Free full text scientific articles...New Search Engine -- Brice Sagot, KnowMade, created a custom search engine in order to find free full text scientific articles in PDF Format. This search engine indexes more than 10 million of free references (mostly to journal articles, conference papers and technical reports). All papers indexed in this search engine are available on open access sources (open archives, universities websites and blogs of researchers…) and are also indexed by Google. The aim of this search engine is to group all these papers in a unique website. No one pay-per-view papers are illegally indexed in this search engine.
Many articles have a “Quick View” link. With this useful Google option, your search keyword is underlined in the PDF documents. It also allows word searching within the documents themselves. The only front end filter is a time frame filter that can search back to the most current 10 years.
You can try it at this address: http://goo.gl/248em If you are aware of any open archives -- contact Brice Sagot directly at KnowMade Ltd., 2405 route des Dolines BP65, 06560 Sophia Antipolis, France, Internet: http://www.knowmade.fr Email: firstname.lastname@example.org Téléphone : +33 (0)4 89 89 16 20 Phone: +33 (0) 4 89 89 16 20
In my opinion, it would also be valuable if the searcher can add customized thesaurus terms post search to export, group and further filter documents. The search engine does not yet offer any document clustering capabilities or document grouping/ranking capabilities by supplemental key words, or IPC classes, but this feature can surely add to the usability of the search results and decrease document evaluation time. Another product produced by his group is the KnowMade Patent Watch which offers an additional analytical toolset. It would be nice to offer some of this post processing capability into the full text search. As well as the group document sharing capability available in other KnowMade products. An additional feature that would make this search engine a bit richer - would be to pull out publicly available patent literature references by examiner and or IPC subclass and be able to use the KWIC highlighting feature in the full text search engine to view literature references and link to the full text if available.