Method and apparatus for processing user behavior data

1 . A method for providing ranked search results to a user based on user behavior data, comprising:

acquiring behavior data of the user from a search engine user behavior log, wherein the search engine user behavior log comprises a plurality of entries comprising:

feature data associated with the behavior data;

label data associated with click behavior; and

a time corresponding to the behavior data being generated according to the search engine;

determining a timeliness factor and a periodicity factor corresponding to the behavior data according to a time at which the behavior data is generated and a current time, wherein

the time at which the behavior data is generated and the current time are an i th day and a j th day in a preset time period having N days;

the timeliness factor provides a higher weight to behavior data closer to the current time;

the periodicity factor provides a higher weight to behavior data closer to a start of a time period and farther from the current time, wherein the behavior data repeats during each time period of a same length; and

determining the periodicity factor comprises:

determining a date difference between the i th day on which the behavior data is generated and the j th day; and

generating the periodicity factor according to the date difference;

weighting the behavior data of the time at which the behavior data is generated and the current time to generate a weighted value, wherein the weighted value is based on at least one of the timeliness factor or the periodicity factor;

adjusting the behavior data using the weighted value;

ranking the search results based on the adjusted behavior data; and

providing the ranked search results to the user.

2 . The method for providing ranked search results to a user based on user behavior data according to claim 1 , wherein adjusting the behavior data comprises:

obtaining a product of the timeliness factor and the behavior data as first behavior data; and

generating second behavior data according to the periodicity factor, the behavior data, and the first behavior data.

3 . The method for providing ranked search results to a user based on user behavior data according to claim 2 , wherein generating second behavior data according to the periodicity factor, the behavior data, and the first behavior data comprises:

generating third behavior data according to the periodicity factor and the behavior data; and

generating the second behavior data according to the first behavior data and the third behavior data.

4 . The method for providing ranked search results to a user based on user behavior data according to claim 3 , wherein the first behavior data is added to or multiplied by the third behavior data to generate the second behavior data.

5 . The method for providing ranked search results to a user based on user behavior data according to claim 1 , wherein determining the timeliness factor comprises:

determining a time difference between a date on which the behavior data is generated and a current date; and

determining the timeliness factor according to the time difference and a time attenuation function.

6 . The method for providing ranked search results to a user based on user behavior data according to claim 5 , wherein the time attenuation function comprises at least one of an exponential function and a power function.

7 . The method for providing ranked search results to a user based on user behavior data according to claim 5 , wherein determining the timeliness factor according to the time difference and a time attenuation function further comprising:

weighting the behavior data according to the time attenuation; and

generating a first weighted value to a first time before the current time and a second weighted value to a second time before the first time, wherein the second weighted value is smaller than the first weighted value.

8 . An apparatus for providing ranked search results to a user based on user behavior data, comprising:

a memory storing a set of instructions;

a processor configured to execute the set of instructions to cause the apparatus to perform:

acquiring behavior data of the user from a search engine user behavior log, wherein the search engine user behavior log comprises a plurality of entries comprising:

feature data associated with the behavior data;

label data associated with click behavior; and

a time corresponding to the behavior data being generated according to the search engine;

determining a timeliness factor and a periodicity factor corresponding to the behavior data according to a time at which the behavior data is generated and a current time, wherein

the time at which the behavior data is generated and the current time are an i th day and a j th day in a preset time period having N days;

the timeliness factor provides a higher weight to behavior data closer to the current time;

the periodicity factor provides a higher weight to behavior data closer to a start of a time period and farther from the current time, wherein the behavior data repeats during each time period of a same length; and

determining the periodicity factor comprises:

determining a date difference between the i th day on which the behavior data is generated and the j th day; and

generating the periodicity factor according to the date difference;

weighting the behavior data of the time at which the behavior data is generated and the current time to generate a weighted value, wherein the weighted value is based on at least one of the timeliness factor or the periodicity factor;

adjusting the behavior data using the weighted value;

ranking the search results based on the adjusted behavior data; and

providing the ranked search results to the user.

9 . The apparatus for providing ranked search results to a user based on user behavior data according to claim 8 , wherein adjusting the behavior data comprises:

obtaining a product of the timeliness factor and the behavior data as first behavior data; and

generating second behavior data according to the periodicity factor, the behavior data, and the first behavior data.

10 . The apparatus for providing ranked search results to a user based on user behavior data according to claim 8 , wherein determining the timeliness factor comprises:

determining a time difference between a date on which the behavior data is generated and a current date; and

determining the timeliness factor according to the time difference and a time attenuation function.

11 . A non-transitory computer readable medium that stores a set of instructions that is executable by at least one processor of a computer to cause the computer to perform method for providing ranked search results to a user based on user behavior data, comprising:

acquiring behavior data of the user from a search engine user behavior log, wherein the search engine user behavior log comprises a plurality of entries comprising:

feature data associated with the behavior data;

label data associated with click behavior; and

a time corresponding to the behavior data being generated according to the search engine;

determining a timeliness factor and a periodicity factor corresponding to the behavior data according to a time at which the behavior data is generated and a current time, wherein

the time at which the behavior data is generated and the current time are an i th day and a j th day in a preset time period having N days;

the timeliness factor provides a higher weight to behavior data closer to the current time;

the periodicity factor provides a higher weight to behavior data closer to a start of a time period and farther from the current time, wherein the behavior data repeats during each time period of a same length; and

determining the periodicity factor comprises:

determining a date difference between the i th day on which the behavior data is generated and the j th day; and

generating the periodicity factor according to the date difference;

weighting the behavior data of the time at which the behavior data is generated and the current time to generate a weighted value, wherein the weighted value is based on at least one of the timeliness factor or the periodicity factor;

adjusting the behavior data using the weighted value;

ranking the search results based on the adjusted behavior data; and

providing the ranked search results to the user.

12 . The non-transitory computer readable medium according to claim 11 , wherein adjusting the behavior data comprises:

obtaining a product of the timeliness factor and the behavior data as first behavior data; and

generating second behavior data according to the periodicity factor, the behavior data, and the first behavior data.

13 . The non-transitory computer readable medium according to claim 12 , wherein generating second behavior data according to the periodicity factor, the behavior data, and the first behavior data comprises:

generating third behavior data according to the periodicity factor and the behavior data; and

generating the second behavior data according to the first behavior data and the third behavior data.

14 . The non-transitory computer readable medium according to claim 13 , wherein the first behavior data is added to or multiplied by the third behavior data to generate the second behavior data.

15 . The non-transitory computer readable medium according to claim 11 , wherein determining the timeliness factor comprises:

determining a time difference between a date on which the behavior data is generated and a current date; and

determining the timeliness factor according to the time difference and a time attenuation function.

16 . The non-transitory computer readable medium according to claim 15 , wherein the time attenuation function comprises at least one of an exponential function and a power function.