Comparison of Academic Search Engines (Guangda Hu)

1. Microsoft Academic

   Search authors, publications, conferences, journals, organizations in Computer Science. Good at dealing with rankings and relations between all these different features.

   Special Features: a) Draw co-author paths between two given authors. But the database is still not complete. It can't find the link between Paul Erdos and Fraces Yao, or their paper. b) For Conferences & Journals: Besides the basic data, there is a calender to see coming conferences and journals. c) For Publication: Has a page for each paper. Very nice interface, but slow and inconvient to dowload.

2. Google Scholar

   Focus on publications (all subjects). The most complete database and most widely used engine to find papers. Good at finding references and download paper. Besides the most common functions, it can be configured to use the Tsinghua library to download paper, and show a link to bibtex right under every paper to help to write references.

3. ArnetMiner

   Seems to be an author-oriented search engine on Computer Science, though it can also find papers. The most impressive feature is to draw a social graph for a list of authors, from which one can find bosses, advisers and co-authors. As the results do not show many familiar relationships, there might be a problem about incomplete database or even a bug in the social path finding algorithm.

4. BiomedExperts

   This is a website to search and build up relationships with other experts in biomedical research. The engine finds experts by searching the research topic, name or location. It draws the experts' distribution on a world map, and the user can choose a place and then select an expert. One can also find publications of an expert, with co-authors and abstrats.