ABSTRACT
Information retrieval (IR) started with the invention of the computer system’s memory. IR spanned the cache, primary memory, secondary memory, database server, flat file system, networked file system, distributed file system, cloud recently, etc. IR’s domain, scope, and technology have gone through a rapid revolution since the inception of the internet and the World Wide Web. IR operations range from simple files to the complex networks of the World Wide Web. It ranges from a single machine to a cluster of servers. Depending on the requirement, IR can be of various types. The IR algorithm also changes with the requirements. The QA system, which has evolved and gradually separated itself from a search engine, is one of the latest forms of IR. Over time, semantic analysis for the perfection of such an IR operation has become inevitable. Novelties or mentionable contributions in this chapter are as follows: (1) discussion about the state-of-the-art research work in QA systems; (2) a detailed survey of recent research on QA systems in global, Indian, and Bangla languages; (3) discussion and systematic comparison of the architecture of the repository, algorithm, and results of various proposed QA systems; and (4) acquiring knowledge of previous work and existing resources and finding research gaps in the previous work. In this chapter, the history and progress of this research domain are discussed first. Then, state-of-the-art QA systems are brought into the scope of this survey. Next, recent research and advancements in the QA systems in various related global, Indian, and Bangla languages are discussed. After that, a comparison of different QA systems based on repository, algorithm, and performance, with emphasis on the Bangla QA system in a separate section, is presented.
