multilingual computing: NLP, I18N, L10N, IR
Jim
1821 18th Ave.
Longmont, CO 80501
(303) 774-1058
jhargr...@msn.com
OBJECTIVE:
To make a difference as a software engineer in the fields of natural
language engineering or multilingual software development.
CORE EXPERTISE:
Software and Web Internationalization
Computer (Aided) Translation
Document Parsing and Extraction
Natural Language Engineering
Information Retrieval
Globalization Management Systems
ORIGINAL CONTRIBUTIONS AND INVENTIONS:
Multilingual Translation Memory, Terminology and Concordance Search
Engine - Inventor
Principle Inventor, Designer and Programmer. Patented technology for
fast fuzzy matching in any human language.
Trainable Multilingual Sentence Breaker - Inventor
Principle Inventor, designer and programmer. A sophisticated C++
program that learns to break sentences in any language. Uses a machine
learning algorithm custom designed for the task.
Multilingual Web Search Engine
Principle designer and programmer. Search 11 locales. C++ and Perl
CGI. Search engine handles all codeset and language issues.
Terminology Recognition
Principle designer and programmer. Java based algorithms use
custom-built string alignment and ngram matching algorithms to find
all fuzzy occurrences of a term within a sentence or paragraph.
Spanish Part-of-Speech Tagger
Principle Programmer. One of the first high coverage Spanish
part-of-speech taggers.
Document Extraction Format
Designer and programmer. Designed a TMX superset to act as an XML
pivot between all extracted formats and internal algorithms. This TMX
superset formed the basis of a second-generation format called GXML.
TMX Parser
Programmer. Custom C++ TMX parser with word count and syntax checker.
Parallel Text Alignment
Principle Designer and Programmer. Automatic alignment of translated
texts using Gale and Church.
Fast Analogical Model
Principle Programmer and Designer. A very fast C++ implementation of
Skousen's Analogical Model a general machine learning program for
symbolic systems.
Vector-Based Morphology
Masters Research. Algorithms for language independent and
automatically generated morphologies.
Syllable-Based Latent Semantic Indexing (LSI) for Semantic
Disambiguation
Original Research. New approach for deriving Latent Semantic Indexing
vectors.
Databases and OSs:
Oracle, MySQL, DB/2, Win32, Unix (Solaris and Linux)
Computer Languages:
Java (4 years), Python/Jython (2 years), C++/C (7 years), Perl (6
years), Pascal (2 years), Prolog (1.5 years), Smalltalk (6 months),
Common Lisp
Human Languages:
English (native), Spanish (Graduate level + 18 months in Central
America), Portuguese, Arabic
Libraries:
ICU4C and ICU4J (Internationalization Components for Unicode), ORO
Regex, Xerces
Tools:
JavaCC, VisualParse++, LEX/YACC, SWIG, Jakarta Ant, JESS, CVS
XML, Localization, Unicode, TMX, XLIFF, Statistical and Corpus-Based
Natural Language Processing, Computational Morphology, Artificial
Intelligence, Machine Learning, Human-Computer Interaction and
Usability Testing.
WORK HISTORY:
5/99 - 6/02 GLOBALSIGHT Boulder, Colorado
Senior Software Engineer of Linguistic Technologies - Highly
successful in a fast turnaround, production quality development
environment. Responsible for design and implementation of linguistic
architecture, code set converters, word and sentence breakers,
document parser/extractor framework, online/offline translation
editors, translation memory and terminology recognition. All code
(Java, Perl, and C++) was fully Unicode enabled and locale sensitive.
On November 28th, 2000 won Globalsight's "Presiden'ts Award" for
outstanding contributions to the company.
11/94 - 4/99 ILE (NOW LIONBRIDGE) Boulder, Colorado
Senior Natural Language Engineer - Responsible for designing and
implementing state-of-the-art linguistic applications and translation
tools. Principle inventor of a patented language independent fuzzy
matching engine used in several translation and localization tools.
Gained experience with the entire localization process and developed
tools and strategies to solve day-to-day production problems.
8/92 - 11/94 COMPUTING RESEARCH LAB Las Cruces, New Mexico
Research Assistant - Developer on the Cibola translator workstation
project. Cibola is a large project underway at CRL that is
researching state-of-the-art techniques and their application to
machine aided translation. Cibola embodies several key technologies
such as multilingual information retrieval, parallel text alignment
and fuzzy matching. The applicant was chosen for the project because
of his expertise in computational morphology and Spanish.
Developer on the Pangloss machine translation project. Pangloss is a
three-site endeavor exploring a multi-paradigm approach to automatic
translation. The applicant was chosen for the project over dozens of
applicants for the position because of his diverse skills in
Linguistics, Computer Science and Spanish.
1/92 - 6/92 CARL VINCENT INSTITUTE OF GOVERNMENT Athens, Georgia
Full Time Analyst/Programmer - Responsible for the design and
implementation of a sophisticated Geographical Information System.
The applicant began the job with no previous knowledge of GIS or the
Arc/Info development system, but quickly learned the technology. In
addition, the GIS system had a very short schedule for completion, but
was delivered on time. The Cobb County Tax Assessors Office, a major
county near Atlanta Georgia, currently uses this system.
5/90 - 1/92 UGA DEPARTMENT OF POULTRY SCIENCE Athens, Georgia
Part Time Programmer - In charge of designing and implementing
software for research and education. One such program, UFLP (User
Friendly Linear Program) has been published with the book: Computer
Applications in Animal Feed Formulation: Principles, Practices and a
Microcomputer Program. UFLP has won several awards for outstanding
educational software.
EDUCATION:
NEW MEXICO STATE UNIVERSITY Las Cruces, New Mexico
MS in Natural Language Engineering and AI - 1994 (All But Thesis)
GPA 3.9/4.0
UNIVERSITY OF GEORGIA Athens, Georgia
BA in Linguistics with Computer Science Minor - 1990.
PUBLICATIONS, PATENTS, CERTIFICATIONS:
Patent Number US5724593: Machine Assisted Translation Tools. Principle
Inventor.
Farwell, D., Helmreich, S., Jin, W., Casper, M., Hargrave, J.,
Salgado, H. and Weng, F.L. (1994). PANGLYZER: Spanish Language
Analysis System in Proceedings of AMTA-94 (Association for Machine
Translation in the Americas).
GlobeMark's Web Globalist Certification (www.webofculture.com)