Search engines and their public interfaces
Frank McCown, Michael L. Nelson · 2007
Researchers of commercial search engines often collect datausing the application programming interface (API) or by"scraping" results from the web user interface (WUI), butanecdotal evidence suggests the interfaces produce differentresults. We provide the first in-depth quantitative analysisof the results produced by the Google, MSN and Yahoo APIand WUI interfaces. After submitting a variety of queriesto the interfaces for 5 months, we found significant discrepanciesin several categories. Our findings suggest that theAPI indexes are not older, but they are probably smaller for Google and Yahoo. Researchers may use our findings tobetter understand the differences between the interfaces andchoose the best API for their particular types of queries.