Using knowledge bases for question answering

• Main Author: LAN, Yunshi
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2020
• Subjects: Knowledge base; knowledge base question answering; textual entailment.; Databases and Information Systems; Data Storage Systems
• Online Access: https://ink.library.smu.edu.sg/etd_coll/261; https://ink.library.smu.edu.sg/cgi/viewcontent.cgi?article=1261&context=etd_coll
• Institution: Singapore Management University

A knowledge base (KB) is a well-structured database, which contains many of entities and their relations. With the fast development of large-scale knowledge bases such as Freebase, DBpedia and YAGO, knowledge bases have become an important resource, which can serve many applications, such as dialogue system, textual entailment, question answering and so on. These applications play significant roles in real-world industry. In this dissertation, we try to explore the entailment information and more general entity-relation information from the KBs. Recognizing textual entailment (RTE) is a task to infer the entailment relations between sentences. We need to decide whether a hypothesis can be inferred from a premise based on the text of two sentences. Such entailment relations could be potentially useful in applications like information retrieval and commonsense reasoning. It's necessary to develop automatic techniques to solve this problem. Another task is knowledge base question answering (KBQA). This task aims to automatically find answers to factoid questions from a knowledge base, where answers are usually entities in the KB. KBQA task has gained much attention in recent years and shown promising contribution to real-world problems. In this dissertation, we try to study the applications of knowledge bases in textual entailment and question answering: We propose a general neural network based framework which can inject lexical entailment relations to RTE, and a novel model is developed to embed lexical entailment relations. The experiment results show that our method can benefit general textual entailment model. We design a KBQA method based on an existing reading comprehension model. This model achieves competitive results on several popular KBQA datasets. In addition, we make full use of contextual relations of entities in the KB. Such enriched information helps our model to attain state-of-art. We propose to perform topic unit linking where topic units cover a wider range of units of a KB. We use a generation-and-scoring approach to gradually refine the set of topic units. Furthermore, we use reinforcement learning to jointly learn the parameters for topic unit linking and answer candidate ranking in an end-to-end manner. Experiments on three commonly used benchmark datasets show that our method consistently works well and outperforms the previous state of the art on two datasets. We further investigate multi-hop KBQA task, i.e., question answering from KB where questions involve multiple hops of relations, and develop a novel model to solve such questions in an iterative and efficient way. The results demonstrate that our method consistently outperforms several multi-hop KBQA baselines.