JP2015197712A - Book recommendation method, book recommendation system, book recommendation server device, program, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for providing recommended book information without receiving any influence from a reader's search keyword setting skill when a purchaser receives a recommendation for an electronic book.SOLUTION: An electronic book recommendation method using a bibliographic information database for storing bibliographic information including a feature word is executed by procedures including: a marker data creation step that a browsing terminal accepts a marker that a browser has entered in an electronic book, and creates marker data including a character string of a portion where the marker is drawn, and transmits the marker data to a recommendation server device; a feature word extraction step that the recommendation server device extracts the feature word from the character string of the marker data; a bibliographic information search step that the recommendation server device searches bibliographic information by referring to the bibliographic information database by using the feature word; a recommended book information transmission step that the recommendation server device transmits the searched bibliographic information as recommended book information to the browsing terminal.

Description

The present invention relates to a book recommendation system for recommending books, a book recommendation server device, a book recommendation method, and a program. The present invention is particularly useful for recommending a new book by using a marker or comment written on the book by an electronic book reader.

In electronic book mail order (net sales) using an information network, recommendation information (recommended content) of an electronic book is provided in response to a request from a purchaser of the electronic book. The online e-book sales system accepts search terms entered by buyers, searches the bibliographic information (book title, author name, publisher name, table of contents, contents summary, price, etc.) of the e-book, The search result is returned to the purchaser as recommendation information and displayed on the purchaser terminal device.

For example, when searching for feature words contained in content metadata (bibliographic information) and providing recommended content, in order to select the recommended content appropriately, the part-of-speech of the keyword entered by the e-book purchaser Patent Document 1 describes a technique for determining feature words having high recommendation accuracy from feature word candidates included in content metadata by using part-of-speech information having a large number of appearances.

Further, Patent Literature 2 describes a technique that provides only recommended upper-level books as recommended information based on searched book information and loan records.

JP 2013-235421 A JP 2012-103759 A

However, as in the technique of Patent Document 1, it is highly likely that a purchaser of an electronic book sets a search keyword by himself because of the skill of keyword setting, and the search result is recommended. There is a high possibility of variations in accuracy.
Also, as in the technique of Patent Document 2, selecting a book with a large number of readers as recommended book information from the search results, conversely, there is a fear that desirable recommendation information cannot be provided because there are few readers. Occurs.

The present invention has been made to solve the above-described problems, and the problem of the present invention is that the purchaser is not affected by the skill of the reader's search keyword setting when receiving the recommendation of the electronic book. It is to provide a system, apparatus, method, and program for providing recommended book information.
Another object of the present invention is to provide a system, an apparatus, a method, and a program that provide recommended book information without being influenced by other readers when a purchaser receives a recommendation of a book.

This invention solves the said subject by the means as described in each following aspect.
That is, the first invention of the present invention is a book recommendation method, and the first aspect is
A book recommendation method using a bibliographic information database for storing bibliographic information,
A marker data creation step in which a browsing terminal creates marker data including a character string of a part where a viewer has pulled a marker in an electronic book, and transmits the marker data to a recommendation server device;
A feature word extraction step in which the recommendation server device extracts a feature word from the character string of the marker data;
A bibliographic information search step in which the recommendation server device searches for bibliographic information by referring to the bibliographic information database using a search query created using a feature word as a search word;
A recommended content transmitting step in which the recommended server device transmits the retrieved bibliographic information to the browsing terminal as recommended content;
This is a book recommendation method characterized by being performed in a procedure including

Since the recommendation server device extracts a feature word from a marker phrase or comment and uses it as a search word, the reader of the electronic book does not need to set the search word.
Therefore, it is possible to provide the recommended content information without being influenced by the technique for setting the search term and the skill of knowledge.

According to a second aspect of the first invention of the present application, in the first aspect,
A feature word weight calculating step in which the recommendation server device calculates the weight of the feature word;
A bibliographic information weight calculating step in which the recommendation server device calculates the weight of the bibliographic information using the weight of the feature word included in the searched bibliographic information;
A recommended content transmitting step in which the recommended server device transmits the bibliographic information selected from the searched bibliographic information to the browsing terminal as the recommended content based on the weight of the bibliographic information;
This is a book recommendation method characterized by being performed in a procedure including

The recommendation server device can weight the bibliographic information of the search result according to the weight of the feature word included in the bibliographic information.
Therefore, since the recommendation server device can select the recommended content from the search results according to the weight of the feature word included in the bibliographic information, it is possible to provide the recommended content that meets the viewer's expectation.

According to a third aspect of the first invention of the present application, in the second aspect,
The bibliographic information search step includes:
The recommended server device searches the bibliographic information by referring to the bibliographic information database using the search query created by using the feature words narrowed down based on the weights of the feature words as the search words.
This is an electronic book recommendation method.

The recommendation server device can improve the search efficiency of the bibliographic information by narrowing down the search word based on the weight of the feature word.

According to a fourth aspect of the first invention of the present application, in the second aspect or the third aspect,
A page weight calculation step in which the recommendation server device calculates the page weight based on the last viewed page information of the marker data;
Made in a procedure that includes
The marker data in the marker data creation step is
Furthermore, it is data including the last viewed page information,
The feature word weight calculation step includes:
A step of calculating weights of feature words based on page weights;
This is a book recommendation method characterized by this.

The recommendation server device can weight the bibliographic information of the search result in accordance with the weight of the feature word based on the page weight based on the last browsing page.
Therefore, the recommended server device can select the recommended content from the search result by strongly reflecting the influence of the feature word existing in the final browsing page.

According to a fifth aspect of the first invention of the present application, in the second aspect or the third aspect,
The recommendation server device refers to the marker entry date / time information of the marker data, identifies the latest marker entry date / time information, and calculates the marker weight of the marker data based on the latest marker entry date / time information. Marker weight calculation step,
Made in a procedure that includes
The marker data in the marker data creation step is
Furthermore, it is data including date and time information that entered the marker,
The feature word weight calculation step includes:
Calculating weights of feature words based on marker weights;
It is the book recommendation method characterized by being.

The recommendation server device can weight the bibliographic information of the search result according to the weight of the feature word based on the marker weight based on the latest marker.
Therefore, the recommendation server device can select the recommended content from the search result by strongly reflecting the influence of the feature word existing in the latest marker.

According to a sixth aspect of the first invention of the present application, in the fourth aspect,
The recommendation server device refers to the marker entry date / time information of the marker data, identifies the latest marker entry date / time information, and calculates the marker weight of the marker data based on the latest marker entry date / time information. Marker weight calculation step,
Made in a procedure that includes
The marker data in the marker data creation step is
Furthermore, it is data including date and time information that entered the marker,
The feature word weight calculation step includes:
Calculating the weight of the feature word based on the page weight and the marker weight,
The recommended book information transmission step is a step in which the recommendation server device transmits bibliographic information extracted based on the weight of the bibliographic information to the browsing terminal as recommended content.
This is a book recommendation method characterized by this.

The recommendation server device can weight the bibliographic information of search results according to the weight of the feature word based on the page weight based on the last viewed page and the marker weight based on the latest marker. is there.
Therefore, the recommendation server device can select the recommended content from the search result by strongly reflecting the influence of the feature word existing in the last browsing page and the feature word existing in the latest marker.

According to a seventh aspect of the first invention of the present application, in the first to sixth aspects of the first invention,
A bibliographic information search step for ranking the searched bibliographic information;
A bibliographic information weight calculation step for calculating bibliographic information weight using the ranking information;
Made in a procedure that includes
The recommended content transmission step includes:
The recommendation server device transmits the bibliographic information selected from the searched bibliographic information based on the weight of the bibliographic information to the browsing terminal as recommended content.
This is a book recommendation method characterized by this.

It is possible to select the recommended content from the bibliographic information of the search results by strongly reflecting the ranking of the bibliographic information.

The first aspect of the second invention of the present invention is:
A book recommendation system configured to be connectable to a browsing terminal equipped with a viewer for browsing an electronic book and a recommendation server device that provides recommendation information of the electronic book,
The viewer of the viewing terminal is
A function to create marker data and send it to the recommendation server device,
A function to receive and display the recommended content;
A viewer comprising
The recommendation server device is:
Storage means for storing a bibliographic database for storing bibliographic information of electronic books;
Marker data receiving means for receiving the marker data and storing it in the marker data storage area;
Word extraction means for extracting a word from a character string of marker data and making it a search word;
A recommended content search means for creating a search query using a search word, searching a bibliographic information database, and acquiring bibliographic information;
Recommended content transmitting means for transmitting book information as recommended content to the browsing terminal device;
A server device comprising:
This is a book recommendation system.

Since the recommendation server device extracts a feature word from a marker phrase or comment and uses it as a search word, the reader of the electronic book does not need to set the search word.
Therefore, it is possible for the recommendation server device to provide the recommended content information without being influenced by the technique for setting the search term and the skill of knowledge.

According to a second aspect of the second invention of the present application, in the first aspect,
The recommendation server device is:
A function for calculating the page weight by identifying the last viewed page with reference to the page number of the marker data of the last viewed page and the storage date and time;
A function for calculating the marker weight including the page number and the marker specifying information by referring to the storage date / time of the marker data of the marker or comment, specifying the storage date / time of the latest marker or comment,
A function of calculating a word weight using a marker word including a page number and marker identification information, a page weight, and a marker weight including a page number and marker identification information;
A word weight calculating means comprising:
Search word selection means for selecting a search word to be used for the search using the calculated word weight;
A function to calculate the ranking of the search results for each retrieved word,
A function for calculating the weight of the search result for each search word using the rank of the search result for each search word and the weight of the search word;
A function for calculating the weight of the book information using the weight of the search result for each search word,
Bibliographic information weight calculation means comprising:
Recommended content determination means for selecting book information based on the weight of the book information;
A server device comprising:
This is a book recommendation system.

The recommendation server device can weight the bibliographic information of search results according to the weight of the feature word based on the page weight based on the last viewed page and the marker weight based on the latest marker. is there.
Therefore, the recommendation server device can select the recommended content from the search result by strongly reflecting the influence of the feature word existing in the last browsing page and the feature word existing in the latest marker.
In addition, it is possible to select the recommended content from the bibliographic information of the search result by strongly reflecting the ranking of the bibliographic information.

The first aspect of the third invention of the present invention is:
Storage means for storing a bibliographic database for storing bibliographic information of electronic books;
Marker data receiving means for receiving the marker data and storing it in the marker data storage area;
Word extraction means for extracting a word from a character string of marker data and making it a search word;
A recommended content search means for creating a search query using a search word, searching a bibliographic information database, and acquiring bibliographic information;
Recommended content transmitting means for transmitting book information as recommended content to the browsing terminal device;
A recommendation server device characterized by comprising:

Since the recommendation server device extracts a feature word from a marker phrase or comment and uses it as a search word, the reader of the electronic book does not need to set the search word.
Therefore, it is possible for the recommendation server device to provide the recommended content information without being influenced by the technique for setting the search term and the skill of knowledge.

According to a second aspect of the third invention of the present application, in the first aspect,
A function for calculating the page weight by identifying the last viewed page with reference to the page number of the marker data of the last viewed page and the storage date and time;
A function for calculating the marker weight including the page number and the marker specifying information by referring to the storage date / time of the marker data of the marker or comment, specifying the storage date / time of the latest marker or comment,
A function of calculating a word weight using a marker word including a page number and marker identification information, a page weight, and a marker weight including a page number and marker identification information;
A word weight calculating means comprising:
Search word selection means for selecting a search word to be used for the search using the calculated word weight;
A function to calculate the ranking of the search results for each retrieved word,
A function for calculating the weight of the search result for each search word using the rank of the search result for each search word and the weight of the search word;
A function for calculating the weight of the book information using the weight of the search result for each search word,
Bibliographic information weight calculation means comprising:
Recommended content determination means for selecting book information based on the weight of the book information;
This is a recommended server device.

The recommendation server device can weight the bibliographic information of search results according to the weight of the feature word based on the page weight based on the last viewed page and the marker weight based on the latest marker. is there.
Therefore, the recommendation server device can select the recommended content from the search result by strongly reflecting the influence of the feature word existing in the last browsing page and the feature word existing in the latest marker.
In addition, it is possible to select the recommended content from the bibliographic information of the search result by strongly reflecting the ranking of the bibliographic information.

The fourth invention of the present invention is:
A computer program that operates as the recommendation server device according to the third invention.

According to the present invention,
(1) It is possible to recommend content information based on marker and comment information drawn by a reader of an electronic book.
(2) The reader of the electronic book can obtain the recommended content without setting a keyword for searching for the recommended report.

It is a figure explaining the whole structure of the book recommendation system. It is a figure explaining the rough procedure of book recommendation. It is explanatory drawing of an electronic book browsing screen. It is explanatory drawing of the phrase 194 which attached the marker. It is an example of associating a marker with a comment. It is explanatory drawing of marker data. It is explanatory drawing of a page weight. It is explanatory drawing of a marker weight. It is explanatory drawing of the weight of the word cut out from the marker / comment. It is explanatory drawing of the weight of a marker word. It is explanatory drawing of the weight of a word. It is explanatory drawing of a bibliographic information database. It is explanatory drawing of a search word and a search result. It is explanatory drawing of weight calculation of the book title of a search result. It is explanatory drawing of the weight of a book title. 3 is a detailed configuration diagram of a recommendation server device 300. FIG. 2 is a detailed configuration diagram of a browsing terminal device 100. FIG.

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

FIG. 1 is an example of a diagram illustrating the overall configuration of a book recommendation system 1 according to an embodiment of the present invention.
The book recommendation system 1 is configured such that a browsing terminal 100 (smartphone) for browsing an electronic book and a recommendation server device 300 that provides recommendation information for the book are connectable.
The browsing terminal 100 is wirelessly connected to the base station 700 and is connected to the communication network 900 via the gateway 800.
The communication network 900 is the Internet.
Note that the connection method of the browsing terminal with the communication network may be a wired connection.

The browsing terminal 100 is an existing smart phone. Alternatively, it may be a tablet terminal, a PC (personal computer), or an electronic book dedicated terminal.

The recommended server device 300 is an existing server device in which a dedicated program described later is installed.

FIG. 2 is a diagram for explaining a rough operation procedure and processing flow of the book recommendation system 1.

(1) The viewer draws a marker on the electronic book displayed on the browsing terminal 100.
At this time, the viewer can associate the input comment with the marker (details will be described later).

(2) The browsing terminal 100 creates marker data including a phrase (character string) of a part where a marker is drawn or a comment (character string) associated with the marker, and transmits the marker data to the recommendation server device 300 (for details) Will be described later).

(3) The recommendation server device 300 extracts feature words (words) from the words or comments of the marker data. Next, the weight of the extracted word (details will be described later) is calculated.

(4) The recommendation server device 300 selects a search word using a word weight as a clue, searches the bibliographic information database, and acquires a search result (details will be described later).

(5) The recommendation server device 300 calculates the weight of the search result based on the weight of the search word (details will be described later).

(6) The recommendation server device 300 selects the search result using the weight of the search result as a clue, and determines the recommended content (for example, the top five weights).

(7) The recommendation server device 300 transmits the determined recommended content information to the browsing terminal 100. The browsing terminal 100 displays recommended content information.
The viewer browses the recommended content information displayed on the browsing terminal 100.

Details of the book recommendation system 1 will be described below.

FIG. 3 is an explanatory diagram of an electronic book browsing screen.
FIG. 3A shows an example of a book viewer screen that displays an electronic book page. On the screen, a plurality of rows 182 of the page and a page number 183 are illustrated.
FIG. 3B is an example of the marker 181 attached to the electronic book. The screen illustrates a marker 181 drawn on a part of a row 182 of the book page.

FIG. 4 is an explanatory diagram of the word / phrase 185 with the marker 181 attached.
The expanded part “words are grammatically correct” in the line 182 and the phrase 185 “grammatically correct” of the part where the marker 181 is subtracted are illustrated.
This word / phrase (word / word of the electronic book with a marker) 185 is a character string including a word.

FIG. 5 is a diagram for explaining in detail a screen for attaching a comment to the marker 181.
FIG. 5A shows an example in which a viewer selects a marker 181a to which a comment is to be written from two markers 181 displayed on the book viewer screen.
In FIG. 5B, when one of the two markers 181a is selected, an operation button list 186 is displayed on the book viewer screen, from which the viewer selects a comment input button. It is an example of the screen of the state which carried out.
FIG. 5C shows an example of a comment input screen.
The comment input screen which displayed the comment 187 "check required!" Which the viewer input is illustrated. When the input operation of the comment is completed, the viewer creates a marker data associated with the marker by pressing the “save button” displayed, and is transmitted to the recommendation server device 300. The
FIG. 5D is a display example of the comment mark 184. A comment mark 184a indicating that a comment is associated is displayed below the marker 181a associated with the comment.

FIG. 6 is an explanatory diagram of the marker data 191.

FIG. 6A is a diagram for explaining the format of the marker data 191.
The marker data 191 is
User ID 1911, device ID 1912, product ID 1913, page number 1914, marker position 1915, marker character string 1916, marker color 1917, comment character string 1918, comment character color 1919, storage date / time (registration date / time) ) 19110, storage date / time (update date / time) 19111, and total page number 1912.
The user ID 1911 is information for identifying a viewer.
The device ID 1912 is information for identifying the browsing terminal 100.
The product ID 1913 is information for identifying an electronic book.
The page number 1914 is a numeric string. This is the last browsing page, or the page of an electronic book in which a marker or comment is registered.
The marker position 1915 is a numeric string. The first position and the last position of the character with the marker are represented by the number of characters (the character position is the number of characters when the first character on the first line in the page is 1).
The marker character string 1916 is a character string with a marker.
The marker color 1917 is marker color information.
The comment character string 1918 is a character string.
The comment character color 1919 is color information of the comment character string.
The storage date (registration date) 19110 is date information. This is the date / time information when the marker or comment is registered.
The storage date (update date) 19111 is date information. This is the date / time information when the marker or comment is updated.
The total page number 1912 is a numeric string.

FIG. 6B is an example of the marker data 191 representing the last viewed page.
The final browsing page is a page when browsing of an electronic book is interrupted, or a page being browsed when browsing is stopped halfway.
In the marker data 191 representing the last viewed page, the page number 1914 describes the number of the last viewed pages, and the marker character string 1916 and the comment character string 1918 are (empty).
For example, when the last viewed page is “20 pages”, the marker data 191 includes a user ID 1911 “0001”, a device ID 1912 “ABC1”, a product ID 1913 “B000-0001”, and a page number 1914 “20”. , The marker position 1915 is “(empty)”, the marker character string 1916 is “(empty)”, the marker color 1917 is “(empty)”, the comment character string 1918 is “(empty)”, and the comment character color 1919 is “( Empty) ”, the registration date 19110 is“ 2013/11/14 19:50:02 ”, the update date 19111 is“ (empty) ”, and the total number of pages 1912 is“ 100 ”.

Since it is normal to interrupt several times before reading one electronic book, marker data 191 representing a plurality of last viewed pages is included in the marker data 191 of one electronic book. Exists. Therefore, in order to know the latest browsing interruption page, the registration date 19110 or the update date 19111 is referred to and compared, and the information of the marker data 191 of the latest date is used.

FIG. 6C shows an example of marker data 191 for registering / updating a marker.
In the marker data for registering or updating the marker word / phrase, the word / phrase with the marker is described in the marker character string 1916 and the comment character string 1918 is (empty).
For example, the marker data 191 in the case where a blue marker is drawn on the phrase “not a civilization citizen, not A person” on page 20 of an electronic book has a user ID 1911 of “0001”, a device ID 1912 is “ABC1”, Product ID 1913 is “B000-0001”, Page number 1914 is “20”, Marker position 1915 is “641-682”, Marker string 1916 is “People of civilization, people of A ”, The marker color 1917 is“ blue ”, the comment character string 1918 is“ (empty) ”, the comment character color 1919 is“ (empty) ”, and the registration date 19110 is“ 2013/11/14 19:44:20 The update date and time 19111 is “2013/11/14 19:44:20”.
Here, since the date information of the registration date 19110 and the date information of the update date 19111 are the same, it can be seen that the marker (word) has been registered.

FIG. 6D is an example of marker data 191 for registering / updating a marker with a comment.
In the marker data of a marker to which a comment is attached, the word / phrase to which the marker is drawn is described in the marker character string 1916, and the word / phrase inputted as a comment is described in the comment character string 1918.
For example, the marker data 191 in the case where the word “check required!” In black letters is input to the phrase “requirement for such a business is the company rule” on page 20 of the electronic book, the user ID 1911 is “0001”, the device data ID 1912 is “ABC1”, Product ID 1913 is “B000-0001”, Page number 1914 is “20”, Marker position 1915 is “1-44”, and Marker string 1916 is “A commitment to such a business is a company rule” The marker color 1917 is “blue”, the comment character string 1918 is “check required!”, The comment character color 1919 is “black”, the registration date 19110 is “2013/11/9 20:25:22”, and the update date 19111 is “2013/11/10 12:11:01”.
Here, since the date / time information of the update date / time 19111 is newer than the date / time information of the registration date / time 19110, it can be seen that the word / phrase of the comment, the word / phrase of the marker, or both are updated.

7 to 10 are diagrams for explaining a detailed procedure for calculating the weight of a word.
The word weight is a value calculated based on the weight of the page on which the marker is described and the weight of the marker from which the word is cut out.

FIG. 7 is an explanatory diagram of page weights.
The page weight is a numerical value set for each page of the electronic book using the last viewed page as a reference page. The page weight is data having a larger value as it is closer to the last viewed page (reference page) and a smaller value as the distance from the reference page is increased.
In addition, the reason why the page weight is larger as it is closer to the last viewed page is that the last viewed page and the marker (or a comment related to this marker) written on the nearby page are the most interesting for the viewer. Based on the assumption that there should be.

FIG. 7A shows an example of the last viewed page and a nearby page.
An electronic book whose last browsed page number is 20 pages in an electronic book having a total number of 100 pages is illustrated.
The marker data 191 of the last browsing page at this time is illustrated in FIG. 6B.
Expression 1 is an expression for calculating the page weight.

ここで、総ページ数とは、最終閲覧を表すマーカー・データの総ページ数１９１１２の値である。
最終閲覧ページ番号とは、最終閲覧を表すマーカー・データのページ番号１９１４の値である。なお、最終閲覧ページ番号は、最終閲覧ページを表すマーカー・データの保存日時欄１９１１０／１９１１１を参照・比較して、最新日時のマーカー・データを選択して、このマーカー・データの閲覧ページ番号を最終閲覧ページ番号とする。

Here, the total number of pages is the value of the total number of pages 19112 of the marker data representing the final browsing.
The last browsing page number is the value of the page number 1914 of the marker data representing the last browsing. The last browsed page number refers to and compares the marker data storage date / time column 19110/19111 representing the last browsed page, selects the latest date / time marker data, and sets the browse page number of this marker data. The last viewed page number.

FIG. 7B is an example of the calculated page weight.
When the last browsing page is 20 pages (example in FIG. 7B), the page weight of each page is as follows. For example, the page weight of the last browsing page number “100” is 100, the page weight of the page number “21” is 99, and the page weight of the page number “100” is 20.

FIG. 8 is an explanatory diagram of marker weights.
The marker weight is a numerical value calculated by using the storage date and time of the latest marker stored most recently as the reference date and time. The marker weight takes a smaller value as it goes back to the old date from the latest date (reference date).
The latest date / time is the latest date / time among the save date / time (registration date / time) 19110 or the save date / time (update date / time) 19111 of the marker data 191.
The reason why the marker weight is larger for the latest date and time and the marker weight is smaller for the older date and time is that the newly saved marker (or a comment related to this marker) should be more interesting to the viewer. based on.

FIG. 8A shows an example of the latest marker.
An electronic book in which the latest marker 1 is written on the electronic book on page 20 is illustrated.
The marker data 191 of the latest marker 1 at this time is illustrated in FIG.
In addition, the marker 2 and the comment 2 (the marker and comment marker data 191 are exemplified in FIG. 6D) are also entered on the 20th page.
Expression 2 is an expression for calculating the marker weight.

最新日時は、マーカーを表すマーカー・データ１９１、あるいは、マーカー／コメントを表すマーカー・データ１９１の保存日時１９１１０／１９１１１を参照・比較して、最新日時のマーカー・データを選択して、このマーカー・データのマーカーを最新マーカーとする。

For the latest date and time, the marker data 191 representing the marker or the saved date and time 19110/19111 of the marker data 191 representing the marker / comment is referred to and compared, and the marker data of the latest date and time is selected. The data marker is the latest marker.

FIG. 8B is an example of the calculated marker weight.
The marker weights of the markers illustrated in FIG. 8B are as follows. For example, the marker specifying information of marker 1 is 20-1, and the marker weight is 100.
The marker specifying information of the marker 2 is 20-641, and the marker weight is 38.36.
Here, the marker specifying information is information for specifying a marker, for example, information in which a page number 1914 and a marker position 1915 are combined.

FIG. 9 is an explanatory diagram of marker words cut out from markers / comments.
The marker word 188 is a character string cut out from the word / phrase 185 of the electronic book with the marker or the word / phrase 187 of the comment.
Marker words 188 cut out from marker 1 are marker word 1, marker word 2, and marker word 3.
The marker words 188 cut out from the marker 2 are a marker word 4, a marker word 5, and a marker word 6.
Marker words extracted from the comment 2 of the marker 2 are a marker word 7 and a marker word 8.
Here, in order to cut out the marker word 188 from the word / phrase 185 of the electronic book with the marker or the word / phrase 187 of the comment, morphological analysis which is a known technique is used.
Morphological analysis divides words into morpheme strings (character strings), collates with the part-of-speech dictionary, examines each part-of-speech, creates a part-of-speech string, and then refers to the grammar rules as a grammar. This is an analysis method in which words in a sequence constituting a correct part-of-speech sequence are correct words.
Note that unnecessary words (for example, particles of particles) are removed from the analysis result words.

This marker word is a search word used for searching for recommended content.
Since the marker word is a word cut out from an electronic book word or comment word, it is possible to set a search word without depending on knowledge or technology for setting the search word.

FIG. 10 is an example of the weight of the marker word cut out from the marker / comment.
This is an example of calculating the weight of the extracted marker word using the marker weight and the page weight.
For example, the weight of the marker word 1 extracted from the marker 1 described on the page 20 is a value obtained by adding the marker weight “100” of the marker 1 and the page weight “100” of the page 20 including the marker 1 “ 200 ".
Further, the word weight of the marker word 4 cut out from the marker 2 described on the page 20 is a value obtained by adding the marker weight “38.36” of the marker 1 and the page weight “100” of the page 20 including the marker 1 “ 138.36 ".

Here, marker word 1, marker word 2,. . . Since the marker word 8 is a word cut out from the marker / comment, the same word usually exists among the marker words.

FIG. 11 is an explanatory diagram of word weights.
Marker word 1, marker word 2,. . . Since the same word exists in the marker word 8, the weights of the same words are aggregated and aggregated from the weights of the extracted marker words to obtain the word weights.
For example, since marker word 1 of marker 1, marker word 5 of marker 2 and marker word 7 of comment 2 are the same word A, the weights “200.0” and “138.36” of the marker words are the same. “476.72” obtained by adding “138.36” is exemplified as the weight of the word A.
Similarly, the weights of word B, word C, and other words are illustrated.
In the example, only marker words on the same page are tabulated. However, the weight of the marker word extracted from the marker / comment on another page may be added to calculate the word weight. At this time, the range of pages to be added may be selected as necessary.

FIG. 12 is an explanatory diagram of a bibliographic information database.
FIG. 12A is a diagram for explaining the format of the bibliographic information database 395.
The bibliographic information database includes a product ID 3951, a book title 3952, a book title Kana 3953, a subtitle 3954, a subtitle Kana 3955, a search keyword 3956, an author name 3957, an author name Kana 3958, an ISBN 3959, and a product remarks. It is composed of bibliographic items such as 39510, table of contents 39511, explanation sentence 39512, label 39513, category 39514, publisher 39515, mark 39516, genre 39517, and release date 39518.
The product ID 3951 is a unique number string.
The book title 3952 is a character string.
The book title Kana 3953 is a Kana character string.
The subtitle 3954 is a character string.
The subtitle Kana 3955 is a Kana character string.
The search keyword 3956 is a character string.
The author name 3957 is a character string.
The author name Kana 3958 is a Kana character string.
ISBN3959 is a numeric string.
The product remark 39510 is a character string.
The table of contents 39511 is a character string.
The explanation sentence 39512 is a character string.
The label 39513 is a character string.
Category 39514 is a character string.
The publisher 39515 is a character string.
The mark 39516 is (please fill in by the inventor).
The genre 39517 is a character string.
The release date 39518 is a numeric string representing a date.

FIG. 12B is an example of the bibliographic information database 395.
Product ID 3951 “4063761827”, book title 3952 “A Spirit”,
Book title Kana 3953 “Aspire”, Subtitle 3954 “This is the beginning of the end of the world”, Subtitle Kana 3955 “Kokohajin Seino Owarino Hajimari”, Search keyword 3956 “Growth strategy”, Author 3957 “S field T Man (Author), A Naga Iko (Original) ”, Author Kana 3958“ Estatio Eiga Aiko ”, ISBN 3959“ 4063761827 ”, Product Remarks 39510“ Latest Volume 2 ”, Table of Contents 39511“
Chapter 1 Sales of Civilizations Chapter 2 Road to A Chapter 3 Progress to Expansion Sales, Commentary 39512 “How Japan's Largest Large Company A has Grown. Is here! ”, Label 39513“ DC Deluxe ”, category 39514“ Comic ”, publisher 39515“ Company D ”,
The mark 39516 “”, the genre 39517 “”, and the release date 39518 “2011.03.31” are illustrated.

The bibliographic information database to be searched may be content data such as paper books, music, videos, and games in addition to electronic books.

FIG. 13 is a diagram illustrating search words and search results.
FIG. 13A is an example of a search word for searching for a book title.
The search word is an example in which, based on the calculated word weight, for example, the words having the highest word weight (words A, B,..., E) are used as the search word.

Since the search word is selected based on the weight of the word, a highly important word is selected.
Therefore, the search result is a result obtained without being affected by the search word setting knowledge and skill of the technique by the viewer.

FIG. 13B is an example of a search result for each search word.
The book title obtained by creating a search query based on one search word and searching for bibliographic items including the book title 3952 of the bibliographic information stored in the bibliographic information database is illustrated.
For example, the search result of the search word “word A” has 5 titles, the book title “title 1” in the rank “1”, the book title “title 2” in the rank “2”, and the rank “3”. The book title “title 3”, the title “title 4” in the rank “4”, and the book title “title 5” in the rank “5” are illustrated.
Here, the number of titles is the number of cases acquired as a search result (a predetermined number of cases).
The ranking is a ranking in which the title of the most recent sale date is ranked higher based on the sale date 39518 of the bibliographic information.
The title order may be set based on the order of the number of copies in combination with the sales information.

14 to 15 are diagrams for explaining a detailed procedure for calculating the weight of the book title.
The weight of the book title is the sum of the weights of the search results of the same book title based on the weights of the search results searched with the search word. Hereinafter, the weight of the search result and the weight of the title will be described.

FIG. 14 is a diagram for explaining the weight of the search result.
Here, the search result weight calculation formula for calculating the weight of the search result is as follows.

順位は、検索結果の順位の値である。
検索タイトル件数は、検索結果として取得する書籍タイトルの件数である。

The rank is a rank value of the search result.
The number of search titles is the number of book titles acquired as a search result.

FIG. 14 illustrates search result weights.
For example, the search result searched for the search word A is the title rank “1”, the book title “title 1”, the search result weight “2383.6”, the title rank “2”, the book title “title 2”, the search Result weight “1906.88”, title rank “3”, book title “title 3”, search result weight “1430.16”, title rank “4”, book title “title 4”, search result weight “953” .44 ”, title ranking“ 5 ”, book title“ title 5 ”, search result weight“ 476.72 ”, and the like.

FIG. 15 is a diagram for explaining the weight of a book title.
The book title weight is calculated by adding the search title weights for each book title.
For example, the book title weight “1906.88” of the book title “title 1” is illustrated.
The weight for each search word of the exemplified book title is the search result weight “2383.6” searched for the search word A, the search result weight “2383.6” searched for the search word B, and the search The search result weight “2383.6” searched for the word C and the total weight of the search results searched for the search word E. Note that the search result searched with the search word A does not include the book title “title 1”.

FIG. 16 is a detailed configuration diagram of the recommendation server device 300.
The recommendation server device 300 includes a CPU 301, a network communication unit 304, a storage unit 309, a dedicated program, and the like.
The BUS 399 connects the CPU 301, the network communication unit 304, the storage unit 309, and the like.

The CPU 301 is a central processing unit.
The network communication unit 304 is a connector (for example, an IEEE 802.3 standard connector).

The storage unit 309 is a semiconductor memory or a magnetic memory.
The storage unit 309 includes a database storage area 3091 and a marker / data storage area 3093, and stores an operating system 391, an HTTP server 392, and a dedicated program.
The database storage area 3091 stores a bibliographic information database 395 (details have been described with reference to FIG. 12).
The marker data storage area 3093 stores marker data 191.

The operating system 391 manages and controls hardware (for example, the CPU 301, the network communication unit 304, the storage unit 309, and the BUS 399) of the recommendation server device 300, and controls application software (for example, a dedicated program). Basic software that provides services that allow these hardware to be used.

The HTTP server 392 is a service program that transmits / receives description data in a markup description language to / from client software (WEB browser) via HTTP communication.

In addition, each means (means described below) realized by each dedicated program is provided.
Each means functions by interpreting and executing each dedicated program by the CPU 301.

The marker data receiving unit 310 is a program module that executes a function of receiving the marker data 191 transmitted from the browsing terminal device 100 and storing it in the marker data storage area 3093.

The word unit 320 has a function of extracting a marker word including a page number and marker specifying information from a marker character string or a comment character string of the marker data 191.
Is a program module that executes (Details are described in FIG. 9)

The word weight calculation means 330
A function of identifying the last viewed page and calculating the page weight with reference to the page number 1914 of the marker data 191 of the last viewed page and the storage date 19110/19111 (details are described in FIGS. 6 and 7) did)
A function for calculating the marker weight including the page number and the marker specifying information by specifying the storage date / time of the latest marker or comment with reference to the storage date / time 19110/19111 of the marker data 191 of the marker or comment; Was explained in FIG. 6 and FIG.
A function of calculating a word weight using a marker word including a page number and marker specifying information, a page weight, and a marker weight including a page number and marker specifying information (details are described in FIG. 10)
Is a program module that executes

The search word selection storage means 340
A function for selecting a search word to be used for search using the calculated word weight (details are described in FIG. 13A)
Is a program module that executes

The recommended content search means 350
A database search function that creates a search query using a search word, searches the bibliographic information database 395, and obtains a search result for each search word; and the details are described in the section (b) of FIG. )
Is a program module that executes

The book title weight calculation means 360
A function for calculating the rank of the obtained search results for each search word (details are described in FIG. 13B)
A function of calculating the weight of the search result content for each search word by using the ranking of the search result for each search word and the weight of the search word (details are described in FIG. 14)
A function for calculating the sum of the weights of the same content using the weights of the search results for each search word (details are described in FIG. 15)
Is a program module that executes

The recommended content determination means 370
A function of selecting recommended content based on the weight of content (details are described in FIG. 2)
Is a program module that executes

The recommended content transmission means 380
A function for the browsing terminal device 100 to transmit recommended content data 398 (part of bibliographic information) including the book title of the selected recommended content;
Is a program module that executes

FIG. 17 is a detailed configuration diagram of the browsing terminal device 100.
The browsing terminal device 100 includes a CPU 101, a display unit 102, an input unit 103, a SIM connector 105, a camera 106, a speaker 107, a microphone 108, a wireless communication circuit 110, a power supply unit 111, and a nonvolatile memory. 121, volatile memory 122, BUS199, and the like.
The BUS 199 includes a CPU 101, a display unit 102, an input unit 103, a SIM connector 105, a camera 106, a speaker 107, a microphone 108, a wireless communication circuit 110, a nonvolatile memory 121, a volatile memory 122, and the like. Connect.

The CPU 101 is a central processing unit.
The display unit 102 is a liquid crystal display device or an organic EL display device.
The input unit 103 is a touch panel.
The wireless communication circuit 110 performs wireless communication with the base station.
The nonvolatile memory 121 is a semiconductor memory.
The volatile memory 122 is a semiconductor memory.
The power supply unit 111 supplies power to the CPU 101, the display unit 102, the input unit 103, the wireless communication circuit 110, the nonvolatile memory 121, the volatile memory 122, the BUS 199 and others.

The nonvolatile memory 121 includes an electronic book storage area 1211 and stores an operating system 195.
The electronic book storage area stores electronic book data.
The volatile memory 122 includes a recommended content storage area 1221.
The recommended content storage area 1221 stores recommended content data 398 including a book title.

The operating system 195 includes hardware of the browsing terminal device 100 (for example, the CPU 101, the input unit 103, the display unit 102, the wireless communication circuit 110, the nonvolatile memory 121, the volatile memory 122, the BUS 199, and the like). Is the basic software that provides services that allow these hardware to be used for application software (for example, the book viewer 197).

The book viewer 197 is an application in which book browsing plug-in software is incorporated into an existing browser that transmits and receives description data in a markup description language to and from the HTTP server 392 through HTTP communication.

The book viewer 197
A function of creating marker data 191 (details are described in FIGS. 3 to 6)
A function of transmitting the marker data 191 to the recommendation server device 300 (details are described in FIGS. 3 to 6)
A function of receiving recommended content data 398 and displaying a book title;
Is provided.

As described above, the search for the recommended content using the word weight obtained by combining the page weight and the marker weight has been described.
Here, the word weight may be calculated using the page weight alone.
Alternatively, the weight of the word may be calculated using the marker weight alone.
Alternatively, the recommended content may be searched using all the words without calculating the word weight.
Alternatively, the recommended content may be searched by calculating a word weight by combining weights including other weights (for example, a weight based on the word frequency).

In addition, it has been described that the recommended content is selected and provided using the weight of the content combining the weight of the word and the weight of the ranking of the search results.
Here, the recommended content may be selected and provided using the page weight alone.
Alternatively, the recommended content may be selected and provided by using the weight of the ranking of the search results alone.
Alternatively, all search results may be provided as recommended content without calculating the content weight.
Alternatively, weights including other weights (for example, weights based on word frequencies) may be combined to calculate content weights and select and provide recommended content.

100 browsing terminal device 191 marker data 197 book viewer 300 recommendation server device 310 marker data receiving means 320 word means 330 word weight calculation means 340 search word selection storage means 350 recommended content search means 360 book title weight calculation means 370 recommended content determination means 380 Recommended content transmission means 395 Bibliographic information database

Claims (12)

A book recommendation method using a bibliographic information database for storing bibliographic information,
A marker data creation step in which a browsing terminal creates marker data including a character string of a part where a viewer has pulled a marker in an electronic book, and transmits the marker data to a recommendation server device;
A feature word extraction step in which the recommendation server device extracts a feature word from the character string of the marker data;
A bibliographic information search step in which the recommendation server device searches for bibliographic information by referring to the bibliographic information database using a search query created using a feature word as a search word;
A recommended content transmitting step in which the recommended server device transmits the retrieved bibliographic information to the browsing terminal as recommended content;
A method for recommending a book, characterized in that it is performed by a procedure including: The book recommendation method according to claim 1, further comprising:
A feature word weight calculating step in which the recommendation server device calculates the weight of the feature word;
A bibliographic information weight calculating step in which the recommendation server device calculates the weight of the bibliographic information using the weight of the feature word included in the searched bibliographic information;
A recommended content transmitting step in which the recommended server device transmits the bibliographic information selected from the searched bibliographic information to the browsing terminal as the recommended content based on the weight of the bibliographic information;
A method for recommending a book, characterized in that it is performed by a procedure including: The book recommendation method according to claim 2, further comprising:
The bibliographic information search step includes:
The recommended server device searches the bibliographic information by referring to the bibliographic information database using the search query created by using the feature words narrowed down based on the weights of the feature words as the search words.
The book recommendation method characterized by this. The book recommendation method according to claim 2 or 3, further comprising:
A page weight calculation step in which the recommendation server device calculates the page weight based on the last viewed page information of the marker data;
Made in a procedure that includes
The marker data in the marker data creation step is
Furthermore, it is data including the last viewed page information,
The feature word weight calculation step includes:
A step of calculating weights of feature words based on page weights;
The book recommendation method characterized by this. The book recommendation method according to claim 2 or 3, further comprising:
The recommendation server device refers to the marker entry date / time information of the marker data, identifies the latest marker entry date / time information, and calculates the marker weight of the marker data based on the latest marker entry date / time information. Marker weight calculation step,
Made in a procedure that includes
The marker data in the marker data creation step is
Furthermore, it is data including date and time information that entered the marker,
The feature word weight calculation step includes:
Calculating weights of feature words based on marker weights;
The book recommendation method characterized by this. The book recommendation method according to claim 4, further comprising:
The recommendation server device refers to the marker entry date / time information of the marker data, identifies the latest marker entry date / time information, and calculates the marker weight of the marker data based on the latest marker entry date / time information. Marker weight calculation step,
Made in a procedure that includes
The marker data in the marker data creation step is
Furthermore, it is data including date and time information that entered the marker,
The feature word weight calculation step includes:
Calculating the weight of the feature word based on the page weight and the marker weight,
The recommended book information transmission step is a step in which the recommendation server device transmits bibliographic information extracted based on the weight of the bibliographic information to the browsing terminal as recommended content.
The book recommendation method characterized by this. The book recommendation method according to claim 1, further comprising:
A bibliographic information search step for ranking the searched bibliographic information;
A bibliographic information weight calculation step for calculating bibliographic information weight using the ranking information;
Made in a procedure that includes
The recommended content transmission step includes:
A book recommendation method, wherein a recommendation server device transmits bibliographic information selected from searched bibliographic information to a browsing terminal as recommended content based on the weight of bibliographic information. A book recommendation system configured to be connectable to a browsing terminal equipped with a viewer for browsing an electronic book and a recommendation server device that provides recommendation information of the electronic book,
The viewer of the viewing terminal is
A function to create marker data and send it to the recommendation server device,
A function to receive and display the recommended content;
A viewer comprising
The recommendation server device is:
Storage means for storing a bibliographic database for storing bibliographic information of electronic books;
Marker data receiving means for receiving the marker data and storing it in the marker data storage area;
Word extraction means for extracting a word from a character string of marker data and making it a search word;
A recommended content search means for creating a search query using a search word, searching a bibliographic information database, and acquiring bibliographic information;
Recommended content transmitting means for transmitting book information as recommended content to the browsing terminal device;
A server device comprising:
A book recommendation system characterized by that. The book recommendation system according to claim 8,
The recommendation server device is:
A function for calculating the page weight by identifying the last viewed page with reference to the page number of the marker data of the last viewed page and the storage date and time;
A function for calculating the marker weight including the page number and the marker specifying information by referring to the storage date / time of the marker data of the marker or comment, specifying the storage date / time of the latest marker or comment,
A function of calculating a word weight using a marker word including a page number and marker identification information, a page weight, and a marker weight including a page number and marker identification information;
A word weight calculating means comprising:
Search word selection means for selecting a search word to be used for the search using the calculated word weight;
A function to calculate the ranking of the search results for each retrieved word,
A function for calculating the weight of the search result for each search word using the rank of the search result for each search word and the weight of the search word;
A function for calculating the weight of the book information using the weight of the search result for each search word,
Bibliographic information weight calculation means comprising:
Recommended content determination means for selecting book information based on the weight of the book information;
A book recommendation system characterized by that. Storage means for storing a bibliographic database for storing bibliographic information of electronic books;
Marker data receiving means for receiving the marker data and storing it in the marker data storage area;
Word extraction means for extracting a word from a character string of marker data and making it a search word;
A recommended content search means for creating a search query using a search word, searching a bibliographic information database, and acquiring bibliographic information;
Recommended content transmitting means for transmitting book information as recommended content to the browsing terminal device;
A recommendation server device comprising: The recommendation server device according to claim 10,
A function for calculating the page weight by identifying the last viewed page with reference to the page number of the marker data of the last viewed page and the storage date and time;
A function for calculating the marker weight including the page number and the marker specifying information by referring to the storage date / time of the marker data of the marker or comment, specifying the storage date / time of the latest marker or comment,
A function of calculating a word weight using a marker word including a page number and marker identification information, a page weight, and a marker weight including a page number and marker identification information;
A word weight calculating means comprising:
Search word selection means for selecting a search word to be used for the search using the calculated word weight;
A function to calculate the ranking of the search results for each retrieved word,
A function for calculating the weight of the search result for each search word using the rank of the search result for each search word and the weight of the search word;
A function for calculating the weight of the book information using the weight of the search result for each search word,
Bibliographic information weight calculation means comprising:
Recommended content determination means for selecting book information based on the weight of the book information;
A recommendation server device characterized by that. A computer program that operates as the recommendation server device according to any one of claims 8 to 11.

Images