Spinque: Patent Search of the Future?

[tweetmeme source=”Intellogist” only_single=false]

The first thing you must be thinking upon seeing the title of this post is: “What’s a Spinque?” I thought the exact same thing when Arjen de Vries initially contacted me to take a look at his upcoming search system.

The short answer: Spinque is a search system that promotes “search by strategy,” a method of parametrization and query processing to allow greater control over the very fundamentals of how a search system functions, making it tailored to your needs as a searcher. Spinque can be utilized with any data set, but patents are one particularly applicable field.

The long answer: You’re going to have to read along as we give you a sneak peak into this interesting search tool that has been in development for two years!

Spinque is a spin-off company of Centrum Wiskunde & Informatica, and it started operations in November 2009. Since then, development has been focused on making a unique and forward thinking system that can best be described as modular and customizable.

The basis of Spinque’s system is is a probabilistic relational database back end, that can process any state of the art ranking/retrieval model, including the well known “best match” BM25 information retrieval approach. With this as the mathematical basis for the rank-centric model, users can construct a search algorithm visually through the use of linked modules and weight different aspects of the search strategy according to their needs. A user can create a custom search strategy that links any document database to search inputs for each field of the document database. To use a patent-centric example, the user could take a document database they had access to, such as DOCDB or Alexandria, and make custom search fields based on any data that those databases contained. The user could then specify the weight and stemming for each combination of keyword and field to be applied to the database. Spinque also allows the search system to be constructed with many simultaneous databases, which can be useful to create a “one stop” system for the searcher.

The result of this system is that there must be one “power user” willing to define custom search strategies to best fit for the future end-users/searchers. The positive trade off for this is a powerful and relevant search system for the end-user, who then does not have to do as much dirty work with complex syntax or weighting functions. My impression upon being shown a live demo of Spinque was that the role of the “power user” controlling the creation of the system was not beyond anyone with search experience–the graphical interface for the back end made clear the relationship between the database, keywords, weighting, inputs, and outputs.

Dynamic linking of modules.

This shows a small portion of the strategy creation interface. Note the linked graphical nature of the modules.

Once the search system has been created, the end-user will see an intuitive interface that has some very nice features. The main feature of Spinque’s search interface is the ever-popular faceted searching approach. Faceted searching is the ability for users to modify sets of results after the query has been performed, drilling down into the records via further limiting criteria. Additionally, users can modify the weighting of (patent) fields on the fly as well as “up or down voting” aspects of the search strategy to immediately re-rank results.

Spinque Faceted Interface

Spinque features a faceted search interface for quick results set refinement (Beta interface shown).

Overall, Spinque is a very interesting system with a lot of potential for users looking to “get their hands dirty” and create their own customized search solution. It has special appeal to those looking to combine large databases into one transparent search system, an aspect often lacking in multi-database tools. The new approach to search promoted by Spinque seems very useful for any information professional dealing with complex data, be it the business analysist, a forensic detective, a real estate agent or a librarian. This holds extra promise for the patent and prior art searching world because of the many avenues and databases the search professional must utilize. I’m eager to see the product in a final state, and we’ll be sure to pass along news when that happens.

Are you curious about Spinque or the general topic of creating your own custom search system? I’m eager to hear if this is the kind of solution you’ve been looking for, so let me know what you think in the comments below!

Read more about Spinque on their website and take a look at their related publications section–I bet you’ll find the information going “behind the scenes” of their product as interesting as I did!

Patent Analysis from Landon IP

This post was contributed by Intellogist Team member Chris Jagalla. The Intellogist blog is provided for free by Intellogist’s parent company, Landon IP, a major provider of patent search, technical translation, and information services.

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: