Search is undergoing a paradigm shift from a keyword-based search that returns a list of documents ranked by relevance, to queries asked in natural language that can retrieve the exact answer from a corpus of documents. The hack session presents an introduction to deep-learning based question-answer models. These models by virtue of the underlying transfer learning layer (using contextualized word embeddings such as BERT) can easily find exact answers to factoid questions from a corpus of documents on which they were not trained.
The outline of the hack session is as follows:
- Introduction and history of Question-Answer Models
- Usage of word embeddings such as BERT in Question-Answer models
- Popular Question-Answer model topologies
- Using a pre-trained Question-Answer model
- Train a simple Question-Answer model
Key Takeaways for the Audience
The hack session will enable an understanding of Question-Answer models to build intelligent search solutions for their business requirements. The session will also introduce the audiences to a powerful application of word embeddings driven transfer learning for a real-life problem.