Metasearch engine integrates the search results from multiple sources, and improves recall in the big data environment. Result merging is a key component which will greatly affect the effectiveness of a metasearch engine. Great progress has been made in this area, however few studies have focused on providing personalized results for different users. This paper proposes a personalized method for merging results of metasearch engine according to a variety of factors, including the user interest distribution, the total number of component search engines exploited, the number of the results that each engine returned, the ranking position of document in each search engine and the number of component search engines who returned the document. Compared with the Borda Fuse method and rCombMNZ method, experimental results show that the proposed model performs better on mean average precision, improving the significance of documents which seldom occurred but are important to the user. By considering the distribution of user interest, the method also has ability of providing personalized result for different users. It is feasible to provide the useful search results more effectively.