TotalPatent Semantic Search

[tweetmeme source=”Intellogist” only_single=false]
After recently gaining more experience and insight into LexisNexis’ patent search system TotalPatent, the Semantic Search feature stands out as one of its best. TotalPatent implements Semantic Search, a latent semantic analysis system, in a particularly open and interesting way.


Unlike some competitors that maintain a “black box” approach to the various algorithms behind latent semantic analysis, LexisNexis has been more open about its approach. When running a Semantic Search, the system first decides which dataset to generate keywords from. Each dataset has a corresponding “brain” that has been built from a pool of about 6 million documents. These source documents include full text patent documents and Elsevier journal articles: a major source of non-patent technical literature that helps make the system more robust. It is this “brain” that analyzes the text input by the user and produces a “query cloud” of search terms. These search terms may be individual words or multiword phrases.

As part of the analysis phase, TotalPatent also assigns weights to each term that conveys the importance of said term in the ensuing search. Terms may be weighted from 4 (mandatory inclusion in the ensuing search) to -1 (mandatory exclusion). Users have the ability to manually edit these weights prior to enacting the search; a nice feature that allows users to add a touch of human intelligence to the traditionally computer dominated technique of latent semantic analysis. (In another example of user weighting controls, FreePatentsOnline allows users to weight search terms in a non semantic search, a slightly different take on the same idea.)

The fact that the process is open and able to be documented is one of the biggest strengths of TotalPatent’s latent semantic analysis and search. Patent searchers can inform their clients of exactly what has been searched, including the assigned term weightings.

Semantic Search through TotalPatent is free for new subscribers; pricing clocks in at $40 per search for users with outdated subscriptions. Contact your LexisNexis representative for details on your subscription.

For further details and example screenshots, see the Intellogist TotalPatent Report.

Have you used TotalPatent Semantic Search? Did it live up to your needs and expectations? Let us know in the comments below.

Patent Searches from Landon IP

This post was contributed by Intellogist team member Chris Jagalla.

9 Responses

  1. Is it financially viable for a singe user to purchase a license? how much, approximately, is the cost?

  2. Social comments and analytics for this post…

    This post was mentioned on Twitter by priorsmart: “TotalPatent Semantic Search” http://u.nu/5ddk8 (at @intellogist) #patent…

  3. An interesting observation of this tool. A phrase that took my attention is: “Patent searchers can inform their clients of exactly what has been searched, including the assigned term weightings.” By sharing this weightings you don’t inform about what you searched but how you searched and that is something a professional searcher should not do to my opinion. Sharing the how of a search is sharing your responsibility.

  4. Communicating exactly what has been searched is important (in most cases) to the client, so they know that the search has been exhaustive. Throwing search terms into a black box is not ideal, since the client is not shown the extent and professionalism of the search. Transparency is important, in my opinion.

  5. […] is available include FreePatentsOnline, SumoBrain, and Delphion. As seen in last week’s TotalPatent Semantic Search post, keyword weighting can also be used in patent analysis tools to narrow the focus on certain areas […]

  6. […] touched on TotalPatent’s Semantic Search in one of our recent posts, but there is much more to investigate in our recent TotalPatent full report review and update. We […]

  7. […] tweetmeme_source = '”Intellogist”'; Semantic searching is nothing new: many patent (like TotalPatent) and non-patent literature (like ProQuest Dialog) search systems provide some form of semantic […]

  8. […] now identifies multiple concepts in a search query.  TotalPatent’s Semantic Search feature impressed the Intellogist writers who first tested it, since it allows users to re-rank terms and customize the semantic query before conducting the […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: