ChemXSeer: A Digital Library for Chemical Kinetics Data & Scientific 
Literature

Prasenjit Mitra

Scientists have digital documents and experimental data that they
want to publish, link, and share.  ChemXSeer is an ongoing NSF-
funded project that aims to establish a digital library for
documents and data related to chemical kinetics.  This talk will
introduce the architecture and algorithms deployed for the
following components: (a) Data extraction: (i) TableSeer : This
tool automatically identifies tables in digital documents and
extracts the contents in the cells of the tables. The contents are
stored in a queryable table in a database. TableSeer extracts table
metadata, and uses a novel ranking function to search for tables
relevant to user queries. (ii) Extraction of data from two-
dimensional plots in figures in digital documents using image
processing techniques, (b) Chemical Entity Search : We seek to
enable improved search capabilities for chemists.  Such tools are
absolutely vital for easy access of information by chemists.  Our
tool identifies chemical formulae and chemical names,
disambiguates the terms from other general terms using
hierarchical Conditional Random Fields, and tags them. Novel
similarity scores, ranking functions and search methods are
introduced to enable searching for chemical entities.

I will also briefly talk about our efforts in populating chemical
kinetics databases with both extracted data and data submitted by
domain scientists.  Our tools can process, store and link data in
multiple formats, e.g., Excel, XML, Gaussian, and Charmm. A
metadata ad-on can help annotate the data and link multiple
datasets. The metadata is then used to link the data to published
articles allow the end-user to search for relevant data and
examine the articles that contain descriptions of the data.


Biography: Prasenjit Mitra is an assistant professor at the College
of Information Sciences and Technology at the Pennsylvania
State University.  He received his Ph.D. in Electrical Engineering
from Stanford University in 2004.  Prior to that, he had received
a Master of Science degree in Computer Science at The
University of Texas at Austin in December 1994. His Bachelor
of Technology (with Honors) degree in Computer Science and
Engineering was received from the Indian Institute of
Technology, Kharagpur in May, 1993. From 1995, he was a
senior member of the technical staff at the Server Technologies
division at Oracle Corporation in Redwood Shores, CA for five
years developing database management systems software. He
also held a position as a senior engineer at Narus, Inc. and
DBWizards working on database applications.  His primary
research interests are in information extraction from the world-
wide web.  He has also pursued research on data mining, digital
libraries, database systems, information retrieval, and artificial
intelligence.  He has published around 50 scholarly articles in
peer-reviewed journals, conferences and workshops including
Science, WWW, WebDb, SIGIR, AAAI, TKDE, PODS, EDBT,
etc.  He has served on the program committees of AAAI, WWW,
CIKM, ICDM and other prestigious conferences.