A Survey of Methods, Tools and Applications of Knowledge Base Construction (KBC)

Abstract

Knowledge Bases (KBs) have recently become very valuable because of their use in many Artificial Intelligence (AI) applications. For example - KBs are a critical part of conversational agents that are rapidly being adopted by the industry. Knowledge Bases are made up of a number of facts about the real world, and the number of such facts is typically very large. The construction of the KB involves identifying the facts from unstructured data such as text, images, videos and speech. Due to the challenges of processing unstructured data, Knowledge Base Construction (KBC) used to be done manually, requiring huge efforts and long timelines. In recent years, intelligent technologies such as Machine Learning and Deep Learning are being employed for the purpose of KBC. In this paper, we provide an introduction to KBC and describe how various cutting edge technologies are being employed for its automation. We then mention some KBC systems created by academic as well as commercial groups. We survey some solutions in the industry that are already using KBC. In addition, we attempt to predict the future research possibilities in this field.

Keywords: Knowledge Base Construction, Information Extraction, Text Processing, Knowledge Graph, Machine Learning

View PDF