TWI525571B - Library collection collaborative recommendation system - Google Patents

Description

Library collection collaborative recommendation system

The invention provides a library collection collaborative recommendation system, in particular to a library audio and video collaborative recommendation system which can reduce the search time according to the popularity ranking generated by the collaborative recommendation method.

Most of the book-reading books in the library must be entered in order to be searched, and the input of the book keywords usually has a bunch of books, and the user must confirm them one by one, so that the user will confirm the desired book. It takes a lot of time, and today there is a technology that uses collaborative recommendation, such as the Republic of China Invention Patent No. 200917762: "Collaborative Media Recommendation and Sharing Technology", which uses media media/devices. To extend the reach of media sharing scenarios that help a user discover media items of interest, such as music, recordings, playlists, and pictures, on nearby media players or devices (devices that can receive, store, and play media) Video game, the collaborative media recommendation and sharing technology determines a user's media preference based on media stored on a pair of media devices, and recommends potential shareable media based on such determined user preferences, but When using preferences to recommend, you can't effectively integrate preference values, only continuous The product cannot be characterized by human memory forgetting, resulting in insufficient system capacity resources and increasing the probability of using the system to crash. Therefore, the creative department provides a library collection collaborative recommendation system, which has a control interface. Enter the keyword from the control interface to enter the library of the video and audio search module to search the database, and the popular module allows all users to recommend the library collection, and calculate the click rate to integrate preferences and popularity, and finally The recommendation category is presented by a presentation module, so that the user can reduce the search time to improve the search efficiency and increase the system utility, and the system has a forgetting preference module and a forgotten hot module, which can be infrequently Click on the data to delete it to save system storage capacity.

The main purpose of the present invention is to provide a library collection collaborative recommendation system, in particular to a library audio and video collaborative recommendation system that can reduce the search time according to the popularity ranking method generated by the collaborative recommendation method, which is a popular module available. All users recommend library collections and calculate their click-through rates for integration preferences and popularity. Finally, a presentation module presents the recommended categories, allowing users to reduce search time to improve search efficiency and increase Its system is practical.


A secondary object of the present invention is to provide a library collection collaborative recommendation system, which has a forgetting preference module and a forgotten hot module, which can delete data that is not often clicked to save system storage capacity.

The specific construction of the present invention and its practical functions are described in detail below.

(11). . . Book video search module

(12). . . Library video library

(13). . . Personalized knowledge module

(14). . . Popular module

(15). . . Rental database

(16). . . Personalized ontology

(17). . . Popular database

(18). . . Integration module

(19). . . Presentation module

(2). . . Control interface

(20). . . main screen

(twenty one). . . Interface menu

(twenty two). . . Page control button

(3). . . Collection query key

(31). . . Collection query bar

(32). . . Query key

(33). . . Movie selection button

(34). . . Borrowing status

(35). . . Reservation button

(36). . . Recommended as popular item button

(361). . . Action class

(362). . . Comedy class

(363). . . Cartoon class

(364). . . Sports class

(365). . . Horror

(366). . . Idol class

(367). . . Kungfu class

(368). . . Exploration class

(37). . . image

(38). . . Movie introduction

(4). . . New film information button

(41). . . New film name

(42). . . Movie introduction

(43). . . Tracking button

(5). . . Hot topic recommendation button

(51). . . Action class

(52). . . Horror

(53). . . Idol class

(54). . . Comedy class

(55). . . Kungfu class

(56). . . Cartoon class

(57). . . Sports class

(6). . . Rental query key

(61). . . Rental loan inquiry column

(62). . . Query key

(63). . . Book picture

(64). . . Book introduction

(65). . . Lending record

(7). . . Back to the library key

(8). . . Viewer data query key

The first figure is a system flow block diagram of the library collection collaborative recommendation system of the present invention.
The second figure is a block diagram of the use flow of the library collection collaborative recommendation system of the present invention.
Third, fourth, fifth, sixth, seventh, and eighth diagrams: A schematic diagram of the operation of the control interface of the present invention.

For a more detailed understanding of the structure of the present invention, please refer to the first to third figures, the present invention provides a library collection collaborative recommendation system, the system is provided with a control interface 2, and the system includes a book video search The module 11 includes a search module and a data processing module, which can be used by a student, a teacher or other users for searching and inputting a library video and audio data after inputting a search. The user 12 can select a personal knowledge module 13 or a hot module 14 , and the personalized knowledge module 13 includes a new concept module, a forgetting preference module, an update preference module, and a The rental processing module has a new concept module that allows users to enter new data. The Forgotten Preferences module can delete data that is not frequently selected in the record. The update preference module can add new concept modules and forgetting preferences. The data of the module is updated, and the rental processing module can record the rental status in the personalized knowledge module 13, and the hot module 14 includes a collaborative recommendation module and a single reading. The module, a forgotten hot module and a rental processing module, the collaborative recommendation module stores data recommended by all users such as students, teachers or other users, and calculates the number of clicks through the click rate module. When the popular data has been retired, the number of clicks calculated by the click rate module is too low and deleted by the forgotten hot module. The rental processing module can record the rental status in the hot module 14 and rent a loan. The database 15 can store the rental status in the rental processing module of the personalized knowledge module 13 and the rental status in the rental processing module in the hot module 14 when the user selects to use the personalized knowledge module. At 13:00, the user can further log in to a personalized knowledge ontology 16 to perform data update and storage of a single individual, and the data in the hot module 14 can be further stored in the popular database 17, and the integration in the library collection collaborative recommendation system. The module 18 can integrate the data in the personalized knowledge body 16 with the data in the popular database 17, and the integration module 18 includes a score integration module and a sorting module. The score integration module integrates the data in the personalized knowledge ontology 16 with the data in the popular database 17 by the score rate, and then the ranking module performs the order of the click rate, and the last presentation module 19 The presentation module 19 includes a classification presentation module and a general presentation module, which allows the user to select the presentation module 19 to display the data classification status or the general display status.

Among them, the popular module 14 can use the update method of the parent concept preference value P _P , and the calculation formula is as follows:

Where α _{P is the} parent concept influence parameter value, which is used to correct the influence degree of the parent concept preference, and N _S is the number of sub-concept nodes, which refers to the number of sub-concepts under the parent concept, the sub-concept preference value The update method of P _S is as follows:


Among them, α _S is the sub-concept influence parameter value, which is used to correct the influence degree of the sub-concept preference, and the update method of the neighbor concept preference P _N is as follows:

Among them, α _N is the influence value of the neighbor concept, which is used to correct the influence degree of the neighbor concept preference, and N _N is the number of neighbor concept nodes, which refers to the number of neighbor concepts around the concept.

The control interface 2 has a main screen 20, an interface menu 21 and a page turning control key 22 (refer to the third figure). The interface menu 21 is provided with a collection inquiry key 3, a new piece of information key 4, and a hot topic. The recommendation key 5, a rent inquiry key 6, a library key 7, a viewer data inquiry key 8, and the main screen 20 displays the content of the selection key in the interface menu 21, when the collection inquiry key 3 is pressed. The main screen 20 displays a collection query field 31, a query button 32 and at least one movie selection button 33. The movie selection button 33 displays a movie name, a video introduction button, a tracking button, and a display. The borrowing status of the movie, when the movie has not been borrowed, displays "on the shelf". Otherwise, the movie has been borrowed to display "borrowing". When a movie selection button 33 is pressed, the main screen 20 The relevant content of the movie is displayed (please refer to the third and fourth figures at the same time), and the related content is a borrowing state 34 of a movie, a reservation button 35, a recommendation for the popular item button 36, a picture 37, and a movie. Introduction 38, and when pressed to recommend as a hot item At 36 o'clock, the main screen 20 displays a plurality of recommended categories, which are action class 361, comedy class 362, cartoon class 363, sports class 364, horror class 365, idol class 366, kung fu class 367, exploration class 368 (please Referring to the fourth and fifth figures at the same time, when the new information button 4 is pressed, as shown in the sixth figure, the main screen 20 displays the new information of the library, which is a new film name 41. A movie profile 42 and a tracking button 43, and when the hot topic recommendation button 5 is pressed (refer to the seventh figure), the main screen 20 displays a plurality of popular topic recommendation categories, which are action classes 51. , horror class 52, idol class 53, comedy class 54, kung fu class 55, cartoon class 56, sports class 57, and the hot topic recommendation category shows the top rank of the category, that is, students, teachers or other users When all the users recommend the information, and when the rent inquiry key 6 is pressed (refer to the eighth figure), the main screen 20 displays the rental query column 61, a query key 62, and the queried book picture 63. Book description 64 and the book's lending record 65.

Please refer to the first and second figures. When the user wants to search for books or audio and video, he inputs the keyword from the control interface 2 to enter the book video search module 11 to search the database, and the system enters the library video library 12 Searching, after the search is completed, the user returns to the personalized knowledge body 16, which allows the personalized knowledge body 16 to determine whether the keyword has an extended concept in the personalized knowledge body. When the judgment is "Yes", Then, the update preference module performs the update preference value, and then returns the updated click value to the personalized knowledge body 16, and when the determination is No, the new concept module can be added. The keyword extension concept and the new keyword extension concept are transmitted back to the personalized knowledge ontology 16, and the library collection collaborative recommendation system integrates the information in the personalized knowledge ontology 16 with the data in the popular database 17 With the popular, finally, a presentation module 19 presents the recommended category to display the search results.

In summary, the library collection collaborative recommendation system of the present invention has a control interface 2, and the system includes a library video search module 11, a library audio and video database 12, and a humanized knowledge module. 13. A hot module 14, an integrated module 18, and a presentation module 19, and the control interface 2 has a main screen 20, an interface menu 21, and a page turning control key 22. The interface menu 21 is provided with a The collection inquiry key 3, a new piece of information key 4, a hot topic recommendation key 5, a rental inquiry key 6, a library key 7, a viewer data inquiry key 8, and the main screen 20 displays the interface menu 21 When the collection inquiry key 3 is pressed, the main screen 20 displays a collection query field 31, a query key 32 and at least one movie selection key 33, and the movie selection key 33 displays a The name of the movie, a video introduction button, a tracking button, and the status of the borrowing of the movie. When the movie has not been loaned, it displays "on the shelf". Otherwise, the movie has been loaned to display "borrowing". ", when pressing a movie selection The key 33, the main screen 20 displays the related content of the movie, the related content is a borrowing state 34 of a movie, a reservation button 35, a recommendation for the popular item button 36, a picture 37 and a movie profile 38, and When the recommended hot item button 36 is pressed, the main screen 20 displays a plurality of recommended categories, which are action class 361, comedy class 362, cartoon class 363, sports class 364, horror class 365, idol class 366, kung fu. The class 367, the exploration class 368, and when the new piece of information button 4 is pressed, the main screen 20 displays the new piece of new information of the library, the new piece of information is a new piece name 41, a movie profile 42 and a tracking button 43. When the hot topic recommendation button 5 is pressed, the main screen 20 displays a plurality of popular topic recommendation categories, which are action class 51, horror class 52, idol class 53, comedy class 54, kung fu class 55, cartoon The class 56 and the sports class 57, and the hot topic recommendation category shows the top ranking of the category, and when the rent inquiry key 6 is pressed, the main screen 20 displays the rental query column 61, a query key 62 and Queryed book picture 63, book About 64 and 65 lending record of the book.

(11). . . Book video search module

(12). . . Library video library

(13). . . Personalized knowledge module

(14). . . Popular module

(15). . . Rental database

(16). . . Personalized ontology

(17). . . Popular database

(18). . . Integration module

(19). . . Presentation module

Claims (9)

A library collection collaborative recommendation system, the system is provided with a control interface, and the system includes a library video search module, the book video search module includes a search module and a data processing module, which can Students, teachers, or other users search for it. After entering the search, they will enter a library video library. Users can choose to use a humanized knowledge module or a popular module, and when users choose to use personalized knowledge. When the module is used, the user can further log in to a personal knowledge body to update and store the data of a single person, and the data in the hot module can be further stored in the popular database, which is characterized in that the popular module can utilize the parent. The way to update the concept preference value P _p is as follows: Where α _{p is the} parent concept influence parameter value, which is used to correct the degree of influence of the parent concept preference, and N _s is the number of sub-concept nodes, which refers to the number of sub-concepts under the parent concept, sub-concept preference The calculation method of the value P _s is as follows: P _s =P _s +α _s , where α _s is the sub-concept influence parameter value, which is used to correct the influence degree of the sub-concept preference, the neighbor concept preference P The update method of _N is as follows: Wherein, α _N o concept impact parameter, which is used to influence the degree of preference correction concept o, o and N _N is the number of nodes concept, which refers to the number of all the surrounding neighborhood concept of this concept, its system The integrated module system integrates the data in the personalized knowledge book with the data in the popular database, and finally displays the data by a presentation module. The library collection collaborative recommendation system according to claim 1, wherein the personalized knowledge module includes a new concept module, a forgetting preference module, an update preference module, and a rental processing module. The new concept module allows the user to enter new data, and the Forgotten Preferences module can delete the data that is not frequently selected in the record. The update preference module can add the data of the new concept module and the forgetting preference module. To update the number of clicks, the rental processing module can record the rental status in the personalized knowledge module, and the popular module includes a collaborative recommendation module, a point rate module, a forgotten hot module and a rental processing module. The collaborative recommendation module stores the information recommended by all users such as students, teachers or other users, and calculates the number of clicks through the click rate module. When the popular data has been retired, it will be due to the point. The reading rate calculated by the reading rate module is too low and is deleted by the forgotten hot module. The library collection collaborative recommendation system according to claim 2, wherein the integration module comprises a score integration module and a sorting module, which can be personalized by the score integration module The data is integrated with the data in the popular database for the number of click-through rates, and then the ranking module performs the ranking of the number of click-through rates. The presentation module includes a classification presentation module and a general presentation module, which allows the user to Selecting the rendering module displays the data classification status or the general display status. The library collection collaborative recommendation system according to claim 3, wherein the control interface has a main screen, an interface menu and a page turning control key. The library collection collaborative recommendation system according to claim 4, wherein the interface menu is provided with a collection inquiry key, a new piece information key, a hot topic recommendation key, a rental inquiry key, and a library key. , a viewer data query button, and the main screen displays the contents of the selection button in the interface menu. The library collection collaborative recommendation system according to claim 5, wherein when the collection inquiry key is pressed, the main screen displays a collection query column, an inquiry key, and At least one movie selection button, the movie selection button displays a movie name, a video introduction button, a tracking button, and displays the borrowing status of the movie, and displays the "on the shelf" when the movie has not been loaned. Conversely, if the movie has been loaned out, it will display "Borrowing". When the video selection button is pressed, the main screen displays the related content of the movie. The related content is the borrowing status of a movie, and a reservation button. One recommended as a popular item button, a picture and a video introduction, and when the recommended item is pressed, the main picture shows a plurality of recommended categories, which are action, comedy, cartoon, sports. , horror, idol, kung fu, and exploration. For example, in the library collection collaborative recommendation system described in claim 5, when the new information button is pressed, the main screen displays the new information of the library, and the new information is a new title. Video introduction and a tracking button. For example, in the library collection collaborative recommendation system described in claim 5, when the hot topic recommendation button is pressed, the main screen displays a plurality of popular topic recommendation categories, which are action classes, horror classes, Idol class, comedy class, kungfu class, cartoon class, sports class, and the popular topic recommendation category shows the popular ranking of the category, that is, the information recommended by all users such as students, teachers or other users. The library collection collaborative recommendation system according to item 5 of the patent application scope, wherein when the rent inquiry key is pressed, the main screen displays a rental loan inquiry column, a query key, a queryed book picture, a book introduction, and The book's lending record.