Abstract

One long-term goal of machine learning research is to produce methods that are applicable to reasoning and natural language, in particular building an intelligent dialogue agent. To measure progress towards that goal, we argue for the usefulness of a set of proxy tasks that evaluate reading comprehension via question answering. Our tasks measure understanding in several ways: whether a system is able to answer questions via chaining facts, simple induction, deduction and many more. The tasks are designed to be prerequisites for any system that aims to be capable of conversing with a human. We believe many existing learning systems can currently not solve them, and hence our aim is to classify these tasks into skill sets, so that researchers can identify (and then rectify) the failings of their systems. We also extend and improve the recently introduced Memory Networks model, and show it is able to solve some, but not all, of the tasks.

Keywords

Computer scienceQuestion answeringChainingSet (abstract data type)Artificial intelligenceForward chainingSimple (philosophy)Reading (process)Human–computer interactionMachine learningExpert systemProgramming language

Affiliated Institutions

Related Publications

Publication Info

Year
2015
Type
preprint
Citations
719
Access
Closed

External Links

Social Impact

Social media, news, blog, policy document mentions

Citation Metrics

719
OpenAlex

Cite This

Jason Weston, Antoine Bordes, Sumit Chopra et al. (2015). Towards AI-Complete Question Answering: A Set of Prerequisite Toy Tasks. arXiv (Cornell University) . https://doi.org/10.48550/arxiv.1502.05698

Identifiers

DOI
10.48550/arxiv.1502.05698

Data Quality

Data completeness: 77%