When a player like Apple co-founder Steve Wozniak joins the advisory board of DeepDyve, it sends a signal to the Internet community that this must be something “hot.”  The “Woz” (why do some tech leaders get names like this – especially from Apple – “Tog,” etc.?) saw the possibilities of this search engine that crawls what is known as “deep web,” or what is not indexed by search engines.  We often think when we use Google, we are tapping into a lot of material.  However, what is so fascinating is that according to a study by the University of California, Berkeley (http://www.physorg.com/news140110066.html), traditional search engines only index .2% of the Internet.  .2%!  That means that 99.8% is not indexed and resides invisibly in the abyss of the “dark web.”  The goal of DeepDyve is to mine the data in the deep web and make it accessible to research.

DeepDyve was founded by two scientists from the Human Genome Project (designing a system for pattern-matching data).  Chris Sherman from Search Engine Land (http://searchengineland.com/deepdyve-explores-the-invisible-web-15417)  comments enthusiastically on DeepDyve’s “genomic researcher” bent (vs. computer scientists and linguists that base search strategies on text and keywords) on setting up “sequence” search strategies, such as indexing words but also computing the “factorial combination of worlds and phrases in the document and uses some industrial strength statistical techniques to assess the ‘information impact’ of these combinations.”  DeepDyve primarily covers scientific, technical and medical research.  With its KeyPhrasetechnology, one can enter up to 25,000 characters – meaning one can also cut and past paragraphs to search for.  A big boon – a user does not have to rely on just an abstract or title to ascertain whether the article is fitting or not.  DeepDyve allows the user to read the full article before buying it.

The DeepDyve article rental plan sounds a lot like Netflix but with a twist – rent viewing time and if so desired, buy the article.  The plans include:

  • One time rental cost of 24 hour article viewing: .99/per article
  • Silver monthly plan: 9.99 for 20 articles to read for 7 days
  • Gold monthly plan: 19.99 for unlimited articles for unlimited time

A problem for researchers with DeepDyve is that in reading the article in its Flash-based proprietary viewer, users cannot capture that information electronically or through print.  If users want to copy any material, it’s down to writing on paper.  Another problem is that publishers that have signed up veer more toward the scholarly societies over for-profit publishers.  What a surprise – Elsevier is not a player.  Certainly, DeepDyve can augment research by providing relatively inexpensive access to those devoid of the free richness of databases and electronic journals found on university campuses.  That is often the shocker when leaving a university – suddenly devoid of those lovely journal and database resources.  Not even alumni get a piece.  However, one alumni association has taken it on.  The Caltech (California Institute of Technology) Alumni Association recently hooked up with DeepDyve (http://www.deepdyve.com/corp/about/press/20100222) to provide their alumni discounted DeepDyve memberships.  Now, that’s a start.