Abstract:Text understanding and reasoning are among the core areas of artificial intelligence. Even a total solution to automatic text understanding and reasoning is still beyond the current techniques, thus it is time to build the stepping stones to solve a more straightforward problem set like question-answering (QA) without ambiguous utterances in the contexts or questions. The reported state-of-the-art approaches to this kind of problem are nearly all connectionist models based on neural networks. Significant progress has even been made in this direction. It is still hard for the pure connectionist models to handle logical reasoning. They generally suffer from the longstanding drawback of poor explainability and sensitivity to data noise and distribution. In this paper, we propose a complementary symbolic approach, GMR (Graph Matching based Reasoner) to QA - it automatically generates reasoning rules in the form of graphs from the training set and uses the generated rules to infer answers to the questions in the test set via graph matching. By employing this symbolic approach, 20 tasks in bAbI are solved with an average accuracy of 99.38%, and it outperforms the state-of-the-art for a real-life QA dataset WikiTableQuestions. After analyzing the accuracy of the basic evaluation indicators, we studied the generalization ability of the model in the paper, including the anti-noise ability, the convergence of the model, the stability of the model, the complexity of the algorithm, and the uncertainty of the parameters. Through comprehensive analysis and comparison, our model is stronger than the neural network model regarding anti-noise interference. Compared with the neural network model, our model performs very well in multi-tasking, and the stability of the model is quite high. The diversity of tasks did not reduce the stability of the model. We have conducted a comprehensive analysis and comparison of the parameter uncertainties. Our model can optimally select parameter configurations and will not cause a sharp drop in performance due to the parameter uncertainties. Finally, we describe the complexity of the GMR method and the optimal configuration of its parameters.

Explainable Question Answering based on Semantic Graph by Global Differentiable Learning and Dynamic Adaptive Reasoning

Dynamic Semantic Graph Construction and Reasoning for Explainable Multi-hop Science Question Answering

Ask to Understand: Question Generation for Multi-hop Question Answering

An Interpretable Reasoning Network for Multi-Relation Question Answering.

Variational Reasoning for Question Answering With Knowledge Graph

Graph Matching Based Reasoner: A Symbolic Approach to Question Answering

GNN2R: Weakly-Supervised Rationale-Providing Question Answering over Knowledge Graphs

Question-aware Dynamic Scene Graph of Local Semantic Representation Learning for Visual Question Answering

Leveraging Structured Information for Explainable Multi-hop Question Answering and Reasoning

Dynamic Dual Graph Networks for Textbook Question Answering

Multi-Hop Reasoning Question Generation and Its Application.

Stepwise relation prediction with dynamic reasoning network for multi-hop knowledge graph question answering

Discrete Subgraph Sampling for Interpretable Graph based Visual Question Answering

Question-Guided Semantic Dual-Graph Visual Reasoning with Novel Answers.

Understanding and Improving Zero-shot Multi-hop Reasoning in Generative Question Answering

Unrestricted multi-hop reasoning network for interpretable question answering over knowledge graph

Reasoning Path Generation for Answering Multi-hop Questions over Knowledge Graph

Reasoning over Hierarchical Question Decomposition Tree for Explainable Question Answering

ReasonFuse: Reason Path Driven and Global–Local Fusion Network for Numerical Table-Text Question Answering

Knowledge Graph Multi-Hop Question Answering Based on Dependent Syntactic Semantic Augmented Graph Networks

Dynamically Fused Graph Network for Multi-hop Reasoning