Search ranking method and system for community users

摘要

Disclosed is a search ranking method for community users. The method includes: calculating a pre-ranking factor and an offline ranking factor according to historical behavior data of users; performing weighted ranking on inverted indices of the users by taking the pre-ranking factor as a weight, to obtain orderly inverted user index data; and with respect to a logged-in search user, in search results obtained according to the index data, performing weighted calculation on the search results according to the offline ranking factor, to obtain final ranking results. Also disclosed is a search ranking system for community users. The method and system can enable a user to obtain more optimized search ranking results.

主权项

1. A search ranking method for community users, comprising:
calculating, by a search ranking system, a pre-ranking factor and an offline ranking factor according to historical behavior data of users searched from a search for community users according to a keyword inputted by a searching user; performing, by the search ranking system, a first weighted ranking on inverted indices of searched users by taking the pre-ranking factor as a weight, to obtain orderly inverted indices data, wherein the inverted indices of searched users are obtained by from said search for community users; in a case that the searching user is a logged-in search user, performing, by the search ranking system, a second weighted ranking on search results of the first weighted ranking according to the offline ranking factor, to obtain final ranking results for community users; and presenting the final ranking results for community users to the searching user, wherein the pre-ranking factor includes user activeness information calculated according to data on a user level of a user in the community, and data on the user's recent log-ins and usages of the community, and the offline ranking factor includes at least one of
user preference information or user category information obtained by analyzing data on articles posted or read by a user in a community, persons with whom the searching user has communicated or become friends, and persons followed by the searching user;user potential friendship chain information obtained by analyzing data on friends of the searching user in the community and persons followed by the searching user in the community; anduser intimacy degree information calculated according to data on interactivities between the searching user and friends.