Abstract:In the question answering task, we usually need to reason the answer according to the question. Question answering tasks mostly use the pretrained language model to obtain the semantic embedding of questions and choices to predict answers, while the pretrained language model cannot accurately represent the potential relationship between entities in the question. Therefore, researchers introduce knowledge graph to realize the reasoning from question entity to answer entity. However, the limitation of knowledge graph lies in the lack of background information of entities, which may lead to wrong reasoning. To solve the above problems, a new question answering system model FeQA is proposed, which adopts large-scale pretrained language model and knowledge graph. The former uses dual-attention mechanism to enhance the semantics of questions by using Wiktionary and other question answering datasets, while the latter uses graph neural network to infer entities. During the interaction of two modal knowledge, the former provides the basis for the reasoning of nodes in the latter, and the latter provides structured knowledge for the former. After several reasoning iterations, the final answer is obtained by using the knowledge of the two modes. The experimental results on the CommonsenseQA and OpenBookQA datasets show that the performance of this model is better than that of the baseline models. Ablation experiments show that the components and knowledge sources included in this model play an important role in the effect of question and answering task. Extended experiments show the model has good application capability.

Enhancing Freebase Question Answering Using Textual Evidence.

Automatic Question-Answering Based on Wikipedia Data Extraction

Open Domain Question Answering Via Semantic Enrichment

Novel Knowledge-Based System with Relation Detection and Textual Evidence for Question Answering Research.

Hybrid Question Answering over Knowledge Base and Free Text.

A General FOFE-net Framework for Simple and Effective Question Answering over Knowledge Bases

Improving Commonsense Question Answering by Graph-based Iterative Retrieval over Multiple Knowledge Sources

Enhancing Complex Question Answering over Knowledge Graphs through Evidence Pattern Retrieval

Pay More Attention to Relation Exploration for Knowledge Base Question Answering

Simple Question Answering by Attentive Convolutional Neural Network.

Complex Factoid Question Answering with a Free-Text Knowledge Graph

Graph-Based Reasoning over Heterogeneous External Knowledge for Commonsense Question Answering

Improving Question Answering Based on Query Expansion with Wikipedia

Improving Knowledge Base Completion by Incorporating Implicit Information.

KBQA: an Online Template Based Question Answering System over Freebase.

Question Answering on Knowledge Bases and Text using Universal Schema and Memory Networks

FeQA: Fusion and enhancement of multi-source knowledge on question answering

Localize, Retrieve and Fuse: A Generalized Framework for Free-Form Question Answering over Tables

Question Answering With Character-Level Lstm Encoders And Model-Based Data Augmentation

Ask Me Anything: Free-form Visual Question Answering Based on Knowledge from External Sources