A Theoretical Survey on Question Answering Systems and the Future

Extracting the right piece of relevant information from the vast and ever-growing volume of textual data across the internet and other knowledge repositories remains significantly challenging and time-consuming task. Question answering (QA) offers an intelligent solution to the information retrieval problem by allowing users to express their quest for information in natural language statements or questions and receive precise answers as response to queries. QA represent one of the most natural ways for human-computer interaction, whereby enabling systems to provide direct response to human queries. Early QA systems were designed for restricted domains and offered limited functionalities, but modern approaches are focused handling general questions using information from multiple data sources to provide more precise answers. Quiet a given number of question answering systems have developed over the years and conducting a comprehensive review is necessary to understand emerging trends and identify future research opportunities. This study presents a theoretical survey of question answering systems and classifying them according to key characteristics, evaluation methodologies and potential future research prospects.