance = the number of articles related to the category of the user / the total number of articles belonging to the category = ((user C for articles belonging to “approval request” in the list 201) Replies to articles B and D, that is, 2) / (the number of articles belonging to “approval request” in the classification category list 102 is articles A, B, and D, ie 3) = 2/3 = 66% It becomes.

Further, the “category relevance” of the user C regarding the classification category “internal regulations” is applied to the above formula, the category relevance = the number of articles related to the category of the user / the total number of articles belonging to the category = ((“ Article E belonging to “internal rules” and bookmarked by user C is article E, ie 1) / (the number of articles belonging to “internal rules” in classification category list 102 is articles C, E ie 2) = 1/2 = 55%.

Next, presentation of the classification category list 102 and category relevance to the user will be described.

First, the classification category list 102 and the category relevance level for each classification category of the user whose mail delivery is to be set are acquired (see the category relevance level 203 for each user in FIG. 2).

Next, the acquired “category category list” and “category relevance for each user category” are presented to the user as setting conditions for the mail delivery setting. At that time, classification categories are presented in descending order of category relevance.

For example, when the mail delivery setting screen 103 shown in FIG. 1 is presented to the user C, “automatic classification category” is “approval request” (category relevance is 66%), “internal regulations” (category relevance is 50 %), And it is possible to select whether or not to deliver (on / off), respectively.

The effect of being able to set the classification category on the mail delivery setting screen 103
By specifying the classification category list 102, it can be classified into any one (however, since the classification category list 102 is some higher-order classification, it may not correspond to any of them).

When a newly arrived article is classified into the category for which the user selects “deliver” in the mail delivery setting, the newly arrived article matches the conditions for mail delivery and is delivered to the user by mail.

FIG. 3 is a block diagram showing the overall configuration of the information system of this embodiment.

A user 301 is a user of the information system 300.

The user interface 302 notifies the user interface control unit 303 of an operation of the user 301 on the user interface.

The user interface control unit 303 transmits the classification category list 102 and category association degree acquisition request from the user interface 302 to the classification category list acquisition unit 304 and the category association degree acquisition unit 305.

The category relevance calculating unit 306 performs automatic classification for all articles based on the information in the past article DB 101 in advance, obtains a classification category list 102, calculates the category relevance for each category and for each user, Information of the category category list 102 and the category relevance level by user for each category is stored in the category DB 307.

The classification category list acquisition unit 304 acquires, from the category DB 307, the classification category list 102 that is the result of automatic classification (previously calculated) in response to the acquisition request of the classification category list 102 from the user interface control unit 303.

The category association degree acquisition unit 305 acquires the category association degree of the user 301 from the category DB 307 for each category in the classification category list 102 in response to the category association degree acquisition request from the user interface control unit 303.

The past article DB 101 is a database that stores (accumulates) articles posted in the past. Stores all information related to the article (article identifier (ID), title, text, contributor, posting time, relations between articles, etc., such as links between articles).

The category DB 307 is a database that stores (accumulates) the classification category list 102 and the category relevance level for each category / user. The category relevance calculation unit 306 determines the category relevance level for each category and user by category from the state of the past article DB 101 at an arbitrary timing (snapshot: calculated once a month, for example). Calculate in advance.

In this embodiment, the user interface 302 and the user interface control unit 303
The category category list acquisition unit 304, the category relevance level acquisition unit 305, and the category relevance level calculation unit 306 can be implemented as a program for causing a computer processing apparatus in the article distribution setting apparatus 308 to execute these processes. Each program is stored in a storage device such as a memory of a computer. The past article DB 101 and the category DB 307 are stored as a database in a storage device such as an HDD in the information system 300. The user interface 302 includes processing for displaying the mail delivery setting screen 103 shown in FIG. 1 via a display for the user 301 and inputting with a keyboard, a mouse, and the like.

Each processing means of the article distribution setting support device 308 is not limited to a single device, and can be implemented in various forms such as a client / server, ASP / SaaS / cloud computing. is there.

FIG. 4 is a flowchart of the category relevance calculation process.

The category relevance calculation unit 306 acquires all article information from the past article DB 101 and executes automatic classification. The category category list 102 shown in FIG. "Category article list") is obtained (step S101).
).

The category relevance calculation unit 306 stores the acquired classification category list 102 and category article list in the category DB 307 (step S102).

  Then, the following is repeated for each article for the entire article list (step S103).

If the article is not a reply to another article but the first article, do nothing and proceed to the next article
(continue) (step S104).

If the article is a reply article, the identifier (ID) of the article poster (user) and the category to which the reply source article belongs are examined, and the obtained category of the reply source article, the user ID of the article poster, the article ID
Is stored (step S105). This holding may be performed anywhere on the internal or external storage medium.

When the article is bookmarked by a certain user, the ID of the bookmarking user and the category to which the article belongs are checked, and the information in which “article category, user ID, article ID” is paired as in the previous step S105. Is held (step S106).

This is the end of the repetition from step S103 (step S107).

Next, for all classification categories, the number of all articles belonging to the category is stored from the category article list (step S108).

For all users, the following is repeated for each user (step S109).

For each classification category, search the stored pairing information by category and user ID, and the corresponding number of items (
The number of information items = the number of articles related to the user in the articles belonging to the category) is divided by the total number of articles belonging to the category, and the quotient is calculated as the “category relevance” for the user category. “Category relevance” is stored in the category DB (step S110).

This is the end of the repetition from step S109 (S111).

FIG. 5 is a flowchart of the process of presenting the classification category list 102 and the category relevance to the user.

The user 301 requests the user interface 302 to acquire the classification category list 102 and the category relevance level of the user. At that time, the user ID of the user is passed to the user interface 302 (step S201).

The user interface 302 passes the user ID to the user interface control unit 303,
The acquisition of the category category list 102 and the category relevance of the user is requested (step S2).
02).

The user interface control unit 303 requests the classification category list 102 from the classification category list acquisition unit 304. At that time, the user ID of the user 301 is displayed in the classification category list acquisition unit 304.
(Step S203).

The classification category list acquisition unit 304 acquires the classification category list 102 from the category DB 307 and returns it to the user interface control unit 303 (step S204).

The user interface control unit 303 passes the category and the user ID to the category relevance obtaining unit 305 for each category of the classification category list 102 and requests acquisition of the category relevance (
Step S205).

The category relevance obtaining unit 305 obtains the category relevance from the category and the user ID, and returns information on “category, user ID, category relevance” to the user interface control unit 303 (step S206).

The user interface control unit 303 acquires the category relevance as many as the number of categories, and returns the result to the user interface 302 (step S207).

The user interface 302 presents the list of classification categories and the category relevance level of the user for each category to the user 301 (step S208).

The user 301 performs mail delivery setting using the category relevance information as a criterion (step S209).

As described above, according to this embodiment, in the information system having the automatic classification function of posted articles, the distribution setting is performed while referring to the category relevance for the presented classification category when performing the mail distribution setting. By setting, the user can reduce the omission of new article distribution to the user by designating the classification without explicitly designating the label or action.

Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and the equivalents thereof.

101 ... Past article DB
102 ... Classification category list 103 ... Mail distribution setting screen 201 ... List of reply source articles and reply articles that match the condition (a) 202 List of articles 203 ... Category relevance for each user 300 ... Information system 301 ... User 302 ... User interface 303 ... User interface control unit 304 ... Classification category list acquisition unit 305 -Category relevance acquisition unit 306 ... Category relevance calculation unit 307 ... Category DB
308 ... Article distribution setting support device

Claims (3)

In an information system having a function for posting an article from a user and a function for distributing an article to a user, and having a storage means for storing the posted article,
Past article storage means for storing past articles that are posted articles and their posting users and related information between articles;
Clustering means for clustering the past articles stored in the past article storage means to obtain classification categories and articles belonging to the classification categories;
Referring to the past article storage means, the degree of relevance with a specific user for a specific classification category is the number of original articles that are the basis of the articles posted by the user in response to the classification category or Category relevance obtaining means for calculating based on at least one of the number of articles bookmarked by the user classified into the classification category ;
An article distribution setting support apparatus comprising: a presentation unit that presents the classification category and the category relevance when setting a condition for distributing the posted article to a user. In an article distribution setting support method in an information system having a function for posting an article from a user and a function for distributing an article to a user, and having a storage means for storing the posted article,
Storing a past article that is a posted article, its posting user, and related information between articles;
Clustering the past articles stored in the storage means to obtain a classification category and articles belonging to the classification category;
Referring to the storage means, the degree of relevance with a specific user for a specific classification category is determined based on the number of original articles that are the basis of the articles posted by the user in response to the classification category or the user Calculating based on at least one of the number of articles that are bookmarked by the classification category ,
A method of presenting the classification category and the category relevance when setting a condition for delivering the posted article to a user. A computer in an information system having a function for posting an article from a user and a function for distributing an article to the user, and having a storage means for storing the posted article.
A procedure for storing past articles that are posted articles, their posting users, and related information between articles;
A procedure for clustering the past articles stored in the storage means to obtain a classification category and an article belonging to the classification category;
Referring to the storage means, the degree of relevance with a specific user for a specific classification category is determined based on the number of original articles that are the basis of the articles posted by the user in response to the classification category or the user A step for obtaining a category relevance calculated based on at least one of the number of articles that have been bookmarked by the classification category ,
A program for executing the procedure of presenting the classification category and the category relevance when setting conditions for distributing the posted article to a user.

Images