Open Domain Question Answering System
Jump to navigation
Jump to search
An Open Domain Question Answering System is a Question Answering System that can extract an answer from unlabeled data in large open-domain knowledge bases.
- AKA: Open QA System, Open Question-Answering System.
- Example(s):
- Counter-Example(s):
- See: Open Information Extraction System, TREC QA Competition, QA Service, Deep Similarity Neural Network, Natural Language Processing, Chatterbot, Turing Test, Knowledge Base, Expert System, Computational Linguistics.
References
2014
- (Fader et al., 2014) ⇒ Anthony Fader, Luke Zettlemoyer, and Oren Etzioni. (2014)."Open question answering over curated and extracted knowledge bases". In: Proceedings of the 20th ACM SIGKDD International Conference on Knowledge discovery and data mining, pp. 1156-1165. ACM.
- QUOTE: In this paper, we present OQA, the first Open QA system to leverage both curated and extracted KBs. A key challenge in Open QA is to be robust to the high variability found in natural language and the many ways of expressing knowledge in large-scale KBs. OQA achieves this robustness by decomposing the full QA problem into smaller sub-problems that are easier to solve. Figure 1 shows an example of how OQA maps the question “How can you tell if you have the flu?” to the answer “chills” over four steps.
Figure 1: OQA automatically mines millions of operators (left) from unlabeled data, then learns to compose them to answer questions (right) using evidence from multiple knowledge bases.
- QUOTE: In this paper, we present OQA, the first Open QA system to leverage both curated and extracted KBs. A key challenge in Open QA is to be robust to the high variability found in natural language and the many ways of expressing knowledge in large-scale KBs. OQA achieves this robustness by decomposing the full QA problem into smaller sub-problems that are easier to solve. Figure 1 shows an example of how OQA maps the question “How can you tell if you have the flu?” to the answer “chills” over four steps.
2013
- (Fader, et al., 2013) ⇒ Anthony Fader, Luke Zettlemoyer, and Oren Etzioni. "Paraphrase-driven learning for open question answering". In: Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), vol. 1, pp. 1608-1618.
- QUOTE: In sum, we make the following contributions:
- We introduce PARALEX, an end-to-end open-domain question answering system.
- We describe scalable learning algorithms that induce general question templates and lexical variants of entities and relations. These algorithms require no manual annotation and can be applied to large, noisy databases of relational triples.
- We evaluate PARALEX on the end-task of answering questions from WikiAnswers using a database of web extractions, and show that it outperforms baseline systems.
- We release our learned lexicon and question-paraphrase dataset to the research community, available at http://openie.cs.washington.edu.
- QUOTE: In sum, we make the following contributions: