Social relationship recommendation method

[0022] Example

[0023] The social relationship recommendation method proposed in this embodiment includes the following steps:

[0024] S1: Data collection: Obtain user data, integrate and classify the data, and analyze the data to obtain user interest and social attention;

[0025] S2: Data search: Search for similar data according to the user's interest level described in S1, and perform preliminary screening on the searched data to obtain target user data;

[0026] S3: Data comparison: Compare the filtered data described in S2 with the user's data, and record the comparison results, and then sort the searched target users based on interest similarity;

[0027] S4: Social attention: According to the user's attention information, collect data of target users similar to the user's attention information, and record the data;

[0028] S5: Data comprehensive analysis: Integrate the target user interest and attention obtained in S3 and S4, and conduct a comprehensive analysis to filter the data again;

[0029] S6: Information screening: final screening of the information after screening described in S5 to obtain the final information;

[0030] S7: Information push: compare the final information described in S6 with the user's data again, sort according to the similarity with the user's data, and push the sorted information to the user;

[0031] S8: Information verification: The user sends a verification request to the target user based on the pushed data described in S7. When sending the verification request, the user’s basic data is also pushed to the target user. After the target user’s consent is obtained, the user can contact the target user. Users establish social relationships.

[0032] In this embodiment, in S4, when collecting social attention data, the target user should be collected based on the proportion of mutual friends and the interaction situation. In S1, when collecting user data, it needs to be based on the user's browsing history. And commonly used keywords are collected. In S1, when collecting user social attention data, first collect user comments, forwarding and favorite information, and then process the collected information to eliminate useless duplicate information. In S6 , When screening information, first make the information into a table, compare the table with the user’s data, and filter out data with large differences in data. When collecting data on interaction, the proportion of interaction should be counted, and Sort the statistical data. In S7, when sorting, the similarity is obtained according to the interest and attention, and the similarity is sorted from high to low. In S2, the searched data is initially filtered It is necessary to eliminate repetitive useless and incomplete information. The beneficial effect of the present invention is to filter the collected information multiple times and sort the information to eliminate useless repetitive information to make the collected information more accurate and effective. It can get the target users more accurately and quickly, saving time.