Unsupervised Natural Question Answering with a Small Model
This work addresses the challenge of adding explicit knowledge to AI systems efficiently, but it appears incremental as it builds on existing unsupervised techniques for question answering.
The paper tackles the problem of enabling small models to answer factoid questions by using external knowledge, achieving results through an unsupervised learning architecture that complements language model training without extensive additional training.
The recent (2019-02) demonstration of the power of huge language models such as GPT-2 to memorise the answers to factoid questions raises questions about the extent to which knowledge is being embedded directly within these large models. This short paper describes an architecture through which much smaller models can also answer such questions - by making use of 'raw' external knowledge. The contribution of this work is that the methods presented here rely on unsupervised learning techniques, complementing the unsupervised training of the Language Model. The goal of this line of research is to be able to add knowledge explicitly, without extensive training.