In this session, Sukriti Sharma from IBM presents some of IBM’s best NLP technologies productionized and available in a Python embeddable library. ‘Watson NLP’ is IBM’s standard NLP library, with a wide range of features with pre-trained models and support for custom training; cross-lingually stable across several languages (30+) ; powering 10+ IBM products, supporting both high quality and high runtime performance use cases and containing innovations from IBM research and other IBM products. Sukriti provides a broad overview of WatsonNLP and deeper insights on entity extraction, text classification, and topic modeling.
Presenter Bio
Sukriti Sharma, ML engineer at IBM
- Manages the team building a python embeddable standard NLP library
- Focus has been on experimenting with different NLP algorithms- particularly for entity extraction; model evaluation, analysis and data collection; building scalable machine learning solutions and productionizing NLP
- Master’s degree in computer science from North Carolina State University, with a specialization in data science