Question Answering system is a method based on natural language, whose aim is to locate a piece of information within a large document. We could describe it as a method to rescue answers to specific questions made by a user (“Who was The Lord Of The Rings written by?” “Which is the capital of Russia?”).
This technology is often seen as a better system than the search engines –even though we can’t deny that it’s connected to them– as it gives us specific information, and we don’t have to read a large text to find what we’re looking for. In addition, this system is not only for an only text, but it can be used in the whole World Wide Web.
Interactive Question Answering system was developed in around 1960 and it was used as a natural language’s interface for expert systems in specific domains.
In this system, we could differentiate between two different methods: Shallow and Deep.
- Shallow: Using keyword-based techniques to locate the interesting information, and then filter the desired answer. Syntactic features are used for this method (word order, word location, similarities…).
- Deep: Named-entity recognition, relation detection, syntactic alternations… it’s a more complicated technique. It needs some knowledge that can be found in ontologies such as WordNet.
To sum up, we could say that nowadays people want to have their information as fast as possible, and that Interactive Question Answering is a great system for this, as it provides as with concrete information. It provides us with an answer to our questions.
References:
- Question Answering. In Wikipedia, The Free Encyclopedia. Retrieved 18:35, May 9, 2009, from: http://en.wikipedia.org/wiki/Question_answering
- Sistemas de búsqueda Question-Answering. In Lycos. Retrieved 18:40, May 9, 2009, from: http://usuarios.lycos.es/questionanswering/
- Búsqueda de respuestas. In Wikipedia, La Enciclopedia Libre. Retrieved 18:45, May 9, 2009, from: http://es.wikipedia.org/wiki/B%C3%BAsqueda_de_respuestas
- Question Answering in Webclopedia. In Webclopedia. Retrieved 18:56, May 9, 2009, from: http://trec.nist.gov/pubs/trec9/papers/webclopedia.pdf
