Hack Session: Enabling Intelligent Search using Question-Answer Models

Nov 15, 2019


Auditorium 2

60 minutes


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:

  1. Introduction and history of Question-Answer Models
  2. Usage of word embeddings such as BERT in Question-Answer models
  3. Popular Question-Answer model topologies
  4. Using a pre-trained Question-Answer model 
  5. 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.

  • Atul Singh (Ph.D.)

    Principal Data Scientist


  • Abhishek Jha

    Data Scientist


  • Priya Shree

    Data Scientist


Copyright 2019 Analytics Vidhya. All rights reserved