JP6492889B2 - Server apparatus, program, and information providing method - Google Patents

Description

  The present invention relates to a server device that provides information to a communication terminal device, a program thereof, and an information providing method.

  In recent years, with the development of networks such as the Internet and communication terminal devices such as personal computers, articles such as books, news, etc., and advertisements of goods or services (hereinafter collectively referred to as “contents”) are transmitted to the network or the network. Electronic browsing through recording media has become common. Recently, electronic books are often browsed through portable portable communication terminal devices such as portable game devices having a communication function, tablet portable terminal devices, and portable telephones.

  Under such circumstances, services for providing content information of content different from the purchased content to users who have purchased content such as books using a network have been provided.

  For example, a user profile is generated by synthesizing a keyword vector of each book previously purchased by the user by category, and a keyword vector is generated for each book based on the book database. And a technique for recommending unpurchased books with a high degree of similarity by calculating the similarity of the keyword vector of each unpurchased book (for example, Patent Document 1).

JP 2001-265808 A

  However, since the system using the technique disclosed in the above-mentioned patent document only calculates a similarity by generating a vector using a keyword, the calculated similarity is constrained by the keyword. As a result, the novelty of the recommended information is gradually lost, and the user cannot obtain unexpected information.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a server device and the like that can provide content information having both usefulness and unexpectedness for the user. is there.

(1) A server device of the present invention for solving the above-described problem is a server device that is connected to a communication terminal device via a network and provides content information regarding content including text, and the content information is registered. A database management means for managing the content information database and a user management database in which user information related to the user is registered, and content information including a sentence stroke characteristic amount obtained by quantifying the character arrangement characteristic in the sentence of the content Content information acquisition means, content information extraction means for analyzing the content and extracting content information including the text stroke feature quantity of the content, the text stroke feature quantity acquired by the content information acquisition means, Content information extraction means is the content Similarity calculation means for calculating similarity by comparing sentence stroke feature amounts extracted from each content registered in the information database, and recommendation information for generating recommendation information based on the calculated similarity It has a configuration comprising generation means and information provision means for providing the generated recommendation information to the user.

  Since the server device of the present invention having this configuration provides the user with information related to the content by calculating the user's preference for the content and the similarity between the content, the server device can provide information that is beneficial to the user. it can.

  In particular, since the similarity is calculated based on the content stroke characteristic amount of the content, it is possible to calculate the similarity with a wide range of content without being limited to categories such as genre and author, which is new to the user. It is possible to provide the user with information that has the benefit of being able to acquire knowledge and the like and the unexpectedness that the user did not expect.

(2) Further, in the server device according to the present invention, the database management unit stores content information including an average sentence writing feature amount calculated based on a sentence writing feature amount of the content purchased by the user. Each content is registered in the user management database, and the content information acquisition unit acquires the average sentence stroke feature amount of a predetermined user from the user information database as the sentence stroke feature amount.

  The server device of the present invention having this configuration can provide information on content that matches the user's preference by using information on content that has been purchased by the user so far.

(3) Moreover, the server apparatus according to the present invention has a configuration in which the sentence stroke feature amount includes a ratio of kana to the entire sentence in the content.

  The proportion of kana has a characteristic that can be called the character density of the sentences that make up the content.Sentences with a high percentage of kana in the whole sentence give the impression that the character density is low, and sentences with a small percentage of kana Gives the impression of high character density.

  Since the server device of the present invention having this configuration calculates the similarity by using the text writing feature amount including the kana ratio, it provides information that matches the user's preference regarding the character density of the text constituting the content. Can do.

(4) Moreover, the server apparatus which concerns on this invention has the structure in which the said text stroke feature-value contains the average value of the number of characters contained in each paragraph which comprises the text in the said content.

  The number of characters included in a paragraph affects the rhythm of the sentence.

  In the server device of the present invention having this configuration, the sentence stroke feature amount includes the average value of the number of characters included in each paragraph, so that it is possible to provide information on content having a rhythm feeling that matches the user's preference.

(5) Further, the server device according to the present invention has a configuration in which the sentence stroke feature amount includes a ratio of a conversation sentence represented by brackets to a sentence in the content.

  The conversational sentence included in the sentence affects the rhythm and image of the sentence.

  The server device of the present invention having this configuration includes the ratio of the sentence stroke feature amount to the sentence in the content of the conversation sentence, and therefore provides information on the content having a rhythm feeling that matches the user's preference. Can do.

(6) In the server device according to the present invention, the database management unit registers the content information including the category to which the content belongs in the content information database for each content, and the purchase by the user. The content information including category information indicating a content category is registered in a user information database, and the recommendation information generation unit generates recommendation information for each category based on the content information acquired by the content information acquisition unit. It has the composition to do.

  Since the server apparatus of the present invention having this configuration generates and provides recommendation information for each category to which the content belongs, it is possible to provide information that matches the characteristics of the category to which the content belongs to the user.

(7) In the server device according to the present invention, the content information registration unit registers the content information further including category information indicating a category of the content purchased by the user in a user information database, and the recommendation The information generation means has only the content having the same category information as the category information or category information different from the category information based on the category information included in the content information acquired by the content information acquisition means. It has a configuration for selecting only content.

  According to the server device of the present invention having this configuration, for example, by generating information on content that belongs to the same category as the content that the user has purchased so far, it is possible to provide information that matches the user's preference. On the other hand, it is possible to provide information that is surprising to the user by generating information on content belonging to different categories.

(8) Further, in the server device according to the present invention, the database management unit registers the content information further including author information specifying the author of the content purchased by the user in a user information database, The recommendation information generation means is based on the author information included in the content information acquired by the content information acquisition means, only the content having the same author information as the author information, or the author information Has a configuration in which only content having different author information is selected.

  According to the server device of the present invention having this configuration, for example, information that matches the user's preference is provided by generating content information about the same author as the author of the content that the user has purchased so far. On the other hand, by providing content information about different authors, the user can be surprised.

(9) Further, in the server device according to the present invention, the database management unit may include the content information further including read information transmitted from the communication terminal device and indicating that the content has been read. The recommendation information generation unit is registered in the database, and has a configuration in which only the content having the read information included in the content information acquired by the content information acquisition unit is selected.

  The user does not always read through all of the purchased content. That is, even if the content is purchased, if the content does not meet the user's preference, the user may stop reading the content halfway. On the other hand, if the user has read all the way to the end, the user can guess that he / she likes the content.

  The server device of the present invention having this configuration can provide the user with information related to content similar to the content that can be presumed to be liked by the user by using already-read information indicating that it has been read. it can.

(10) In the server device according to the present invention, the content information acquisition unit acquires content information including a sentence writing feature amount specified based on the content information related to the content transmitted from the communication terminal device. It has the composition to do.

  Since the server device of the present invention having this configuration uses content information including text writing feature quantities specified based on content information related to content transmitted from the communication terminal device as described above, the user then Can provide information on the desired content.

(11) A program according to the present invention for solving the above-described problem is registered in a server device that is connected to a communication terminal device via a network and provides content information related to content including text. Database management means for managing the prepared content information database and a user management database in which user information related to the user is registered, and acquiring content information including a sentence writing feature value obtained by quantifying the character arrangement characteristic in the sentence of the content Content information acquisition means, content information extraction means for analyzing the content and extracting content information including the text stroke feature quantity of the content, the text stroke feature quantity acquired by the content information acquisition means, and the content information extraction Means said content Similarity calculation means for calculating similarity by comparing sentence stroke feature amounts extracted from each content registered in the information database, and recommendation information generation for generating recommendation information based on the calculated similarity And an information providing means for providing the generated recommendation information to the user.

  Since the program of the present invention having this configuration provides the user with information related to the content by calculating the user's preference for the content and the similarity between the content, it is possible to provide information that is beneficial to the user. .

In particular, since the similarity is calculated based on the content stroke characteristic amount of the content, it is possible to calculate the similarity with a wide range of content without being limited to categories such as genre and author, which is new to the user. It is possible to provide the user with information that has the benefit of being able to acquire knowledge and the like and the unexpectedness that the user did not expect.

(12) Further, an information providing method according to the present invention for solving the above-described problem provides content information related to content including text from a server device to a communication terminal device connected to the server device via a network. A database management step for managing a content information database in which the content information is registered and a user management database in which user information about the user is registered; and content information including a sentence writing feature amount of the content A content information acquisition step for acquiring, a content information extraction step for analyzing the content, and extracting content information including a sentence stroke characteristic amount obtained by quantifying the character arrangement characteristic in the text of the content, and the content information acquisition The sentence stroke characteristic amount acquired by the means The similarity calculation step of calculating the similarity by comparing the content stroke extraction feature amount extracted from each content registered in the content information database by the content information extraction means, and the calculated similarity A recommendation information generating step for generating recommendation information based on the information, and an information providing step for providing the generated recommendation information to the user.

  The method providing method of the present invention having this configuration provides the user with information related to the content by calculating the user's preference for the content and the similarity between the content, and thus provides information that is beneficial to the user. Can do.

  In particular, since the similarity is calculated based on the content stroke characteristic amount of the content, it is possible to calculate the similarity with a wide range of content without being limited to categories such as genre and author, which is new to the user. It is possible to provide the user with information that has the benefit of being able to acquire knowledge and the like and the unexpectedness that the user did not expect.

  The present invention can provide content information that is not only useful to the user but also surprising to the user.

1 is a system configuration diagram showing a configuration of a communication system in an embodiment of the present invention. It is a block diagram which shows the structure of the information provision server apparatus in one Embodiment. It is a figure which shows an example of the data recorded on the user management database in the information provision server apparatus in one Embodiment. It is a figure which shows an example of the data recorded on the electronic book database in the information provision server apparatus in one Embodiment. It is explanatory drawing for demonstrating the preference of each user's content with respect to a content, and the calculation of the similarity with a content in the similarity calculation part of one Embodiment. It is a flowchart which shows the operation | movement of the process performed in the server apparatus of one Embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  In the following embodiments, a server device, a program, and information provision according to the present invention are provided for a communication system having a communication terminal device and an information providing server device connected to the communication terminal device via a network. It is embodiment at the time of applying a method.

  In addition, the “content” of the present invention includes paper media documents such as books, pamphlets, and advertisements, and digital contents such as electronic books and electronic advertisements. However, hereinafter, the present embodiment will be described using an electronic book as a representative example of the content of the present invention.

[1] Overview of Communication System First, an overview of the communication system 1 according to the present embodiment will be described with reference to FIGS. 1 and 2. FIG. 1 is a system configuration diagram showing a configuration of the communication system 1 in the present embodiment, and FIG. 2 is a conceptual diagram showing an operation outline of the communication system 1 in the present embodiment. Further, in order to prevent the figure from becoming complicated, only some users are shown in FIGS. 1 and 2. That is, in the actual communication system 1, there are a larger number of users and communication terminal devices 10 than displayed.

  The communication system 1 according to the present embodiment distributes electronic books that are digital contents, calculates a sentence writing feature amount that quantifies a sentence writing characteristic indicating the character arrangement characteristics of each electronic book, and calculates the calculated sentence writing match. This is a system for distributing information recommending an electronic book to a user based on a feature amount.

  In the present embodiment, each user is described as being registered as a member in a network service that distributes electronic books (hereinafter referred to as “electronic book distribution service”).

  In order to realize such a function, the communication system 1 according to the present embodiment provides a plurality of communication terminal devices 10 possessed by each user, a network 20, and an electronic book distribution service, as shown in FIG. Information providing server device 30.

  Each communication terminal apparatus 10 is connected to the network 20 via the base station 50. The network 20 is configured by interconnecting a public telephone network including a mobile telephone network and an IP (Internet Protocol) network, for example. However, the configuration of the network 20 is not limited to this.

  The communication terminal device 10 has a function of an electronic book reader. For example, a portable music player, a portable game machine, a tablet information terminal device, a smartphone, a portable telephone, a personal computer, or the like that can be carried by a user It is a terminal device.

  In particular, the communication terminal apparatus 10 receives resource data for the WWW system described in a markup language such as XML (extensible Markup Language) and a unique URL (Uniform Resource Locator) indicating a network address of the resource data. While using, it has a browsing function for performing data communication with the information providing server device 30.

  Moreover, the communication terminal device 10 receives the resource data from the information providing server device 30 using the browsing function, and displays the content configured by the resource data so that the user can browse.

  Then, the communication terminal device 10 is connected to the information providing server device 30 via the base station 50 and the network 20, and logs into the electronic book distribution service based on the user's operation. As a “logged-in user”), it is configured to receive various data from the information providing server device 30 and receive provision of an electronic book distribution service.

  In particular, when the user has purchased an e-book using the communication system 1 so far, the communication terminal device 10 is beneficial to the user and has unexpectedness based on the purchase history of the e-book. It is configured to be able to receive provision of content information related to an electronic book.

  Further, the communication terminal device 10 is configured such that the user himself / herself distributes predetermined content information to the information providing server device 30 to obtain content information related to the electronic book selected based on the content information. Has been.

  The information providing server device 30 is a system used for providing an electronic book distribution service, and has various databases (hereinafter also referred to as “DB”) in order to provide the electronic book distribution service. And the information provision server apparatus 30 performs the various processes for delivering the content information regarding the electronic book which has usefulness and unexpectedness for a user.

And this information provision server apparatus 30 is:
(1) Configuration for managing a content information database in which content information related to content such as electronic books is registered;
(2) A configuration for acquiring content information including a sentence stroke characteristic amount of content,
(3) Configuration for analyzing content and extracting content information including sentence stroke feature amount of the content,
(4) Sentence writing feature amount extracted from each content registered in the content information database with a configuration for extracting content information including the sentence writing feature amount acquired by the configuration for acquiring content information and the content writing stroke feature amount of the content And a configuration for calculating the similarity by comparing
(5) A configuration for generating recommendation information for recommending content information to the user based on the calculated similarity,
(6) a configuration for providing the generated recommendation information to the user;
have.

  That is, the information providing server device 30 compares the text stroke feature amount of the content such as the electronic book registered in the content information database with the average value of the text stroke feature amount of the content such as the electronic book that the user has already purchased. The similarity between the two is calculated.

  However, the information providing server device 30 extracts the text stroke feature amount based on the content information transmitted from the communication terminal device 10, and the extracted text stroke feature amount and an electronic book registered in the content information database, etc. It is also possible to compare the writing stroke feature amount of the content of the content and calculate the similarity between the two.

  And the information provision server apparatus 30 can generate | occur | produce the information regarding the electronic book which has both usefulness and unexpectedness for a user based on the calculated similarity degree, and can provide the produced | generated information to a user now. Yes.

  In the present embodiment, “sentence writing” means “sentence writing” or “sentence form” representing the rhythm of the sentences constituting the content. In the present embodiment, “sentence writing” does not mean writing or carrying a brush. The sentence stroke feature amount is a value that quantitatively represents the feature of the sentence stroke that represents the feature of the sentence constituting the content.

In this embodiment, as the text stroke feature amount, for example,
(1) The ratio of kana to the entire text of the electronic book,
(2) The average number of characters included in each paragraph constituting the text of the electronic book,
(3) The ratio of the conversation sentence represented by the brackets to the text of the electronic book,
The numerical arrangement of the character arrangement in a sentence such as

  By having the above configuration, the communication system 1 of the present embodiment is beneficial to the user from among electronic books registered in the database based on the purchase history of the user's past electronic books and the content information proposed by the user. It is possible to provide information on various electronic books.

  In particular, since the communication system 1 according to the present embodiment calculates the similarity based on the text stroke feature amount, it is possible to prevent the information provided to the user from being ruined and provide the user with unexpected information. Be able to.

[2] Information Providing Server Device [2.1] Configuration Next, the configuration of the information providing server device 30 of the present embodiment will be described with reference to FIG. FIG. 2 is a block diagram showing the configuration of the information providing server device 30 of this embodiment.

  As shown in FIG. 2, the information providing server device 30 of the present embodiment includes a communication control unit 310 connected to the network 20, a ROM / RAM 320 functioning as various memories, and various databases (hereinafter, “DB”). And a server management control unit 340 for controlling the entire apparatus, distribution of electronic book data, collection of electronic book history information, registration of comment information, and provision of the comment information A data processing unit 350.

  The above-described units are connected to each other by a bus B, and data transfer is performed between the components.

  The communication control unit 310 is a predetermined network interface, establishes a communication channel with the communication terminal device 10, and exchanges various data.

  Various programs necessary for driving the information providing server device 30 are recorded in the ROM / RAM 320. The ROM / RAM 320 is used as a work area when various processes are executed.

  The recording device 330 is configured by an HDD or an SSD (Solid State Drive). The recording device 330 includes a user management DB 331 that stores information necessary for the user to receive the electronic book distribution service, and a content information DB 332 that stores content information about the electronic book to be distributed. Yes.

  Note that the user management DB 331 of this embodiment constitutes a predetermined database of the present invention, and the content information DB 332 constitutes a content information database of the present invention. Details of the recording apparatus 330 of this embodiment will be described later.

  The server management control unit 340 is mainly configured by a central processing unit (CPU), and integrally controls each unit of the information providing server device 30 by executing a program.

  When the data processing unit 350 obtains a login request based on a user instruction from the communication terminal device 10 in conjunction with the recording device 330 under the control of the server management control unit 340, the data processing unit 350 receives the electronic information of each user based on the request. Log in to the book distribution service.

Then, the data processing unit 350
(1) Processing for managing the user management DB 331 in which user information is registered and the content information DB 332 in which content information such as electronic books is registered;
(2) Acquisition processing for acquiring content information including an average sentence writing feature amount for each user or a sentence writing feature amount obtained based on information distributed from the user;
(3) An extraction process of analyzing content such as an electronic book and extracting text information of content information including a text stroke feature amount of the content,
(4) A calculation process for calculating the similarity by comparing the acquired sentence writing feature quantity with the extracted sentence writing feature quantity;
(5) Generation processing for generating recommendation information related to an electronic book based on the calculated similarity, and
(6) Providing processing for providing the generated recommendation information to the user,
It is carried out.

The data processing unit 350 further includes
(7) Register content information regarding an electronic book including an average sentence writing feature amount calculated based on a sentence writing feature amount of content subscribed by a user in a predetermined database for each user.
Processing is in progress.

  Specifically, the data processing unit 350, by executing an application, a DB management unit 351 that manages each DB, a content information extraction unit 353 that executes an extraction process that extracts text information from content information, and content information A content information acquisition unit 352 that executes an acquisition process for acquiring a similarity, a similarity calculation unit 354 that executes a calculation process that calculates a similarity, a recommendation information generation unit 355 that executes a generation process that generates recommendation information, And an information providing unit 356 that provides recommendation information.

[2.2] Various DBs
Next, each DB of this embodiment will be described with reference to FIGS. 3 and 4. 3 and 4 are diagrams illustrating examples of data recorded in the user management DB 331 and the content information DB 332 provided in the information providing server device 30, respectively.

The user management DB 331 of this embodiment is a database in which various types of information for managing each user registered for use in the electronic book distribution service is stored as data. For example, in the user management DB 331, as shown in FIG.
(1) a user ID;
(2) login information (login ID, password);
(3) Information on the purchased book list (book ID);
(4) Information on the average sentence stroke feature amount (a kana ratio, average number of paragraph characters, conversation sentence ratio) of a purchased book;
Are registered (recorded) in association with each other.

  For example, in FIG. 3, “login ID: ○ Δ ×... / Password: 123abc...” Is registered and purchased in the user management DB 331 as login information corresponding to the user with the user ID “user101”. The book IDs “DC001”, “DC003”, “DC004”, and “DC010” are registered as electronic books, and the kana ratio “0.43”, the average number of paragraph characters “36” as the average sentence stroke feature amount of the purchased books. "And the conversational sentence ratio" 0.21 "are registered.

The content information DB 332 is a DB in which content information such as an electronic book provided to the user via the communication terminal device 10 is recorded. For example, in the content information DB 332, as shown in FIG. 4, for each electronic book (content), (1) an electronic book ID and
(2) the title of the electronic book;
(3) Body data of the electronic book (image data, text data, and control data in the electronic book reader),
(4) Image data such as a cover,
(5) Author name and author ID;
(6) Category name and category ID;
(7) Sentence writing features (kana ratio, average number of paragraph characters, conversation sentence ratio),
Are registered in association with each other.

  For example, in FIG. 4, the content information DB 332 is associated with the electronic book ID “DC001”, the title of the electronic book “Introduction to XX Law”, the body data of “DB001”, and the “image data” Image data of “1”, the author name and author ID of “Thu XX: AT001”, the category name and category ID of “law document: CT010”, and the kana ratio “0. 12 ”, the average number of paragraph characters“ 56 ”, and the conversation sentence ratio“ 0.08 ”are recorded. The electronic book ID, author ID, and category ID are assigned automatically by the information providing server device 30 when electronic book data is newly registered in the content information DB 332 or based on electronic book data already registered. It is supposed to be.

[2.3] Details of Data Processing Unit [2.3.1] DB Management Unit First, details of the DB management unit 351 of the data processing unit 350 of this embodiment will be described.

  The DB management unit 351 in the present embodiment constitutes a database management unit in the present invention, and performs processing such as writing and reading information to the user management DB 331 and reading information from the content information DB.

  Specifically, the DB management unit 351 executes processing for registering content information related to the electronic book in the content information DB 332 in advance. At that time, the DB management unit 351, for each electronic book, for example, as shown in FIG. 4, “book ID”, “title”, “text data”, “cover image”, “author / author ID”. ”,“ Category / Category ID ”, and sentence writing feature amount (kana ratio, average number of paragraph characters, conversation sentence ratio) are registered as content information in the content information DB 332.

  The DB management unit 351 also executes processing for additionally registering content information related to the electronic book.

  On the other hand, when the user purchases an electronic book, the DB management unit 351 executes processing for registering content information regarding the electronic book in the user management DB 331 for each user. At that time, the DB management unit 351 determines, for each user, the average sentence stroke feature amount calculated based on the sentence stroke feature amount of the electronic book purchased by each user and the sentence stroke feature amount of the electronic book purchased this time. Register with.

  In addition, the DB management unit 351 executes processing for reading content information including the average sentence writing feature amount registered in the user management DB 331 and delivering the content information to a content information acquisition unit described later.

[2.3.2] Content Information Acquisition Unit Next, details of the content information acquisition unit 352 of the data processing unit 350 of the present embodiment will be described.

  The content information acquisition unit 352 of the present embodiment constitutes content information acquisition means of the present invention, acquires content information for a predetermined user from the user management DB 331, and is transmitted from the user via the communication terminal device 10. A process of acquiring content information obtained based on the information is executed.

  In particular, the content information acquisition unit 352 performs processing for acquiring, from information registered in the user management DB 331, an average sentence writing feature amount unique to the user including a kana ratio, an average number of paragraph characters, and a conversation sentence ratio. .

  Note that the content information acquisition unit 352 also performs processing for acquiring content information based on information transmitted from the user via the communication terminal device 10. In that case, based on information (for example, the title of the electronic book) transmitted from the user, the content information extraction unit 353 described later extracts sentence stroke feature quantities including a kana ratio, an average number of paragraph characters, and a conversation sentence ratio. The content information acquisition unit 352 acquires the sentence stroke feature amount extracted by the content information extraction unit 353.

  However, when the information transmitted from the user via the communication terminal device 10 is a sentence writing feature amount including the kana ratio, the average number of paragraph characters, and the conversation sentence ratio, the content information acquisition unit 352 is connected to the user via the communication terminal device 10. The information sent from is acquired as it is.

[2.3.3] Content Information Extraction Unit Next, details of the content information extraction unit 353 of the data processing unit 350 of the present embodiment will be described.

  The content information extraction unit 353 constitutes content information extraction means of the present invention, and executes a process of extracting sentence writing feature amounts by analyzing content such as an electronic book. The sentence stroke feature amount extracted by the content information extraction unit 353 is information that forms the core of the content information in the process executed by the information providing server device 30 in the present embodiment.

Specifically, the content information extraction unit 353 analyzes sentences constituting content such as each electronic book,
(1) extracting the ratio of kana to the whole sentence, such as how much the kana occupies in the whole sentence;
(2) extracting an average value of the number of characters included in each paragraph constituting the sentence;
(3) Each process of extracting the ratio of the conversation sentence represented by the brackets to the whole sentence is executed. Details of each process will be described later.

  The sentence stroke feature amount extracted by the content information extraction unit 353 is registered in the content information DB 332 for each electronic book by the DB management unit 351 (see FIG. 4).

[2.3.4] Similarity Calculation Unit Next, details of the similarity calculation unit 354 of the data processing unit 350 of the present embodiment will be described.

  The similarity calculation unit 354 constitutes similarity calculation means of the present invention. The similarity calculation unit 354 compares the text stroke feature amount included in the content information acquired by the content information acquisition unit 352 with the text stroke feature amount included in the information of each electronic book registered in the content information DB 332. Then, a process of calculating the similarity of the sentence stroke feature amount between the two is executed.

  The similarity calculation unit 354 uses the cosine similarity, the calculation method using the Euclidean distance, and the like, and the sentence stroke feature amount included in the content information acquired by the content information acquisition unit 352 and the content information DB 332 are registered. The degree of similarity between the text stroke feature amount included in the information of each electronic book is calculated. Details of the similarity by the similarity calculator 354 will be described later.

[2.3.5] Recommendation Information Generation Unit Next, details of the recommendation information generation unit 355 of the data processing unit 350 of the present embodiment will be described.

  The recommendation information generation unit 355 constitutes content recommendation information generation means of the present invention, selects a predetermined electronic book based on the similarity calculated by the similarity calculation unit 354, and performs various types of information related to the selected electronic book. Executes processing to generate information. For example, a threshold is set in advance for the similarity, and only an electronic book showing a value with a higher similarity calculated than the threshold is selected. Alternatively, the calculated similarities are compared, and electronic books up to the top Nth (for example, the top third) having a high similarity are selected.

  In addition, when a threshold value is set in advance for the degree of similarity and an electronic book having a higher degree of similarity calculated than the threshold value is selected, if there are a large number of selected electronic books, It can also be configured to select the top Nth (for example, top 3) electronic books with the highest similarity.

  And the recommendation information generation part 355 produces | generates the recommendation information regarding the selected electronic book. For example, the information included in the recommendation information can include the title, cover image, author, and category of the electronic book. The recommendation information may include the user's average sentence stroke feature amount.

[2.3.6] Information Providing Unit Next, the information providing unit 356 of the data processing unit 350 of the present embodiment will be described.

  The information providing unit 356 constitutes information providing means of the present invention. The information providing unit 356 provides the communication terminal device 10 with recommendation information regarding the electronic book generated by the recommendation information generating unit 355 via the network 20.

  The means for the information providing unit 356 to provide information to the communication terminal device 10 is not particularly limited, and for example, the recommendation information can be distributed using e-mail. However, the provision of information is not limited to the use of e-mail, but is performed using an application for browsing books provided in the system of the electronic bookstore or the communication terminal device 10 or using various functions of the SNS. It can also be configured.

[2.4] Details of Processing Contents of Content Information Extraction Unit Next, details of processing executed by the content information extraction unit 353 will be described.

The content information extraction unit 353 extracts a sentence stroke feature amount for each electronic book. Specifically, as described above,
(1) A process of extracting the ratio of the kana to the entire sentence as to how much the kana occupies the entire sentence (hereinafter referred to as “a kana ratio extraction process”);
(2) A process for extracting the average value of the number of characters included in each paragraph constituting the sentence (hereinafter referred to as “average paragraph number of characters extraction process”), and (3) a conversation sentence represented by brackets is a sentence. A process of extracting the proportion of the total (hereinafter referred to as “conversation sentence ratio extraction process”),
Each process is executed.

[2.4.1] Kana Ratio Extraction Process The content information extraction unit 353 is a process for extracting how much the kana occupies in the entire sentence constituting the electronic book as the kana ratio extraction process. This kana ratio extraction process extracts the feature quantity of sentence writing that should be referred to as the character density of the entire sentence constituting the electronic book, and reflects the ratio of kana in the entire sentence of the electronic book in the information provided to the user. Information that matches user preferences.

  That is, in the kana ratio extraction process, in order to provide a user who prefers a sentence with a low character density and a large ratio of kana to the whole sentence, in order to provide information about an electronic book composed of sentences with a low character density, This is a process of extracting information used to provide information related to an electronic book composed of sentences having a high character density to a user who prefers sentences having a low character density and a high character density.

  The content information extraction unit 353 uses a kana dictionary (hereinafter referred to as “kana dictionary”) prepared in advance, and sorts the entire sentence constituting the electronic book into kana and characters other than kana. In the present embodiment, characters other than kana mean characters other than hiragana, such as kanji, katakana, alphabets, and symbols.

  Then, the content information extraction unit 353 counts the number of characters selected for the kana and the characters other than the kana, respectively, and the total number of characters constituting the text of the electronic book for each character selected for the kana and the characters other than the kana The ratio of the kana to the whole and the ratio of the characters other than the kana to the whole are calculated.

  In addition, the content information extraction unit 353 transfers the value of the ratio of the calculated pseudonym to the whole to the DB management unit 351, and the DB management unit 351 registers it in the content information DB 332. The information registered in the content information DB 332 is a numerical value such as “0.43” as shown in FIG.

[2.4.2] Extraction process of average number of paragraph characters The content information extraction unit 353 extracts an average value of the number of characters per paragraph in the text of the electronic book as the extraction process of the average number of paragraph characters. This process of extracting the average number of paragraph characters is to extract the feature amount of sentence writing, which should be called rhythm, and to reflect the rhythm feeling of the sentence in the information provided to the user to provide information that matches the user's preference This is a process for extracting information used in the process.

  Specifically, the content information extraction unit 353 decomposes the entire sentence into paragraphs by using line breaks provided in the sentences constituting the electronic book.

  Then, the content information extraction unit 353 counts the number of characters constituting each decomposed paragraph, and counts the number of decomposed paragraphs.

  In addition, the content information extraction unit 353 adds the number of characters constituting each paragraph by all the paragraphs, obtains the total number of characters constituting the entire sentence of the electronic book, and divides the total number of characters by the number of paragraphs. Calculate the average number of paragraph characters per paragraph.

  Further, the content information extraction unit 353 delivers the calculated average number of paragraph characters to the DB management unit 351, and the DB management unit 351 registers it in the content information DB 332. The information registered in the content information DB 332 is a numerical value such as “36”, for example, as shown in FIG.

[2.4.3] Conversation Sentence Ratio Extraction Process The content information extraction unit 353 extracts, as a conversational sentence ratio extraction process, how much a conversational sentence occupies in the entire sentence constituting the electronic book. In this conversation sentence ratio extraction process, information used to provide information that matches the user's preference is reflected by reflecting in the information provided to the user the ratio that the conversation sentence occupies in the entire sentence constituting the electronic book. It is processing to do.

  Specifically, the content information extraction unit 353 searches for sentences constituting the electronic book, selects brackets from the sentences, and extracts characters included in the brackets.

  Then, the content information extraction unit 353 counts the number of characters of all extracted characters.

  Further, the content information extraction unit 353 calculates the conversational sentence ratio by dividing the number of characters of all the extracted characters by the total number of characters constituting the text of the electronic book.

  Furthermore, the content information extraction unit 353 delivers the calculated conversation sentence ratio to the DB management unit 351, and the DB management unit 351 registers the content information DB 332 as “conversation sentence ratio”. The information registered in the content information DB 332 is a numerical value such as “0.21” as shown in FIG.

[2.4.4] Calculation process of average sentence stroke feature amount The content information extraction unit 353 also executes an average sentence stroke feature amount calculation process of each user.

  As described above, the average sentence writing feature amount is a value obtained by calculating, for each user, an average value of sentence writing feature amounts of electronic books purchased by each user so far. Therefore, each time the user purchases an electronic book, the content information extraction unit 353 converts the average sentence writing feature amount registered in the user management DB 331 and the sentence writing feature amount of the electronic book purchased this time. Based on this, a new average sentence stroke feature amount is calculated.

  In addition, the content information extraction unit 353 passes the newly calculated average text stroke feature amount to the DB management unit 351, and the DB management unit 351 updates the registration content of the user management DBB 331.

[2.5] Details of Processing Content of Similarity Calculation Unit Next, details of processing content of the similarity calculation processing unit 354 will be described.

  The similarity calculation unit 354 determines whether the predetermined user's preference for the electronic book and each electronic book based on the average sentence writing feature amount of the predetermined user and each sentence writing feature amount of each electronic book. Calculate similarity.

For example, the similarity calculation unit 354 calculates the similarity using the cosine similarity method.
(1) For the user, the vector “UX” shown in (Equation 1) is calculated based on the “Kana ratio”, “Average number of paragraph characters”, and “Conversation sentence ratio”, which are the average sentence stroke feature quantities of a predetermined user. Generate
(2) For electronic books, the vector “Di” shown in (Equation 2) is calculated based on the “sentence ratio”, “average number of paragraph characters”, and “conversation sentence ratio”, which are the sentence stroke feature quantities of each electronic book. Generate
(3) Using a vector space model of the vector “UX” of the predetermined user and the vector “Di” of each electronic book, the similarity (sim) with the predetermined user is expressed by (Equation 3) for each electronic book. As shown, the inner product of the vector “UX” and the vector “Di” is obtained by dividing by the product of the vector “UX” and the vector “Di”.

  In (Expression 1), “X” represents a user such as user A or user B (in Expression 2, “i” represents an ID number of an electronic book).

  A process in which the similarity calculation unit 354 calculates the similarity using the cosine similarity method will be specifically described with reference to FIG. FIG. 5 is an explanatory diagram for explaining processing for calculating similarity using a cosine similarity method.

  As shown in FIG. 5, consider a case where the similarity between the user A and the electronic book is calculated.

  As for the average sentence stroke feature amount of the user A, the kana ratio is “0.33”, the average number of paragraph characters is “54”, and the conversation sentence ratio is “0.53”. Note that the average sentence writing feature amount of the user A is a value obtained by averaging the sentence writing feature amounts of the electronic books that the user A has purchased so far for each item of the kana ratio, the average number of paragraph characters, and the conversation sentence ratio.

  On the other hand, the text stroke feature amount of each electronic book registered in the content information DB is that the book A has a kana ratio “0.43”, an average number of paragraph characters “36.00”, and a conversation sentence ratio. In book B, the kana ratio is “0.32”, the average number of paragraph characters is “50.00”, and the conversation sentence ratio is “0.60”. In other words, the kana ratio is “0.80”, the average number of paragraph characters is “21.00”, and the conversation sentence ratio is “0.55”.

  Based on these values, the cosine similarity with respect to each electronic book of user A is obtained by (Equation 3). As a result, the cosine similarity with respect to the book A is “0.905490442”, the cosine similarity with respect to the book B is “0.999511003”, and the cosine similarity with respect to the book C is “0.815811036”. In the example shown in FIG. 5, the user A has the highest similarity to the book B.

[3] Operation of Communication System Next, processing executed in the information providing server device 30 of the present embodiment will be described using FIG. FIG. 6 is a flowchart showing processing executed in the information providing server device 30 of the present embodiment.

  In this operation, it is assumed that information regarding each user is already registered in the user management DB 331 and information regarding content such as an electronic book is already registered in the content information DB 332. Further, it is assumed that predetermined electronic book data is stored in the communication terminal device 20 and that a predetermined user logs in to the electronic book distribution service using the communication terminal device 20.

  First, the information providing server device 30 determines whether or not a condition for executing this operation is satisfied (S101). For example, in the case where the content information is provided to the user every several days, the condition for executing this operation is determined as to whether or not it is the timing for executing this operation.

  In addition, regardless of whether or not it is the timing to execute this operation, when content information is transmitted from a predetermined user via the communication terminal device 10, the distribution from the user is detected and this operation is executed.

  Next, the information providing server device 30 determines whether or not the condition for executing this operation is distribution from the user (S102).

  When the condition for executing this operation is distribution from the user, the content information acquisition unit 352 of the information providing server device 30 acquires the stroke feature amount obtained based on the distributed information (S103). For example, when the distributed information is the title of the electronic book, the DB management unit 351 searches the content information DB 332 based on the distributed title to read the text of the electronic book, and the content information extraction unit 353 Then, the sentence stroke feature amount including the average number of paragraph characters and the conversation sentence ratio is extracted. The content information acquisition unit 352 acquires the extracted text stroke feature amount.

  However, when the transmitted information is a sentence writing feature amount including the kana ratio, the average number of paragraph characters, and the conversation sentence ratio, the content information acquisition unit 352 acquires the transmitted information as it is.

  On the other hand, when the condition for executing this operation is the timing for executing this operation set in advance, the DB management unit 351 determines the kana ratio, the average number of paragraph characters, and the conversation sentence of a predetermined user from the user management DB 331. The average sentence writing feature amount including the ratio is read, and the content information acquisition unit acquires the read average sentence writing feature amount (S104).

  Next, the information providing server device 30 reads the sentence stroke feature amount of each electronic book from the content information DB 332 (S105).

  Next, the similarity calculation unit 354 calculates the preference between the predetermined user's electronic book and the similarity between each electronic book.

  That is, the similarity calculation unit 354 calculates the similarity based on the text stroke feature amount obtained from the information distributed from the user and the text stroke feature amount of each electronic book registered in the content information DB 332. Alternatively, the similarity calculation unit 354 calculates the similarity based on the average sentence stroke feature amount read from the user management DB 331 and the sentence stroke feature amount of each electronic book registered in the content information DB 332.

  Next, the recommendation information generation unit 355 generates recommendation information based on the calculated similarity (S107). For example, a threshold is set in advance for the similarity, an electronic book whose calculated similarity exceeds the threshold is selected, and a title, a cover image, a category, and the like of the selected electronic book are generated as recommendation information. Alternatively, the top three electronic books with high similarity are selected, and the title, cover image, category, and the like of the selected electronic book are generated as recommendation information.

  Next, the information providing unit 356 provides the generated recommendation information to a predetermined user (S108), and ends this operation.

  As described above, in the present embodiment, information regarding content such as an electronic book, in particular, recommended information regarding the content is provided to the user using the sentence stroke feature amount including the kana ratio, the average number of paragraph characters, and the conversation sentence ratio. Not only is it beneficial, it can provide surprising information.

[4] Modified Example Next, a modified example in the above embodiment will be described.

[4.1] Modification 1
In the information providing server device 30 of the above embodiment, the recommended information is provided to the user by using the text writing feature amount regardless of the category. However, sentence writing feature amounts such as the kana ratio, the average number of paragraph characters, and the conversation sentence ratio vary depending on the category to which the electronic book belongs. For example, literary works and utility books have greatly different characteristics of writing stroke feature quantities.

  Therefore, recommendation information may be provided to the user using a category to which the electronic book belongs.

  Specifically, when registering the average sentence writing feature amount in the user management DB 331, the average sentence writing feature amount is registered for each category.

  When calculating the similarity, the content information acquisition unit 352 acquires the average text stroke feature amount from the user management DB 331 for each category, while reading the text stroke feature amount for each category from the content information DB 332, The outgoing similarity calculation unit 354 calculates the similarity using both values.

  The recommendation information generation unit 355 generates recommendation information for each category, and the information providing unit 356 provides the generated recommendation information for each category to the user.

  As described above, by providing the recommendation information for each category, the accuracy of the recommendation information can be improved, and high-value information can be provided to the user.

  In addition, when providing recommendation information using a category, not only the similarity is calculated using the text stroke feature quantity belonging to the same category as the category to which the target electronic book belongs, The similarity can also be calculated by using the text writing feature amount of the electronic book belonging to a different category. For example, when an electronic book belonging to a pure literature category is targeted, the similarity is calculated using the sentence writing feature amount of an SF novel belonging to a different category.

  The information providing server device 30 can provide information that matches the user's preference by calculating the degree of similarity using the writing stroke feature amount of the electronic book belonging to the same category and providing the recommended information. On the other hand, the information providing server device 30 provides unexpected information to the user by calculating the degree of similarity using the writing stroke feature amounts of electronic books belonging to different categories and providing recommendation information. Can do.

[4.2] Modification 2
You may make it provide recommendation information to a user using the author of an electronic book.

  For example, when generating recommendation information, information on an electronic book of the same author as the author of the electronic book that has been purchased so far by a predetermined user and registered in the user management DB 331 is generated. In this case, the information providing server device 30 can provide information that matches the user's preference.

  On the other hand, when generating the recommendation information, the recommendation information is generated excluding the author of the electronic book that has been purchased by a predetermined user and registered in the user management DB 331. Until now, the information providing server device 30 can provide unexpected information to the user by intentionally excluding the author of the electronic book purchased by the predetermined user.

[4.3] Modification 3
In the information providing server device 30 of the above embodiment, recommendation information is provided using information on an electronic book purchased by a user. However, the information provision server device 30 may provide recommendation information using information of the user's already read books.

  When providing recommendation information using information of a user's already read book, the communication terminal device 10 uses an electronic book (hereinafter referred to as a “reader application”) that functions as an electronic book reader. Specifically, the read notification information indicating that the electronic book has been read when the last page of the electronic book being read is displayed, as well as a function of displaying the electronic book data) Is delivered to the information providing server device 30. On the other hand, the information providing server device 30 registers the read book in the user management DB 331.

  Since the electronic book to which the read notification information is distributed from the communication terminal device 10 has been read by the user to the end, the user is compared with the electronic book to which the read notification information is not distributed from the communication terminal device 10. There is a high probability that it is a favorite e-book.

  By providing recommendation information using information of the user's already read books, the information providing server device 30 can provide information that the user likes.

[4.4] Modification 4
In the above embodiment, when calculating the kana ratio, the kana ratio is calculated by dividing into hiragana and characters other than hiragana. In calculating the kana ratio, in addition to hiragana, katakana, alphabets, numbers, and symbols can be incorporated into the elements of the category.

  When incorporating katakana, alphabets, numbers, and symbols in addition to hiragana into the elements of the category, the information providing server device 30 prepares a dictionary for each element, and calculates the ratio of each element to the entire text of the electronic book. Hold it.

  In addition to hiragana, katakana, alphabets, numbers, and symbols are incorporated into the elements of the category, not only the ratio of only one of the elements in the whole sentence is the kana ratio, but also hiragana, katakana, alphabets, numbers. The ratio of two or more elements arbitrarily combined from the symbols to the entire sentence can be set as the kana ratio.

  For example, alphabets, numbers, and symbols are used more frequently in horizontally written books than in vertically written books. In addition to hiragana, katakana, alphabets, numbers, and symbols are taken into account to calculate the kana ratio and analyze e-books purchased by a given user. Suppose that it turns out that only books are purchased. Since this user is considered to prefer the vertically written electronic book, the user is provided with information related to the vertically written electronic book as recommendation information.

  In this way, by incorporating katakana, alphabets, numbers, and symbols in addition to hiragana, it is possible to analyze user preferences in more detail and provide information that is highly beneficial to users. Can do.

[4.5]
In the information providing server device 30, the feature of each part of speech can be used for recommendation information provided to the user.

  Specifically, the morphological element analysis is performed on the entire sentence of the electronic book, and the sentences constituting the electronic book are classified into parts of speech and phrases. Then, for example, the ratio of hiragana, katakana and alphanumeric characters included in the noun phrase is obtained and held.

  By generating recommendation information to be provided to the user using this result, it is possible to provide information that matches the user's preference. For example, when an electronic book is a novel set in overseas, a relatively large number of noun phrases including katakana and alphanumeric characters are often used. Up to now, it is considered that users who have purchased many e-books that use a relatively large number of katakana and noun phrases including alphanumeric characters prefer to purchase novels set in overseas. The information providing server device 30 can provide such users with information on novels set in overseas by using the characteristics of each part of speech.

[4.6] Modification 6
In the above embodiment, each DB is provided in the information providing server device 30, and these DBs are managed and controlled only by the information providing server device 30. However, each DB management and control entity is separately provided. A simple computer system may be used.

DESCRIPTION OF SYMBOLS 1 ... Communication system 10 ... Communication terminal device 30 ... Information provision server device 310 ... Communication control part 320 ... ROM / RAM
330 ... Recording device 331 ... User management DB
332 ... Content information DB
340 ... Server management control unit 350 ... Data processing unit 351 ... DB management unit 352 ... Content information acquisition unit 353 ... Content information extraction unit 354 ... Similarity calculation unit 355 ... Recommended information generation unit 356 ... Information provision unit

Claims (12)

A server device connected to a communication terminal device via a network and providing content information related to content including text,
Database management means for managing a content information database in which the content information is registered and a user management database in which user information about the user is registered;
Content information acquisition means for acquiring content information including a sentence stroke feature amount obtained by quantifying the character arrangement characteristics in the content sentence;
Content information extraction means for analyzing the content and extracting content information including a sentence stroke feature amount of the content;
By comparing the sentence stroke feature amount acquired by the content information acquisition unit with the sentence stroke feature amount extracted from each content registered in the content information database by the content information extraction unit, similarity is obtained. Similarity calculating means for calculating;
Recommendation information generating means for generating recommendation information based on the calculated similarity;
An information providing means for providing the generated recommendation information to the user;
A server device comprising: The server device according to claim 1,
The database management means registers content information including an average sentence writing feature amount calculated based on a sentence writing feature amount of the content purchased by the user in the user management database for each user,
The server apparatus in which the content information acquisition unit acquires the average sentence stroke feature amount of a predetermined user as the sentence stroke feature amount from the user information database. In the server apparatus according to claim 1 or 2,
The server device, wherein the sentence writing feature amount is a ratio of kana to the entire sentence in the content. In the server apparatus in any one of Claims 1-3,
The server device, wherein the sentence writing feature amount is an average value of the number of characters included in each paragraph constituting the sentence in the content. In the server apparatus in any one of Claims 1-4,
The server device, wherein the sentence writing feature amount is a ratio of a conversation sentence represented by brackets to a sentence in the content. In the server apparatus in any one of Claims 1-5,
The database management means registers the content information including the category to which the content belongs in the content information database for each content, and the content information including category information indicating the category of the content purchased by the user. Register in the user information database,
The recommendation information generation unit is a server device that generates recommendation information for each category based on the content information acquired by the content information acquisition unit. In the server apparatus in any one of Claims 1-5,
The content information registration means registers the content information further including category information indicating a category of the content purchased by the user in a user information database;
The recommendation information generation means is based on the category information included in the content information acquired by the content information acquisition means, only content having the same category information as the category information, or category information different from the category information Server device that selects only content having In the server apparatus in any one of Claims 1-7,
The database management means registers the content information further including author information for identifying an author of the content purchased by the user in a user information database;
The recommendation information generation means is based on the author information included in the content information acquired by the content information acquisition means, only the content having the same author information as the author information, or the author information Is a server device that selects only content having different author information. In the server apparatus in any one of Claims 1-8,
The database management means registers the content information further including read information transmitted from the communication terminal device and indicating that the content has been read, in a user information database,
The recommended information generation unit is a server device that selects only content having the read information included in the content information acquired by the content information acquisition unit. In the server apparatus according to any one of claims 1 to 9,
A server device, wherein the content information acquisition unit acquires content information including a sentence writing feature amount specified based on content information related to the content transmitted from the communication terminal device. A server device connected to a communication terminal device via a network and providing content information related to content including text,
Database management means for managing a content information database in which the content information is registered and a user management database in which user information about the user is registered;
Content information acquisition means for acquiring content information including a sentence stroke characteristic amount obtained by quantifying the character arrangement characteristics in the content sentence;
Content information extraction means for analyzing the content and extracting content information including a sentence stroke feature amount of the content;
By comparing the sentence stroke feature amount acquired by the content information acquisition unit with the sentence stroke feature amount extracted from each content registered in the content information database by the content information extraction unit, similarity is obtained. Similarity calculation means for calculating,
Recommendation information generating means for generating recommendation information based on the calculated similarity,
Information providing means for providing the generated recommendation information to the user;
A program characterized by functioning as An information providing method for providing content information relating to content including text from a server device to a communication terminal device connected to the server device via a network,
A database management process for managing a content information database in which the content information is registered and a user management database in which user information about the user is registered;
A content information acquisition step for acquiring content information including a sentence stroke feature value obtained by quantifying the character array characteristics in the content sentence;
A content information extraction step of analyzing the content and extracting content information including a sentence stroke feature amount of the content;
By comparing the sentence stroke feature amount acquired by the content information acquisition unit with the sentence stroke feature amount extracted from each content registered in the content information database by the content information extraction unit, similarity is obtained. A similarity calculation step to calculate,
A recommendation information generating step of generating recommendation information based on the calculated similarity;
An information providing step of providing the generated recommendation information to the user;
An information providing method comprising:

Images