NeuralQA: A Usable Library for Question Answering (Contextual Query Expansion + BERT) on Large Datasets

Existing tools for Question Answering (QA) have challenges that limit their\nuse in practice. They can be complex to set up or integrate with existing\ninfrastructure, do not offer configurable interactive interfaces, and do not\ncover the full set of subtasks that frequently comprise the QA pipeline (query\nexpansion, retrieval, reading, and explanation/sensemaking). To help address\nthese issues, we introduce NeuralQA - a usable library for QA on large\ndatasets. NeuralQA integrates well with existing infrastructure (e.g.,\nElasticSearch instances and reader models trained with the HuggingFace\nTransformers API) and offers helpful defaults for QA subtasks. It introduces\nand implements contextual query expansion (CQE) using a masked language model\n(MLM) as well as relevant snippets (RelSnip) - a method for condensing large\ndocuments into smaller passages that can be speedily processed by a document\nreader model. Finally, it offers a flexible user interface to support workflows\nfor research explorations (e.g., visualization of gradient-based explanations\nto support qualitative inspection of model behaviour) and large scale search\ndeployment. Code and documentation for NeuralQA is available as open source on\nGithub (https://github.com/victordibia/neuralqa}{Github).\n