The overlap among Social Networks Extracted from Major Web Search Engines

Abstract

The performance and functionality of web search engines is one of the popular issues considered by researchers. One of the new research areas is extraction of social networks from website by search engines. This study aimed to evaluate the overlap among three Google, Yahoo, and Bing search engines in extracting social networks. For this purpose, the overlap among search engines results links was investigated. It was expected that the overlap would be very low. Then, the textual content of links was extracted and texts were processed to extract the related words to queries. These words are the nodes of our social networks. By conducting this process for each of the search engines and extracting social networks from search engines, the overlap of these social networks can be evaluated. The findings showed that the percentage of unique links in search engines is 59 percent and the percentage of unique nodes in social networks is 26 percent. The percentage of shared links in two and three search engines is 20 percent and 3 percent and the shared nodes are 37% and 21%, respectively. This indicated higher overlap among search engines in extracting social networks than overlap among search engines in results links