TW202011749A - Multimedia data recommending system and multimedia data recommending method - Google Patents

Abstract

A multimedia data recommending system includes a storage device and a processor. The storage device includes a first storage unit and a second storage unit. The first storage unit is configured to store the multimedia data fragment. The first storage unit is coupled to at least one client electronic device via network. The first storage unit is configured to store the operating data generated by at least one client electronic device interacting with the multimedia data system. The processor is coupled to the storage device. The processor includes a response analysis unit and a timing unit. The processor is configured to analyze the multimedia data fragment in the first storage unit to generate a relevance link between the multimedia data fragment. The processor is also configured to analyze the operating data in the second storage unit to generate a recommended list according to the multimedia data fragment and the operating data, the recommended list includes the multimedia data fragment.

Description

Multimedia data recommendation system and multimedia data recommendation method

This disclosure document relates to a multimedia data recommendation system and multimedia data recommendation method, in particular to a multimedia data recommendation system and multimedia data recommendation method based on user habits and preferences.

With the development of distance teaching, the boundaries of space are broken, and users can choose to study in ways such as TV broadcasting or the Internet according to their own preferences and convenience. However, there are so many course materials that it is not easy for users to quickly learn the suitable course materials. Most of the common learning list arrangement methods are that the system has a certain preset order according to the theme or administrator setting. The user cannot determine whether the current course is suitable for his current preferences and needs.

There are many kinds of multimedia materials. Everyone likes different types of courses, themes, methods, and teaching methods. Different people have different acceptances of the same course. Course materials must meet the preferences of users to have good learning results.

In an embodiment of the present disclosure, a multimedia data recommendation system includes a storage device and a processor. The storage device includes a first storage unit and a second storage unit. The first storage unit is used to store multimedia data paragraphs. The second storage unit is connected to at least one user device through a network, and is used to store operation data generated by at least one user device interacting with the multimedia data system. The processor is coupled to the storage device, and the processor is used to analyze the multimedia data in the first storage unit to generate a correlation link between the multimedia data, and analyze the operation data in the second storage unit, and connect and operate according to the correlation The data generates a corresponding recommendation list for at least the user device to display the recommendation list, and the recommendation list records multimedia data. The processor includes a response analysis unit and a timing unit. The operation data includes at least one response to the multimedia data segment. The response analysis unit is used to analyze to obtain at least one response corresponding to at least one question in the multimedia data segment. The timing unit is used To calculate the playing time of the multimedia data paragraph.

In another embodiment of the present disclosure, the multimedia data recommendation method includes the following operations: storing multimedia data through the first storage unit; storing operation data generated by at least one user device interacting with the multimedia data system through the second storage unit; and analyzing The multimedia data segments in the first storage unit to generate the related links between the multimedia data segments; analyze the operation data in the second storage unit, and generate a corresponding recommendation list according to the related links and the operation data, the recommendation list records multimedia Information paragraph.

In summary, the multimedia data system stores multimedia data through the first storage unit, stores operation data through the second storage unit, and analyzes and calculates the multimedia data in the first storage unit and the operation data in the second storage unit, And generate the corresponding recommendation list data.

The words "including", "having", etc. used in this article are all open terms, which means "including but not limited to". In addition, "and/or" used in this article includes any one or more of the items listed in the relevant list and all combinations thereof.

In this article, when an element is called "connected" or "coupled", it can be referred to as "electrically connected" or "electrically coupled." "Link" or "Coupling" can also be used to indicate the operation or interaction of two or more components. In addition, although terms such as "first", "second", etc. are used in this document to describe different elements, the terms are only used to distinguish elements or operations described in the same technical terms. Unless the context clearly dictates, the term does not specifically refer to or imply the order or order, nor is it intended to limit the present disclosure.

Please refer to FIG. 1, which illustrates a functional block diagram of a multimedia data system according to an embodiment of the present disclosure. The communication connection between the multimedia data system 100 and the user device 200 can be connected to each other via wired or wireless communication, so as to achieve data transmission. In an embodiment, the multimedia data system 100 may be an English online learning platform, the user device 200 may be a personal desktop computer, a notebook computer device, a tablet computer device, or a smart mobile communication device, etc. The multimedia data may be online Courses, including movies, articles, sounds, slides or other media with learning materials.

In one embodiment, the multimedia data system 100 includes a processor 120 and a storage device 140. The processor 120 and the storage device 140 are communicatively connected to the user device 200, and the processor 120 transmits data with the user device 200, and stores the data in the storage device 140 or reads the data from the storage device 140. In an embodiment, the processor 120 may be a central processing unit, a microprocessor, or other components with data processing functions. The storage device 140 may be a hard disk, a disk array, a tape drive, a non-volatile memory, or other electronic storage media.

In one embodiment, the operation data includes at least one response to the multimedia data segment. The processor 120 includes a response analysis unit 122 and a timing unit 124. The response analysis unit 122 is used to analyze to obtain at least one corresponding to the multimedia data segment At least one response to a question, such as obtaining answers to online quizzes from users on English teaching videos. The timing unit 124 is used to calculate the playing time of the multimedia data paragraphs, for example, the playing time of the user watching an English teaching video.

In one embodiment, the storage device 140 includes a first storage unit 142 and a second storage unit 144. The user device 200 is communicatively connected to the second storage unit 144, and the second storage unit 144 stores data corresponding to the user device 200. The first storage unit 142 is used to store multimedia data paragraphs. The second storage unit 144 is used to store operation data generated by the interaction between the user device 200 and the multimedia system 100, such as logs or logs. The processor 120 is used to analyze and calculate the multimedia data segments and operation data stored in the first storage unit 142, and generate corresponding recommendation list data for the user device 200 to display the recommendation list.

In an embodiment, the storage device 140 further includes a third storage unit (not shown). The third storage unit is used to store multimedia data segments corresponding to the first storage unit 142 (eg, actual video mp4 files and video files) Meta data after uploading to the cloud platform database in the multimedia data system 100. In an embodiment, the third storage unit may be integrated with any one of the first storage unit 142 and the second storage unit 144, which is not limited to the above.

In one embodiment, the first storage unit 142 may be a course database for storing online courses, and the second storage unit 144 may be a back-end database for storing operation data of a personal desktop computer on an English online learning platform. The operation data is, for example, when the user clicks the online course data to be viewed with the mouse, the back-end database will store the clicked item as operation data.

It should be noted that the implementation of the devices and components in the multimedia data system 100 is not limited to those disclosed in the above embodiments, and the connection relationship is not limited to the above embodiments. Anything sufficient to enable the multimedia data system 100 to achieve the following Both the connection and implementation of the technical content can be applied to this case.

Please refer to Figure 1 and Figure 2 at the same time. FIG. 2 illustrates a flowchart of a multimedia data processing method according to an embodiment of the present disclosure. The processing method 300 in FIG. 2 includes step S310, step S320, step S330, and step S340. First, in step S310, the multimedia data segment is stored through the first storage unit 142 in the storage device 140. In one embodiment, the multimedia data segment includes non-segmented multimedia data, such as a complete movie or article, or multiple segmented multimedia data segments, such as an article segment segmented according to the theme.

Next, in step S320, through the second storage unit 144 in the storage device 140, the operation data generated by the interaction between the user device 200 and the multimedia system 100 is stored.

The following describes the method for generating operating data in the multimedia data system 100. In an embodiment, the method for generating the operation data may be based on the number of questions asked in the discussion area. For example, when a user views multimedia data in an English online learning platform, in addition to displaying an interface of the multimedia data content, the platform also includes a discussion area corresponding to the multimedia data. The discussion area provides users with questions or discussions about the content. For example, when the user has questions about the content of the video, he can ask questions in the discussion area, and when he asks, he can mark the time of the video to let the person who wants to answer more quickly know which paragraph of the video the problem appears in.

For example, if the user has questions about the 1:00 in the video, he can ask questions in the discussion area and mark the 1:00 in the video. Other users can see this question in the discussion board when watching this video, and can answer it below. People who watch this movie in the future can find out whether the same question has been answered in the discussion board if they have the same question at 1:00 in the movie.

In an embodiment, the method of generating the operation data may be based on the number of marks. For example, when viewing a paragraph of multimedia data, a user may add a mark to the content, and the mark may further add a comment. In addition to the annotations added by the users themselves, they can also view the marks added by other users or apply the marks and annotations used by professionals through settings. For example, when watching a video, the user thinks that the video is at 3:00, and he can directly mark it at 3:00 and attach a comment to mark it as the key. Other users can view the marks and comments of other users through the settings, and the system can also count the most user marks to provide user application.

In an embodiment, the method for generating the operation data may be based on the viewing record. For example, when the processor 120 calculates the user's record of viewing multimedia data paragraphs, that is, statistical viewing records. The statistical results have the highest percentage of videos and the lowest percentage of articles, indicating that the user prefers to use videos for learning, has a higher acceptance of the videos, or wants to improve English listening skills, but does not like to read text or need to improve English reading ability. The processor 120 learns the user's preferred learning method and theme by analyzing the online learning course that the user has viewed. For example, the user’s viewing records are mostly travel-type English articles, indicating that the user is more interested in the subject of travel and prefers to read or wants to improve reading ability. This analysis result is stored in the second storage unit 144 as operation data .

In one embodiment, the method for generating the operation data may be based on the performance of the online test or the analysis result of the response to the question. In other words, the aforementioned operation data includes the user's response to the aforementioned multimedia data paragraph during operation. If the questionnaire is included in the paragraph of the multimedia data, the response is "answer". For another example, the online quiz allows users to understand their own learning results. The online quiz can be true or false, multiple choice, essay, essay or oral test. In addition to the online test scores, the time and speed of answering questions are one of the parameters of the response analysis unit 122 in the processor 120 to evaluate the user's learning results. The processor 120 analyzes the above parameters to generate an analysis result and stores the analysis result For operation information. For example, during the online test, the analysis result may be the correctness of the user's answer when performing the online test, and the analysis result will have both correct and incorrect results.

In an embodiment, the method for generating the operation data may be based on the operation when watching the online video, such as the actual playing time of the multimedia data. When users watch English online courses of different difficulty, they will have different operations. For example, English beginners (e.g. NBC Elementary Level) who have difficulty viewing English online courses (e.g. NFL Intermediate Level) will need to spend more time to watch, such as repeatedly watching specific passages or pausing videos, actually playing The time will be greater than the time of the movie itself. The timing unit 124 in the processor 120 will calculate the actual playback time of the user. The second storage unit 144 will store these records and the actual playback time as operation data. The processor 120 can determine to watch repeatedly Of online courses are learning topics that users are interested in or need to strengthen.

For example, online quiz performance includes quiz score or quiz time. The processor 120 may refer to the playing time, the reference answer time, and the reference score. Judging the user's learning level by calculating the playing time when the user watches the online course or the score and answer time of the online quiz. When the playing time is higher than the reference playing time, it is difficult for the user to judge the online course, and when the playing time is lower than the reference playing time, the judgment is simple for the user. It is difficult for the user to judge the online test when the answer time is longer than the reference answer time or when the score is lower than the reference score. When the answer time is lower than the reference answer time or when the test score is higher than the reference score The user is simple. Therefore, the processor 120 can know how difficult the current multimedia data segment or online test is for the user through the above determination method. In this embodiment, the nature of the multimedia data paragraph may be the difficulty of the multimedia data paragraph. The setting method of the reference score can be the average score of all users who have passed the test. The setting method of the reference answer time can be the average time spent by all users who have passed the test. The setting method is not limited to the above, but can also be other Setting method.

The method for generating the operation data of the present disclosure is not limited to the above. The above is an example of the method for generating the operation data. All operations performed by the user on the multimedia system 100 by the user device 200 are included in the description of the present disclosure. Operating information. The operation data represents the user's usage habits and behaviors, and subsequent evaluation data is generated by analyzing the operation data, and the processor 120 can further know the user's preference or degree through the evaluation data.

In step S330, the multimedia data segment in the first storage unit 142 and the operation data in the second storage unit 144 are analyzed and calculated. In one embodiment, step S330 includes further steps S331 and S332. In step S331, the content of the multimedia data segment in the first storage unit 142 is relatedly linked. In step S332, evaluation data is generated according to the operation data in the second storage unit.

The following describes the analysis method of the correlation between the content of the multimedia data paragraphs in step S330. The content of the multimedia data paragraph takes a video as an example, and also includes video subtitles, discussion board content, video tags, and so on. The relevance of the content of different multimedia data paragraphs may be due to the similarity of topics, subtitles, course descriptions or difficulty, and high similarity. In addition, taking the film as an example, two films with different themes may also have some content relatedness, such as the 2:00~3:00 of the A film of the travel theme and the 5 of the B film of the food theme: From 00 to 6:00, the content is related, and the content may all describe the travel mode and food culture of the same country. Therefore, it is necessary to segment the content of different videos and establish related links between related paragraphs.

The following describes the method for establishing the relevance link in step S331. Please refer to FIG. 3, which is a schematic diagram illustrating the relevance of the content of multimedia data paragraphs according to an embodiment of the present disclosure. Figure 3 contains multimedia data A1, multimedia data B1 and multimedia data C1, and corresponding paragraphs A11~A14, paragraphs B11~B13 and paragraphs C11~C14. In this embodiment, A1 is a movie, B1 is an article, and C1 is a movie as an example. The following describes how to analyze the correlation between online course content.

In the multimedia data A1, the processor 120 may first analyze the video content of A1 and use the video content as a segment reference to divide into a plurality of unit segments. Here, taking four paragraphs A11, A12, A13, and A14 as examples, the number of multimedia data segments of the disclosed document is not limited to four, and can be divided into other number of unit paragraphs according to the actual situation. The method of analyzing the segmentation of the video content can be analyzed according to the way that the video subtitles, discussion forums, video tags, etc. in the multimedia data A1 can know the video content. The method of analyzing the segmentation can be, for example, for keywords, such as The beginning or ending sentences commonly used by speakers in the film to distinguish different unit paragraphs. In addition to the analysis by the processor 120, it can also be segmented by the person who makes the multimedia data A1 or an expert with considerable knowledge. The processor 120 can segment different paragraphs by the above-mentioned marking method, allowing the user to directly click on the marking when viewing the multimedia data A1 or jump to the paragraph to be viewed by viewing the annotation, which makes the operation more convenient , Other different types of multimedia data can also apply the same marking method.

In the multimedia data B1, the method of analyzing the content and paragraphs of the article is used as a reference to distinguish different paragraphs of the unit. Generally, different paragraphs in the article content will be separated by two spaces at the beginning of the paragraph as a way to distinguish different paragraphs. The test text paragraph has two empty spaces as the beginning of the paragraphs of different units. In this embodiment, taking three paragraphs B11, B12, and B13 as an example. The segmentation method of the multimedia data C1 is similar to that of the multimedia data A1. Here, the example is divided into four paragraphs C11, C12, C13, and C14.

The processor 120 divides the different multimedia data A1, B1, and C1 into sections according to the content, and then analyzes the correlation between the multimedia data A1, B1, and C1. There may be a correlation between the content of different paragraphs in different multimedia data. In this embodiment, the paragraph A12 in the multimedia data A1 is related to the paragraph B12 in the multimedia data B1 and the paragraph C11 in the multimedia data C1. The processor 120 establishes an associative connection between paragraph A12, paragraph B12 and paragraph C11, as shown by the curved arrows in FIG. 3. For example, after watching the movie content of paragraph A12, in addition to continuing paragraph A13 in the preset order, it is also suitable to continue to read the article content of paragraph B12, and then watch the movie content of paragraph C11. The article content of paragraph B12 may be an article paragraph describing the movie content of paragraph A12, and is one of the knowledge bases of paragraph C11. In this case, it is not necessary to follow paragraph A11 in the order of multimedia data A1 as the traditional way of watching movies , A12, A13, and A14 after watching the multimedia material B1.

In one embodiment, the related links between the multimedia data segments can also be established in advance when uploading the multimedia data. For example, after the uploading is completed, the professional who makes the video will first associate the related content in different videos. To establish associative links. For example, after an English lecturer uploads multiple English teaching videos that he has made, he will first establish a related link to the related parts of the content.

In an embodiment, the multimedia data system 100 does not segment the multimedia data, and directly generates related links according to the contents of the multimedia data A1, the multimedia data B1, and the multimedia data C1.

In step S332, the operation data includes evaluation data, and the processor 120 further generates evaluation data corresponding to the current user by analyzing the operation data generated in step S320 and the difficulty corresponding to each multimedia data segment. The evaluation data includes evaluation Corresponding to the user's degree of the multimedia data paragraph, the multimedia data paragraph suitable for the user is judged. In other words, the processor 120 can judge the degree of the user by evaluating the data. For example, in the operation data, the proportion of the multimedia data watched by the user is the highest in the national primary examination, then the processor 120 judges that the level corresponding to the user is the national primary examination and generates corresponding evaluation data, and then uses the The assessment of the degree of the person is the first level of the British national inspection. The assessment method is not limited to the above, and it can also be other statistical methods and judgment methods.

For example, the method for generating the evaluation data may be a result of further evaluation based on analyzing the operation data generated by the various methods in the above embodiments. For example, if the performance of the online quiz is lower than the average score of the Intermediate National Examination Intermediate Test, the current user’s assessment is generated as the primary evaluation data of the National General Examination. The processor 120 can learn the current user through the evaluation data. Degree.

In other embodiments, in the case where the user device 200 and the multimedia system 100 are connected to access the multimedia data segment, the multimedia data system 100 generates a user operation record, and the operation record is stored in the second storage unit 144. The operation record includes the record generated when the user uses the multimedia data segment in the first storage unit 142. The operation record represents the historical record of the multimedia data segment used by the user. Analyzing the result of the operation record can know the user's favorite multimedia The type of data paragraph or the degree of user. By analyzing the operation record, the processor 120 will evaluate the operation record, evaluate the preferences and degree of the multimedia data segments corresponding to the operation record, so as to determine the multimedia data suitable for the user.

Finally, in step S340, the corresponding recommendation list data is generated for the user device 200 to display the recommendation list according to the related links and operation data in step S330, and the content of the recommendation list is changed according to the updated related links and operation data. The following description uses an embodiment as an example.

Please refer to FIG. 4, which is a schematic diagram of a learning path of a recommendation list according to an embodiment of the present disclosure. Figure 4 contains multimedia data A2~F2. With the above-mentioned related links and operation data, the processor 120 generates a corresponding recommendation list. In this embodiment, the initial recommendation list includes multimedia data A2 and B2. After the user views the multimedia data A2 and B2, corresponding connection and operation data will be generated. Again, through the connection and operation data, the processor 120 judges that the multimedia data C2 and the multimedia data E2, the user is more suitable to continue View the multimedia data C2, and add the multimedia data C2 to the recommendation list, and then judge the multimedia data D2 and the multimedia data F2 again. The judgment result is that the multimedia data D2 is more suitable than the multimedia data F2. Therefore, the final learning path is multimedia data A2, B2, C2 and D2.

For example, multimedia materials A2 and C2 are the first-level course videos of the British National Examination, and multimedia materials B2 and E2 are the intermediate-level course videos of the British National Examination. The user did not pause the video or ask questions in the discussion area when viewing the multimedia data A2. The online test score was higher than the average score. When viewing multimedia data B2, pause 5 times, and post 2 times in the discussion forum. The score of the online test is lower than the average score and the time is higher than the average time. The processor 120 generates the operation data for the above operation. The processor 120 determines that the content of the multimedia data B2 and the multimedia data C2 and E2 are related and establishes a related link. With the above-mentioned operation data, the processor 120 generates evaluation data that judges the current user's level as the primary level of the UK national inspection rather than the intermediate level of the UK national inspection, and then changes the learning path to connect the multimedia data C2 instead of the multimedia data E2. The situation of the multimedia data D2 and F2 is similar to the above. The processor 120 determines that the current user is more suitable for the multimedia data D2 than the multimedia data F2. After the above operations, the multimedia data displayed in the recommended list at the end are multimedia data A2~D2, and the recommended learning path for the user is multimedia data A2 to multimedia data D2.

In an embodiment, in addition to changing the order of the multimedia data in the recommendation list, it is also possible to add or remove multimedia data in the recommendation list to change the content of the recommendation list. For example, in the above embodiment, after the user has watched the multimedia data D2, the processor 120 generates corresponding evaluation data according to the user's operation data after viewing the multimedia data D2 to determine the multimedia data D2 for the current user. Said the difficulty is simple, you can add and recommend other multimedia data related to the multimedia data D2 and more difficult to connect.

In one embodiment, the storage device 140 further includes an analysis database to store the recommendation list and the results of the learning path change, and to store the user's results in real time. The processor 120 further analyzes the recommendations in the analysis database List materials and learning paths and update recommended lists. Therefore, the data of the recommendation list is updated in real time or regularly with the user's behavior and the result after using the multimedia data. For example, the processor 120 may set an interval to update the recommendation list, for example, update every hour, and the manner of setting the interval is not limited to the above.

In summary, the multimedia data system analyzes the correlation between the content of the multimedia data segments and establishes a correlation link, and analyzes the operation data of the interaction between the user device and the multimedia data system to generate evaluation data. Related links and evaluation data to generate a recommendation list suitable for the user, and update the recommendation list in real time or periodically as the user uses the multimedia data paragraphs and usage behaviors on the recommendation list, and the result of the learning path selected by the user Save as one of the basis for updating the recommendation list later.

100: Multimedia data system 120: processor 140: storage device 142: First storage unit 144: Second storage unit 200: user device 300: data processing method S310, S320, S330, S331, S332, S340: steps A1, B1, C1, A2, B2, C2, D2, E2, F2: multimedia data A11~A14, B11~B13, C11~C14: paragraph

FIG. 1 is a functional block diagram of a multimedia data system according to an embodiment of the present disclosure. FIG. 2 illustrates a flowchart of a multimedia data processing method according to an embodiment of the present disclosure. FIG. 3 is a schematic diagram of multimedia data content relevance according to an embodiment of the present disclosure. FIG. 4 is a schematic diagram of a learning path of a recommendation list according to an embodiment of the disclosed document.

100: Multimedia data system

120: processor

122: Response analysis unit

124: timing unit

140: storage device

142: First storage unit

144: Second storage unit

200: user device

Claims (14)

A multimedia data recommendation system, including: A storage device, including: A first storage unit for storing a plurality of multimedia data segments; and A second storage unit connected to at least one user device via a network and used to store operation data generated by the interaction of the at least one user device with the multimedia data system; and A processor, coupled to the storage device, is used to analyze the multimedia data segments in the first storage unit to generate a plurality of related links between the multimedia data segments, and to analyze the second storage unit in the second storage unit Some operation data, and generate a corresponding recommendation list according to the related links and the operation data, wherein the recommendation list records the paragraphs of multimedia data. The multimedia data recommendation system according to claim 1, wherein the processor is used to segment the content of at least one multimedia data, generate the multimedia data segments, and generate the multimedia data segments according to the multimedia data segments These related links between. The multimedia data recommendation system according to claim 2, wherein the processor is used to calculate the similarity between the multimedia data segments, and generate the correlations for the content-related parts of the multimedia data segments according to the calculation result Sexual connection. The multimedia data recommendation system according to claim 1, wherein the processor includes: A response analysis unit, wherein the operation data includes at least one response to the multimedia data segments, the response analysis unit is used to analyze to obtain the at least one response corresponding to at least one question in the multimedia data segments, The operation data further includes the response analysis unit analyzing the at least one question corresponding to the multimedia data paragraphs and the analysis result of the at least one response. The multimedia data recommendation system according to claim 1, wherein the processor includes: A timing unit is used to calculate a playing time of the multimedia data segments, wherein the operation data includes the playing time of the multimedia data segments. The multimedia data recommendation system according to claim 5, wherein the multimedia data paragraphs each correspond to a property, and the operation data is updated according to the property. The multimedia data recommendation system according to claim 6, wherein the processor changes the content of the recommendation list provided to the at least one user device according to the updated operation data and the related links. A multimedia data recommendation method, including: Through a first storage unit, a plurality of multimedia data segments are stored; Through a second storage unit, storing operation data generated by at least one user device interacting with a multimedia data system; Analyzing a plurality of related links between the multimedia data segments in the first storage unit; and Analyze the operation data in the second storage unit, and generate a corresponding recommendation list according to the association links and the operation data, where the recommendation list records the multimedia data paragraphs. The multimedia data recommendation method described in claim 8 further includes: The content of at least one multimedia data is segmented to generate the multimedia data segments, and the related links between the multimedia data segments are generated according to the multimedia data segments. The multimedia data recommendation method described in claim 9 further includes: Calculate the similarity between the multimedia data segments, and generate the related links for the content-related parts of the multimedia data segments according to the calculation result. The multimedia data recommendation method according to claim 8, wherein the steps of analyzing the operation data in the second storage unit and generating a corresponding recommendation list according to the related links and the operation data include: Analyze at least one response corresponding to at least one question in the multimedia data paragraphs to generate an analysis result. The multimedia data recommendation method according to claim 8, wherein the step of analyzing the operation data in the second storage unit and generating the corresponding recommendation list according to the related links and the operation data further includes: Calculate a playing time of the multimedia data paragraphs. The multimedia data recommendation method described in claim 12, further includes: According to the nature of the paragraphs corresponding to the multimedia data, the operation data is updated. The multimedia data recommendation method according to claim 13, wherein the steps of analyzing the operation data in the second storage unit and generating a corresponding recommendation list according to the related links and the operation data include: After generating the recommendation list, the content of the recommendation list provided to the at least one user device is changed according to the updated operation data and the related links.

Images