System, method and computer-readable medium for recommendation
The present disclosure relates to a system, a method and a computer-readable medium for recommendation. The method includes providing a first content according to a first recommendation logic to a user terminal of a viewer; providing a second content according to a second recommendation logic to the user terminal; obtaining interaction data from the user terminal; and adjusting an allocation of contents from the first recommendation logic and the second recommendation logic to be shown on the user terminal according to the interaction data.
1 . A method for recommendation, executed by a server, comprising:
providing a first content according to a first recommendation logic to a user terminal of a viewer;
providing a second content according to a second recommendation logic to the user terminal, wherein the second recommendation logic is different from the first recommendation logic, and the first content and the second content are displayed at the user terminal of the viewer at the same time;
obtaining interaction data from the user terminal;
calculating respectively a click rate and an average time length with respect to the first recommendation logic and the second recommendation logic based on the interaction data; and
adjusting an allocation of contents from the first recommendation logic and the second recommendation logic to be shown on the user terminal according to the interaction data, wherein adjusting the allocation of contents comprises displaying more contents recommended from the first recommendation logic or displaying contents recommended from the first recommendation logic in a higher order, compared with contents recommended from the second recommendation logic.
2 . The method according to claim 1 , further comprising:
determining a first recommendation score for the first content according to content attribute data of the first content and viewer attribute data of the viewer, by the first recommendation logic;
determining the first recommendation score to be higher than a first threshold;
determining a second recommendation score for the second content according to content attribute data of the second content and the viewer attribute data of the viewer, by the second recommendation logic; and
determining the second recommendation score to be higher than a second threshold.
3 . The method according to claim 2 , further comprising:
determining a third recommendation score for the first content according to the content attribute data of the first content and the viewer attribute data of the viewer, by the second recommendation logic;
determining the third recommendation score to be lower than a third threshold;
determining a fourth recommendation score for the second content according to the content attribute data of the second content and the viewer attribute data of the viewer, by the first recommendation logic; and
determining the fourth recommendation score to be lower than a fourth threshold.
4 . The method according to claim 2 , further comprising:
determining a first mixed score for the first content according to the first recommendation score and the interaction data;
determining a second mixed score for the second content according to the second recommendation score and the interaction data; and
adjusting the allocation of contents from the first recommendation logic and the second recommendation logic according to the first mixed score and the second mixed score.
5 . The method according to claim 1 , wherein the adjusting the allocation of contents includes adjusting an order of contents from the first recommendation logic and the second recommendation logic to be shown on the user terminal.
6 . The method according to claim 1 , wherein the adjusting the allocation of contents includes adjusting respective numbers of contents from the first recommendation logic and the second recommendation logic to be shown on the user terminal.
7 . The method according to claim 1 , further comprising:
providing a first group of contents according to the first recommendation logic to the user terminal;
providing a second group of contents according to the second recommendation logic to the user terminal;
determining the viewer to have interacted more with the first group of contents according to the interaction data; and
adjusting the allocation of contents to display more contents from the first recommendation logic on the user terminal.
8 . The method according to claim 1 , wherein the interaction data includes click rates or retention lengths with respect to contents from the first recommendation logic and contents from the second recommendation logic.
9 . A system for recommendation, comprising one or a plurality of processors, wherein the one or plurality of processors execute a machine-readable instruction to perform:
providing a first content according to a first recommendation logic to a user terminal of a viewer;
providing a second content according to a second recommendation logic to the user terminal, wherein the second recommendation logic is different from the first recommendation logic, and first content and the second content are displayed at the user terminal of the viewer at the same time;
obtaining interaction data from the user terminal;
calculating respectively a click rate and an average time length with respect to the first recommendation logic and the second recommendation logic based on the interaction data; and
adjusting an allocation of contents from the first recommendation logic and the second recommendation logic to be shown on the user terminal according to the interaction data, wherein the adjusting the allocation of contents comprises displaying more contents recommended from the first recommendation logic or displaying contents recommended from the first recommendation logic in a higher order, compared with contents recommended from the second recommendation logic.
10 . A non-transitory computer-readable medium including a program for recommendation, wherein the program causes one or a plurality of computers to execute:
providing a first content according to a first recommendation logic to a user terminal of a viewer;
providing a second content according to a second recommendation logic to the user terminal, wherein the second recommendation logic is different from the first recommendation logic, and the first content and the second content are displayed at the user terminal of the viewer at the same time;
obtaining interaction data from the user terminal;
calculating respectively a click rate and an average time length with respect to the first recommendation logic and the second recommendation logic based on the interaction data; and
adjusting an allocation of contents from the first recommendation logic and the second recommendation logic to be shown on the user terminal according to the interaction data, wherein the adjusting the allocation of contents comprises displaying more contents recommended from the first recommendation logic or displaying contents recommended from the first recommendation logic in a higher order, compared with contents recommended from the second recommendation logic.