Researched, developed, and implemented new algorithms and applications for NLM's Indexing Initiative Project. Worked on mechanisms to streamline MetaMap update, user interface, and back-end software for the DCMS computer augmented retrieval system. MetaMap develops an knowledge based ordered list of UMLS Metathesaurus concepts for each citation, based on the noun phrases extracted from that text. Current environments include Windows, Macintosh OS/X, and Linux. Applying machine learning using TensorFlow and PyTorch for NLP tasks.

Researched, developed, and implemented new algorithms and applications for NLM's Indexing Initiative Project. Ported the Journal descriptor indexing system from in preparation for the conversion of the systems from development to production status. Worked on mechanisms to streamline MetaMap update, user interface, and back-end software for the DCMS computer augmented retrieval system. MetaMap develops an knowledge based ordered list of UMLS Metathesaurus concepts for each citation, based on the noun phrases extracted from that text. Current environments include Sun Solaris, Windows NT, Macintosh OS/X, and Linux.

Produced software for Natural Language Processing and Statistical Information Retrieval.

As part of the System Team for Project NBUS (A Budget System), analyzed GemStone remote method invocation (RMI) performance. Designed and implemented a NBUS connection authentication facility (EJB, X.509 based). Analyzed and optimized the application code, and tuned the GemStone application server. For Project Baobab, designed and implemented a Use Case Application Product for the Zope Content Management Application Server in Python. For Project FoneCap, designed and implemented a Mobile Device Profile Repository for the Fonecap toolkit using XML/DOM (eXtensible Markup Language/Document Object Model). The platforms used for development include Windows NT, Solaris, OSF/1, Linux.

For NIST's Written Natural Language Processing (Now the Information Retrieval) Group, developed and implemented new algorithms for the TREC text retrieval project and the PRISE full-text retrieval system. Acquired technical expertise in the WWW and related networking technologies. Served as primary software developer of integrated data retrieval systems, and coordinated transfer of these technologies with other agencies. Performed system and network administration for the Written Natural Language Processing Group (including NCSA and CERN http servers). Developed software using C, C++, Java, Tcl/Tk, Perl, on the following platforms: Sun OS 4.1, Solaris, AIX, OpenBSD.

Designed and implemented a new, user-friendly, front-end interface for a raster image editor, providing easy-to-use point and click drawing commands with time-saving features for accessing drawing and font files.

Provided user support to assist faculty, assisted in developing a User Guide, and tested applications programs. Installed and maintained software packages; coordinated installation of communication hardware and configured related host system software; and installed and augmented a document typesetting package. Installed, modified, and designed site-supported software.