A social network-based meta search engine

Social networks as an application of social web have attracted many users and have became one of the most important services of Internet. Interaction of users in a social network generates huge amount of data. In this paper, a novel meta-search engine is proposed to exploit social network data to improve web search results. The system modifies meta-search engine's multi-agent based architecture by adding new agents to gather interaction data of users and process them to create user profiles based on the previous researches. These profiles are used to re-rank top search results of a web search engine and increase effectiveness of retrieval. Normalized Discounted Cumulative Gain (NDCG) measure is used to evaluate our system. Experimental results show the potential usefulness of social network data for improvement of web search effectiveness.