TWI715488B - A time-based recommendation system applied to iptv and method thereof - Google Patents

Abstract

The invention provides a time-based recommendation system applied to IPTV and method thereof, including a processing module, a calculation module and a playback module. The invention cuts IPTV programs according to time to provide different time-based recommendation lists, making it suitable for different users in the same home. And calculate preset scores, clicks, revenue and dynamic parameters to get the time-based recommendation lists that matches users' using preferences and increase the revenue of IPTV platform operators to improve the system benefit of IPTV.

Description

A time-sharing recommendation system and method applied to IPTV

The present invention relates to a recommendation technology for audio-visual programs, in particular to a time-sharing recommendation system and method applied to IPTV.

Many video and audio streaming platforms have appeared in the current market, such as Netflix, Youtube, Spotify, KKBOX, etc., and these video and audio streaming platforms will recommend videos or music that users may be interested in through various methods to provide User-specific audio-visual services, thereby increasing the user’s reliance on specific platforms and commercial benefits.

For example, a user A likes to watch action movies or albums on an audio-visual streaming platform A. Therefore, the audio-visual streaming platform A collects the types of movies that user A frequently watches, and further recommends more action movies or albums to User A to meet the needs of user A.

However, the current technology usually has two shortcomings. (1) These audio-visual streaming platforms use accounts created by each user to recommend the account’s usage habits. Therefore, they cannot be applied to network protocols. Television (Internet Protocol Television, IPTV), that is, when people watch TV programs together, the existing recommendation method cannot make recommendations without the usage habits of specific users. Therefore, it is easy to have a recommendation list that cannot match all users; (2) ) Today’s video streaming platforms are Taking the user as the starting point, that is, the video or music recommended by the video streaming platform is based on the user experience as the first consideration. However, for the IPTV platform industry, this type of exclusive user recommendation method cannot reflect the revenue in time In other words, IPTV platform operators cannot effectively recommend other types of programs or movies to users, and thus cannot increase revenue.

Therefore, how to provide a recommendation method suitable for IPTV, which can promptly reflect the problem of the income of the IPTV platform industry, and how to improve the system efficiency of IPTV is one of the issues that needs to be solved urgently.

The present invention provides a time-sharing recommendation system applied to IPTV, which includes: a processing module and a calculation module, the processing module is used to obtain a first time-sharing recommendation list within a time range and establish a share Table for the processing module to obtain a table of the number of times a user clicks each IPTV program in the first time-sharing recommendation list, and the calculation module is based on the first time-sharing recommendation list and the profit sharing table And the click count table calculates a weighted value to generate a second time-sharing recommendation list based on the weighted value, wherein the processing module continuously updates the user clicking each IPTV program in the second time-sharing recommendation list Click the number of times table for the calculation module to continuously update the second time-sharing recommendation list according to the profit sharing table and the updated click number table, so as to provide the updated second time-sharing recommendation list to the user .

The present invention provides a time-sharing recommendation method applied to IPTV, which includes: obtaining a first time-sharing recommendation list within a time range and creating a profit table; obtaining a user clicking each of the first time-sharing recommendation list A click count table of an IPTV program; a weighted value is calculated according to the first time-sharing recommendation list, the share table, and the click count table to generate a second time-sharing recommendation based on the weighted value List; continuously update the click count table of each IPTV program in the second time-sharing recommendation list by the user; and continuously update the second time-sharing recommendation list according to the profit sharing table and the updated click count table To provide the updated second time-sharing recommendation list to the user.

In one embodiment, it further includes a playback module, which transmits each IPTV program in the second time-sharing recommendation list to a playback device for display, playback, and subscription or viewing by the user to generate clicks, and then The click count is transmitted to the playback module to update the click count table, wherein the processing module obtains the click count table from the playback module.

In one embodiment, the processing module obtains the first time-sharing recommendation list within the time range from a database, and the first time-sharing recommendation list has the IPTV programs and their corresponding presets Score; the profit sharing table has each IPTV program and its corresponding revenue; and the click count table has each IPTV program and its corresponding click count. In addition, the processing module has a preset dynamic parameter.

In an embodiment, the calculation module uses the preset score of each IPTV program, the number of clicks in the click count table, the revenue, and the dynamic parameter to perform calculations on the first time-sharing recommendation list to obtain each IPTV The weighted value of the program, wherein the calculation module increases the weighted value of the IPTV program with higher revenue in the profit sharing table according to the dynamic parameter, and reduces the weighted value of the IPTV program with higher clicks in the click table, In this way, the calculation module reorders the recommended rankings of the IPTV programs through the weighted value to generate the second time-sharing recommendation list.

In one embodiment, the processing module continuously updates the click count table of each IPTV program in the second time-sharing recommendation list by the user, and the calculation module is based on the profit sharing table and the updated value The number of clicks in the click count table performs deep learning on the dynamic parameter to obtain a new dynamic parameter, and the calculation module uses the new dynamic parameter, the preset score, and the updated click count table. The number of clicks and the revenue are recalculated for the weighted value of each IPTV program to continuously update the second time-sharing recommendation list.

In summary, the time-sharing recommendation system and method of the present invention applied to IPTV replace the recommended list of exclusive users by time-sharing. Therefore, the time-sharing is used to cut IPTV programs according to time so that the recommendation list can be applied to Different users in the same home, and have trust. In addition, based on the user’s original preferences, the recommended list is adjusted using the revenue obtained by the IPTV platform industry and the number of user clicks on each IPTV program, thereby increasing the revenue of the IPTV platform industry and the recommended IPTV programs are also in line with the user’s needs. Preferences, thereby enhancing the system benefits of IPTV.

1: A time-sharing recommendation system applied to IPTV

10: Processing module

11: Calculation module

12: Database

13: Play module

S21~S25: steps

Figure 1 is a schematic diagram of a time-sharing recommendation system applied to IPTV according to the present invention; and

Figure 2 is a flowchart of a time-sharing recommendation method applied to IPTV according to the present invention.

It should be noted that the structures, proportions, sizes, etc. shown in the drawings in this manual are only used to match the contents disclosed in the manual for the understanding and reading of those familiar with the art, and are not intended to limit the implementation of the present invention Therefore, it does not have any technical significance. Any structural modification, proportional relationship change, or size adjustment should still fall within the scope of the present invention without affecting the effects and objectives that can be achieved. The technical content disclosed by the invention can be covered. At the same time, the terms such as "on", "first", "second" and "one" cited in this manual are only for ease of description and are not used to limit the implementation of the present invention. The change or adjustment of the scope and the relative relationship shall be regarded as the scope within which the present invention can be implemented without substantially changing the technical content.

Figure 1 is a schematic diagram of a time-sharing recommendation system 1 applied to IPTV of the present invention, which includes: a processing module 10, a calculation module 11, a database 12, and a playback module 13, and these The module is set in the computer, host, server (such as remote server, cloud server, network server), etc.

Specifically, the processing module 10 obtains a first time-sharing recommendation list within the time range from the database 12 according to a set time range (for example, 6:00 pm to 8:00 pm), and the first time-sharing recommendation The list has each IPTV program and its corresponding preset score, and the processing module 10 establishes a profit sharing table of IPTV programs, and the profit sharing table has each IPTV program and its corresponding revenue, and the processing module 10 obtains a user Click a click count table of each IPTV program in the first time-sharing recommendation list, and the click count table has each IPTV program and its corresponding click count.

Furthermore, the calculation module 11 uses the preset scores of each IPTV program, the click count table, the profit and a dynamic parameter preset by the processing module 10 to calculate the first time-sharing recommendation list to obtain the weighted value of each IPTV program. Among them, the calculation module 11 increases the weighted value of the IPTV program with a higher profit in the profit sharing table according to the dynamic parameters, and reduces the weighted value of the IPTV program with a higher number of clicks in the click table, whereby the calculation module 11 passes The weighted value reorders the recommended rankings of each IPTV program to generate a second time-sharing recommendation list. And the playback module 13 transmits each IPTV program in the second time-sharing recommendation list to a playback device (such as a TV, a smart phone, a tablet, etc.) for display and playback, and provides users with subscription or viewing to generate clicks , Thereby enhancing the exposure of IPTV programs that have not been subscribed or watched, and increasing the revenue of IPTV platform operators.

In addition, in order to avoid that the recommended IPTV program is not the user’s favorite IPTV program, which may cause the user to reduce their willingness to watch the IPTV program, the processing module 10 continues to update the user’s second time-sharing push For the click count table of each IPTV program in the recommended list, the calculation module 11 performs deep learning on the dynamic parameters based on the revenue of the profit sharing table and the click count in the updated click count table to obtain a new dynamic parameter. Group 11 recalculates the weighted value of each IPTV program with new dynamic parameters, preset scores, updated clicks table, and revenue, to update the second time-sharing recommendation list, and provide users with preferences and access to IPTV platform operators Recommended list of higher-yield IPTV programs.

In one embodiment, the processing module 10 obtains the click count table from the playback module 13, wherein after the click count is generated, the click count is transmitted to the playback module 13 to update the click count table.

In one embodiment, the calculation module 11 performs deep learning on the dynamic parameters based on the profit of the profit sharing table and the number of clicks in the updated clicks table to obtain a new dynamic parameter. Among them, the deep learning used by the calculation module 11 Methods such as deep neural network, convolutional neural network, deep belief network and loop neural network, and the calculation module 11 will continue to use the click count and share table of the click count table generated by the user clicking each IPTV program In order to continue deep learning, we can continuously update dynamic parameters and adjust the second time-sharing recommendation list.

In one embodiment, the calculation module 11 calculates the weighted value of each IPTV program through formula (1), where the dynamic parameter or new dynamic parameter includes a first dynamic parameter (a) and a second dynamic parameter ( b). The click count table contains the click count of each IPTV program, and is based on the preset score, click count, revenue, first dynamic parameter (a) and second dynamic parameter (b) of each IPTV program through formula (1) Calculate to generate the weighted value of each IPTV program, the formula (1) is as follows:

Weighted value = a×revenue-b×number of clicks+preset score (1)

Figure 2 is a flowchart of a time-sharing recommendation method applied to IPTV according to the present invention, and please refer to Figure 1 for description.

In step S21, the processing module 10 obtains a first time-sharing recommendation list within a time range and creates a profit sharing table.

In step S22, the processing module 10 obtains a click count table for a user to click each IPTV program in the first time-sharing recommendation list.

In step S23, the calculation module 11 calculates a weighted value according to the first time-sharing recommendation list, the share table, and the click count table to generate a second time-sharing recommendation list according to the weighted value.

In step S24, the processing module 10 continuously updates the user's click count table of each IPTV program in the second time-sharing recommendation list.

In step S25, the calculation module 11 continuously updates the second time-sharing recommendation list according to the profit sharing table and the updated click count table, thereby providing an updated updated list that meets user needs and that IPTV platform operators can also obtain higher revenue The second time-sharing recommendation list to users.

In addition, the following is an embodiment of the time-sharing recommendation system 1 applied to IPTV of the present invention, and please refer to FIGS. 1 and 2 for description. When a user watches an IPTV program provided by an IPTV platform provider on TV from 6:00 pm to 8:00 pm on the same day, a processing module 10 obtains the one from a database 12 from 6:00 pm to 8:00 pm on that day. The first time-sharing recommendation list, and the first time-sharing recommendation list has each IPTV program and its corresponding preset score, to recommend the corresponding IPTV program to the user according to the level of the preset score, and the processing module 10 creates the IPTV program One-point profit table, and the profit-sharing table includes each IPTV program and its corresponding income. In addition, the processing module 10 obtains a click count table of each IPTV program in the first time-sharing recommendation list by the user through the playback module 13 , Among them, the first time-sharing recommendation list, profit sharing table and click count table are shown in Tables 1~3.

Table 1: The first time-sharing recommendation list

Table 2: Profit Sharing Table

Table 3: Clicks table

Then, for recommending users to subscribe or watch other unclicked IPTV programs to increase the revenue of the IPTV platform industry, the calculation module 11 is based on the preset scores of each IPTV program, the number of clicks in the click table, the revenue, and the preset The first dynamic parameter (a=1) and the second dynamic parameter (b=3) are The time-sharing recommendation list calculates the weighted value to generate the second time-sharing recommendation list. Therefore, the calculation module 11 uses the formula (1) (weighted value=a×revenue-b×number of clicks+preset score) to calculate Obtain the weighted value of each IPTV program, and sort the IPTV programs from large to small according to the weighted value to recommend other unsubscribed or watched IPTV programs with higher revenue to users. For example, the weighted value of program A is 1. ×20-3×3+10=21, the weighted value of program B is 1×20-3×5+9=14, etc. Therefore, the second value generated by program A~program G after calculating the weighted value The time-sharing recommendation list is shown in Table 4.

Table 4: Second time sharing recommendation list

As shown in Table 4, the calculation module 11 uses the first dynamic parameter and the second dynamic parameter to increase the profit of the profit sharing table and reduce the weighted value of the IPTV program with the higher number of clicks in the click table, thereby playing The module 13 transmits other unclicked IPTV programs and provides them on the TV to recommend to users to subscribe or watch, thereby increasing the exposure of unsubscribed or watched IPTV programs, and increasing the revenue of the IPTV platform industry.

In addition, in order to make the recommended IPTV programs in line with the user’s preferences, so as to avoid reducing the user’s willingness to watch IPTV programs, the processing module 10 continuously updates the user’s clicks on each IPTV program in the second time-sharing recommendation list from the playback module 13 The frequency table is shown in Table 5.

Table 5: Clicks table after update

The calculation module 11 performs deep learning on the first dynamic parameter and the second dynamic parameter according to the profit of the profit sharing table and the number of clicks in the updated click frequency table to obtain the new first dynamic parameter (a=1) and the new second dynamic parameter Parameter (b=2), and the calculation module 11 reuses the new first dynamic parameter (a=1), the new second dynamic parameter (b=2), the preset score, the number of clicks in the updated click table and Revenue recalculates the weighted value of each IPTV program to update the second time-sharing recommendation list, thereby providing a recommended list of IPTV programs that meet user preferences and IPTV platform operators can also obtain higher revenue. The updated second time-sharing The recommended list is shown in Table 6.

Table 6: Updated second time-sharing recommendation list

Furthermore, the processing module 10 in the time-sharing recommendation system 1 applied to the IPTV of the present invention will continuously obtain the clicks of each IPTV program in the second time-sharing recommendation list by the user's clicks to update the clicks table, and then through The calculation module 11 continuously performs deep learning to generate new first dynamic parameters And new second dynamic parameters, and then continuously provide the updated second time-sharing recommendation list, and make the updated second time-sharing recommendation list meet the user's preferences and increase the income of the IPTV platform industry at the same time.

On the other hand, when the processing module 10 obtains the click count table of each IPTV program in the first time-sharing recommendation list by the user through the play module 13 for the first time, the click count of each IPTV program in the click count table is 0. , Then, the calculation module 11 generates a second time-sharing recommendation list based on the preset scores of each IPTV program, the number of clicks in the click table (the number of clicks is 0), revenue, and dynamic parameters, which are then provided by the playback module 13 The user subscribes or watches the second time-sharing recommendation list. Therefore, the playback module 13 only provides the second time-sharing recommendation list calculated by the calculation module 11 for the user to subscribe or watch. In addition, the processing module 10 will base the click count on the list Adjust the preset score of the first time-sharing recommendation list to sort the IPTV programs.

In summary, the time-sharing recommendation system and method of the present invention applied to IPTV replace the recommended list of exclusive users by time-sharing. Therefore, the time-sharing is used to cut IPTV programs according to time so that the recommendation list can be applied to Different users in the same home, and have a degree of trust. In addition, based on the user’s original favorite recommendation list, the income obtained by the IPTV platform industry and the number of user clicks on each IPTV program are used to adjust the recommendation list, thereby improving the IPTV platform industry The profitable and recommended IPTV programs are also in line with users' preferences, so as to enhance the system benefits of IPTV.

Furthermore, a time-sharing recommendation system and method applied to IPTV of the present invention can have the following advantages or technical effects.

1. The present invention time-shares IPTV programs according to time, so that different recommended lists can be provided at different time periods to meet the needs of watching programs of different users in the same home, and further achieve that the recommended program list is not limited to specific User effect.

2. The present invention uses deep learning technology to adjust dynamic parameters according to the number of times the user clicks on the IPTV program and the income of the IPTV platform operator, thereby providing a recommendation list that meets the user's viewing preferences and improves the income of the IPTV platform operator.

1: A time-sharing recommendation system applied to IPTV

10: Processing module

11: Calculation module

12: Database

13: Play module

Claims (10)

A time-sharing recommendation system applied to IPTV, including: A processing module is used to obtain a first time-sharing recommendation list within a time range and create a profit sharing table for the processing module to obtain a user clicking each IPTV in the first time-sharing recommendation list A table of the number of clicks of the program; and A calculation module, which calculates a weighted value according to the first time-sharing recommendation list, the profit sharing table, and the click count table, so as to generate a second time-sharing recommendation list according to the weighted value; Wherein, the processing module continuously updates the click count table of each IPTV program in the second time-sharing recommendation list that the user clicks, so that the calculation module continuously based on the share table and the updated click count table The second time-sharing recommendation list is updated to provide the updated second time-sharing recommendation list to the user. For example, the time-sharing recommendation system described in item 1 of the scope of patent application further includes a playback module, which transmits each IPTV program in the second time-sharing recommendation list to a playback device for display and playback, and provides the The user subscribes or watches and generates clicks, and then transmits the clicks to the play module to update the clicks table. The processing module obtains the clicks table from the play module. For example, the time-sharing recommendation system described in item 1 of the scope of patent application, wherein the processing module obtains the first time-sharing recommendation list within the time range from a database, and the first time-sharing recommendation list is Has the IPTV programs and their corresponding preset scores, and the profit sharing table has the IPTV programs and their corresponding revenues, and the click count table has the IPTV programs and their corresponding click counts, and the The processing module has a preset dynamic parameter. For example, the time-sharing recommendation system described in item 3 of the scope of patent application, wherein the calculation module uses the preset score of each IPTV program, the number of clicks in the click count table, the revenue and the dynamic parameter The first time-sharing recommendation list is calculated to obtain the weighted value of each IPTV program, wherein the calculation module increases the weighted value of the IPTV program with the higher profit of the share table according to the dynamic parameter, and reduces The weighted value of the IPTV program with the higher number of clicks in the click count table, so that the calculation module reorders the recommended ranking of each IPTV program through the weighted value to generate the second time-sharing recommendation list. For example, the time-sharing recommendation system described in item 3 of the scope of patent application, wherein the processing module continuously updates the table of the number of clicks that the user clicks on each IPTV program in the second time-sharing recommendation list, and the calculation module is based on it The profit of the profit sharing table and the number of clicks in the updated clicks table perform deep learning on the dynamic parameters to obtain a new dynamic parameter, and then the calculation module uses the new dynamic parameter, the preset score, and the update The number of clicks in the subsequent clicks table and the weighted value of the revenue for each IPTV program are recalculated to continuously update the second time-sharing recommendation list. A time-sharing recommendation method applied to IPTV includes: Obtain a first time-sharing recommendation list within a time range and establish a profit sharing table; Obtain a click count table for a user to click each IPTV program in the first time-sharing recommendation list; Calculating a weighted value according to the first time-sharing recommendation list, the profit table and the click count table, so as to generate a second time-sharing recommendation list according to the weighting value; Continuously update the click count table of each IPTV program in the second time-sharing recommendation list by the user; and The second time-sharing recommendation list is continuously updated according to the profit sharing table and the updated click count table to provide the updated second time-sharing recommendation list to the user. For example, the time-sharing recommendation method described in item 6 of the scope of patent application, wherein each IPTV program in the second time-sharing recommendation list is transmitted to a playback device for display and playback, and the user is provided with subscription or viewing The click times are generated to update and obtain the click times table. For example, the time-sharing recommendation method described in item 6 of the scope of patent application, wherein the first time-sharing recommendation list within the time range is obtained from a database, and the first time-sharing recommendation list has the IPTV programs And its corresponding preset score, and the profit sharing table has each IPTV program and its corresponding revenue, and the click count table has each IPTV program and its corresponding click count, and a dynamic parameter is preset . For example, the time-sharing recommendation method described in item 8 of the scope of patent application, wherein the preset score of each IPTV program, the number of clicks in the click count table, the revenue and the dynamic parameter are used for the first time-sharing recommendation list Perform calculations to obtain the weighted value of each IPTV program, wherein according to the dynamic parameter, increase the weighted value of the IPTV program with higher revenue in the profit sharing table, and decrease the weighted value of the IPTV program with higher clicks in the click table The weighted value is used to re-sort the recommended rankings of the IPTV programs through the weighted value to generate the second time-sharing recommendation list. For example, the time-sharing recommendation method described in item 8 of the scope of patent application, wherein the user clicks on each IPTV program in the second time-sharing recommendation list continuously updated according to the profit sharing table and the update The number of clicks in the subsequent clicks table performs deep learning on the dynamic parameter to obtain a new dynamic parameter, and then uses the new dynamic parameter, the preset score, the number of clicks in the updated clicks table, and the revenue The weighted value of the IPTV program is recalculated to continuously update the second time-sharing recommendation list.

Images