Relevance through Machine Learning-based Data Enrichment and Enhanced Visualization
Christopher Ball • Location: Theater 4 • Back to Haystack 2020
Relevance through Machine Learning-based data enrichment and enhanced visualization - i.e. Helping users better understand and navigate a body of content.
Presentation will provide an overview of a broad spectrum of techniques:
- Context-aware Parsing
- Fine grained Named Entity Recognition
- Transfer learning from Transformer Models
- Information Extraction using Bayesian Generative Models
- Semantic Similarity Queries: by word, phrase, sentence, or paragraph
Presentation will introduce concepts, demonstrate working examples and provide tips on pursuing through open source libraries. Work extends from a joint research project with the Dartmouth College and the University of Southern California.
Christopher is the Lead Data Science Solutions Architect at meta Heuristica, a company specializing in an interdisciplinary approach to data science. Christopher helps companies and government agencies make their unstructured content more actionable across the data science life cycle: from context aware content parsing and enrichment, through assorted forms of machine learning, to search and visual analytics. He is also a regular participant in National Science Foundation funded research projects.