System and method for ranking search results using social information
An improved system and method for ranking search results using social information is provided. A social information ranking engine may be provided to rerank a list of URLs of web pages in search results from a search engine using social information. To do so, the search query may be applied to a social information domain and a second list of references to web pages may be obtained as search results. A dimensional vector of users with the plurality of terms may be constructed using search results from social information domain, and a similarity measure may be computed between the references to web pages obtained from search results by the search engine and the references to web pages obtained from search results by the social information ranking engine. The score from the similarity measure may then be used to rerank the list of references to web pages.
1. A computer system for providing search results for a query, comprising:
a search engine that receives a search query and obtains a list of URLs of web pages as search results;
a social information ranking engine operably coupled to the search engine to rank the list of URLs of web pages using social information;
wherein the search engine receives from a client device a social information domain to be used to rerank the search results using the social information;
wherein the search engine reranks the search results using the social information;
wherein reranking the search results using social information further comprises generating a bipartite graph with a plurality of users connected by weighted edges to a plurality of terms, wherein the edges are weighted by a number of times the user referenced the term, and generating a dimensional vector of a plurality of users with a plurality of terms from the bipartite graph with the plurality of users connected by weighted edges to the plurality of terms, wherein the edges are weighted by a number of times the user referenced the term; and
a web page storage operably coupled to the social information ranking engine that stores a network of users associated with a plurality of terms.
2. The system of claim 1 further comprising a web browser operably coupled to the search engine to display the list of URLs of web pages ranked using social information.
3. A computer-implemented method for providing search results for a query, comprising:
receiving a search query to obtain a plurality of references to web pages as search results;
obtaining the plurality of references to web pages as search results;
receiving from a client device a social information domain to be used to rerank the plurality of references to web pages as search results using social information from the social information domain;
reranking the plurality of references to web pages using social information;
wherein reranking the plurality of references to web pages using social information further comprises generating a bipartite graph with a plurality of users connected by weighted edges to a plurality of terms, wherein the edges are weighted by a number of times the user referenced the term, and generating a dimensional vector of a plurality of users with a plurality of terms from the bipartite graph with the plurality of users connected by weighted edges to the plurality of terms, wherein the edges are weighted by a number of times the user referenced the term; and
sending the reranked plurality of references to web pages as search results to a client device for display to a user.
4. The method of claim 3 wherein the social information domain comprises a social network website.
5. The method of claim 3 wherein reranking the plurality of references to web pages using social information further comprises computing a similarity between the plurality of references to web pages with a plurality of terms and a plurality of users with a plurality of terms.
6. The method of claim 5 wherein computing a similarity between the plurality of references to web pages with a plurality of terms and a plurality of users with a plurality of terms comprises computing a dot product of a dimensional vector representing the plurality of references to web pages with the plurality of terms and a dimensional vector representing the plurality of users with the plurality of terms.
7. The method of claim 3 wherein reranking the plurality of references to web pages using social information further comprises obtaining a second plurality of references to a second plurality of web pages as search results from a social network website.
8. The method of claim 7 further comprising constructing a dimensional vector of a plurality of users with a plurality of terms from the second plurality of references to the second plurality of web pages as search results from the social network website.
9. The method of claim 8 further comprising computing a dot product of a dimensional vector representing a plurality of references to web pages with a plurality of terms and the dimensional vector representing the plurality of users with the plurality of terms from the second plurality of reference to the second plurality of web pages.
10. A non-transitory computer-readable storage medium having computer-executable instructions for performing the method of claim 3 .
11. A non-transitory computer-readable storage medium having computer-executable instructions for performing the steps of:
receiving a search query to obtain a plurality of references to web pages as search results; obtaining the plurality of references to web pages as search results;
obtaining from a client device a second plurality of references to a second plurality of web pages as search results from a social information domain;
reranking the plurality of references to web pages using the second plurality of references to the second plurality of web pages as search results from the social information domain;
wherein reranking the plurality of references to web pages using the second plurality of references to the second plurality of web pages as search results from the social information domain further comprises generating a bipartite graph with a plurality of users connected by weighted edges to a plurality of terms, wherein the edges are weighted by a number of times the user referenced the term, and generating a dimensional vector of a plurality of users with a plurality of terms from the bipartite graph with the plurality of users connected by weighted edges to the plurality of terms, wherein the edges are weighted by a number of times the user referenced the term; and
sending the reranked plurality of references to web pages as search results to a client device for display to a user.
12. The non-transitory computer-readable storage medium of claim 11 further comprising computer-executable instructions for performing the step of computing a dot product of a dimensional vector representing a plurality of references to web pages with a plurality of terms and the dimensional vector representing the plurality of users with the plurality of terms.
13. The non-transitory computer-readable storage medium of claim 11 further comprising computer-executable instructions for performing the step of receiving a reference to the social information domain for obtaining the second plurality of references to the second plurality of web pages as search results.
14. The non-transitory computer-readable storage medium of claim 11 wherein the computer-executable instructions for performing the step of obtaining the second plurality of references to the second plurality of web pages as search results from the social information domain comprises computer-executable instructions for performing the step of obtaining the second plurality of references to the second plurality of web pages as search results from a social network website.