PIUG Conference--Tuesday's blog

The Tuesday morning sessions at the 2012 Annual Meeting of the Patent Information Users Group (PIUG) focused on challenges, opportunities, and developments in patent analysis through visualization tools and developments in data dissemination by the patent offices.  Nils Newman of Search Technology, Inc. noted that, even though patent analysis relies upon mathematical and algorithmic techniques, the visualization of the data provided by these techniques is critical to understanding the results.   Mike Baycroft of Fairview Research, LLC, described the challenge researchers  face in reviewing the enormous volumes of published patents.  He found that more than 70% of a patent searcher’s time is spent not in searching, but in determining the relevance of patents retrieved by searching.   Searchers must spend valuable time reading the titles, abstracts, drawing sheets, and independent claims to sort out documents that are truly relevant.   He described a text mining and visualization tool called Treparel’s KMX Patent Analytics solution.  This tool improves searcher productivity by surfacing semantic relationships  between significant terms and phrases in the retrieved patents.

An outstanding presentation by Marla Grossman, Executive Director of the Coalition for Patent and Trademark Information Dissemination (CPTID), provided an overview of current plans at the U.S. Patent and Trademark Office (USPTO).  LexisNexis, a longtime member of CPTID, has a major interest in the USPTO’s plans for its data.  A significant part of LexisNexis’s content revenue depends on U.S. patents.  Marla Grossman covered the USPTO’s Public Dissemination Solution project.  This project is planned to meet President Obama’s Open Government Initiative to provide increased transparency of government operations by making more electronic patent data made available in bulk on the internet.   This goal will involve using Optical Character Recogntion (OCR) to create electronic data from text.  It is known that OCR technology can create corrupted data.  Corrupted data has an impact on search systems such as Lexis that rely on correctly spelled keywords.   Ms. Grossman explained the possible impact of the USPTO releasing many terabytes of OCR’d data into the bulk database and discussed alternative approaches.

Another current project at the USPTO is the transition to the Cooperative Patent Classification (CPC) system.   The CPC is the result of an agreement between the USPTO and the European Patent Office (EPO) to jointly develop a patent classification scheme that is based on the International Patent Classification (IPC) scheme.   Currently, the USPTO and the EPO use different classification systems.  This creates disconnects in the work between the two offices.   Jerry Lorengo, the Co-Program Coordinator for CPC at the USPTO described the progress being made to release the CPC in 2013.  Mr. Lorengo is a patent examiner and classification expert who has served as Supervisory Patent Examiner in the Chemical and Materials Engineering Technology Center.  His presentation underscored the growing cooperation of the patent offices to create the ultimate international classification system, the Common Hybrid Classification (CHC) system.  I think the patent world will be revolutionized by the CHC.   The CHC will make communication between world patent offices much more streamlined.

A highlight of the afternoon session was the talk by Dan Shalloe, Editor-in-Chief of EPIDOS News for the EPO.  Mr. Shalloe discussed the Obama administration’s “open data” policy and the open data platform, data.gov.  He noted that many countries have followed the same pattern, providing many open data sites.  Mr. Shalloe noted that open data policies create new issues on information policy for the patent offices, specifically in pricing and licensing.  Also, under the open data model, data itself will be no cost.  Therefore, this may create an opportunity for adding value to patent data through linking it to financial, company, and scientific data.  This added linking will create further possibilities for visualization models to use in predictive analysis of patent competitive value.   Tomorrow’s sessions will focus on competitive intelligence tools for patents!

View Monday's Blog.