Natural Language Processing, Machine Learning, AI

My current passion is natural language processing (NLP), text analysis, corpus linguistics, machine learning (ML), knowledge representation (KR), automated question-answering (QA), logical inference, and how to make the above AI / AGI technologies scalable and massively distributed.

My career has see-sawed between technical leadership roles and applied research, with experience in large companies, executive roles in small companies, and in academia. I'm crazy about open source! Can't live without it! Strong background in mathematics and physics, with recent academic publications.

Experience

Senior Researcher - Novamente

Privately Held; Research industry

December 2007 – Present (1 year 11 months)

Development of systems for natural language processing and reasoning based on a text corpus. Work on word-sense disambiguation, automated ontology extraction, anaphora and reference resolution, and statistical/machine-learning based on corpus linguistics. Maintain the Link Grammar parser, the RelEx semantic relation extractor. Contribute to the OpenCog artificial general intelligence (AGI) system.

Research Scientist - Link Grammar

Open Source Project

December 2007 – Present (1 year 11 months)

Maintain and develop the CMU Link Grammar NLP parser. Expand the English-language parsing dictionaries. Add support for corpus-based statistical parse ranking. Add support for runtime, table lookup-based word-sense disambiguation (WSD) based on the grammatical constructs used in a sentence. Ongoing research into statistical methods to improve parse coverage and accuracy.

Lead developer - RelEx

Open Source Project

December 2007 – Present (1 year 11 months)

Develop software for providing NLP grammatical dependency relations and semantic frames for the English language. Develop code for anaphora (pronoun reference) resolution via Hobbs' algorithm. RelEx is vaguely comparable to the Stanford parser, and is an important part of the NLP pipeline for the OpenCog AGI project.

Research Scientist - OpenCog

Open Source Project

December 2007 – Present (1 year 11 months)

Develop the NLP pipeline for the OpenCog AGI project. This includes the integration of an NLP parser subsystem, a WSD and anaphora resolution system, integration with (upper) ontologies (primarily the MIT OpenMind/CommonSense corpus, but also the OpenCyc dataset), and the use of probabilistic reasoning (forward and backward chainers) for NLP and semantic deduction/induction. Develop/maintain core system components, including object persistence and distributed processing capabilities.