Question Answering: Definition and Examples

Question Answering (QA) is a field of artificial intelligence that involves developing systems capable of understanding a question formulated in natural language and providing an exact answer. Unlike a traditional search engine that returns a list of documents, a QA system directly extracts or generates the expected answer.

Several types of Question Answering exist. Extractive QA identifies and extracts a precise passage from an existing document corpus. Generative QA, on the other hand, formulates an original response by synthesizing available information — this is how large language models like Claude or GPT work. There is also open-domain QA, which answers questions on any topic, and closed-domain QA, specialized in a specific knowledge area.

With the rise of LLMs, Question Answering has become one of the most common use cases of generative AI. Retrieval-Augmented Generation (RAG) techniques combine document retrieval and generation to provide answers that are both accurate and grounded in verifiable sources. In prompt engineering, mastering QA allows you to formulate questions that maximize the quality and reliability of responses.

Applications are numerous: enterprise virtual assistants, customer support chatbots, medical research systems, educational tools, and interactive knowledge bases. The quality of a QA system is mainly measured by the relevance, accuracy, and completeness of its answers.