Structured retrieval for question answering

Bilotti, Matthew W.; Ogilvie, Paul; Callan, Jamie; Nyberg, Eric

doi:10.1145/1277741.1277802

Cited by 73 publications

(61 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most approaches to extracting the expected answer type perform some sort of syntactic analysis on the question (by chunking, shallow parsing, or probabilistic deep parsing) in order to find the question focus. Based on the question focus, the question word, and named entity classification, the expected answer type is then determined via semantic generalization using lexical semantic resources such as WordNet, either by manually defined mappings of WordNet hyponym subhierarchies to answer taxonomies (Harabagiu et al 2000; see also Section 3) or by feature-based classifiers resting on machine learning techniques (Li/Roth 2006) or statistical methods (Ittycheriah 2006 Nyberg et al (2005) and Bilotti et al (2007) try to achieve this goal by shallow semantic parsing whereas Harabagiu et al (2000) and Mollá/Gardiner (2004) transform the results of a syntactic parser into shallow logical forms (conjunctive predicate-argument structures). These approaches make use of publicly available probabilistic parsers trained on annotated corpora.…”

Section: Methodological Aspectsmentioning

confidence: 99%

79. Electronic dictionaries for question answering and natural language database access

Helbig¹,

Osswald²

2013

Dictionaries. An International Encyclopedia of Lexicography

View full text Add to dashboard Cite

Section: Methodological Aspectsmentioning

confidence: 99%

79. Electronic dictionaries for question answering and natural language database access

Helbig¹,

Osswald²

2013

Dictionaries. An International Encyclopedia of Lexicography

View full text Add to dashboard Cite

“…The version of Indri used throughout this thesis extends the notion of a field to include a parent pointer, through which a field can optionally point to another field within the same document [43,8]. This pointer provides a means of checking relationships between fields that do not enclose one another.…”

Section: Indri Search Enginementioning

confidence: 99%

“…The experiments presented here and originally published in [8], evaluate whether the structured retrieval approach can provide a better quality passage ranking when compared to a baseline retrieval approach consisting of keyterms drawn from question, with named entity support for the expected answer type, which is considered to be a strong baseline for QA.…”

Section: Structured Retrievalmentioning

confidence: 99%

Linguistic and semantic passage retrieval strategies for question answering

Bilotti

2011

SIGIR Forum

Self Cite

View full text Add to dashboard Cite

Question Answering (QA) is the task of searching a large text collection for specific answers to questions posed in natural language. Though they often have access to rich linguistic and semantic analyses of their input questions, QA systems often rely on off-the-shelf bag-of-words Information Retrieval (IR) solutions to retrieve passages matching a set of terms extracted from the question.There is a fundamental disconnect between the capabilities of the bag-of-words retrieval model and the retrieval needs of the QA system. Bag-of-words IR retrieves documents matching a query, but the QA system really needs documents that contain answers. Through question analysis, the QA system has compiled a sophisticated information need representation for what constitutes an answer to the question. This representation is composed of a set of linguistic and semantic constraints satisfied by answer-bearing passages. Unfortunately, off-the-shelf IR libraries commonly used in QA systems can not, in general, check these types of constraints at query-time. Poor quality retrieval can cause a QA system to fail if no answer-bearing text is retrieved, if it is not ranked highly enough, or if it is outranked or overwhelmed by false positives, text that matches the query well, yet supports a wrong answer.This thesis proposes two linguistic and semantic passage retrieval methods for QA, one based on structured retrieval and the other on rank-learning techniques. In addition, a methodology is proposed for mapping annotated text consisting of labeled spans and typed relations between them into an annotation graph representation. The annotation graph supports query-time linguistic and semantic constraint-checking, and serves as a unifying formalism for the QA system's information need and for retrieved passages. The proposed methods rely only on the relatively weak assumption that the QA system's information need can be represented as an annotation graph. The two approaches are shown to retrieve more answer-bearing text, more highly ranked, compared to a bag-of-words baseline for two different QA tasks. Linguistic and semantic passage retrieval methods are also shown to improve end-to-end QA system accuracy and answer MRR.iv Acknowledgments I would like to thank my advisor, Eric Nyberg, for all he has done, and the rest of my committee, Jamie Callan, Jaime Carbonell and Eric Brown, for helping to guide this work.

show abstract

“…Bilotti, et. al., use structured queries to retrieve text satisfying PropBank-style semantic constraints [3]. Their method is shown to be effective in certain cases, yet poor at combining evidence from bag-of-words and structured features and not robust to ranking partial matches.…”

Section: Related Workmentioning

confidence: 99%

Rank learning for factoid question answering with linguistic and semantic constraints

Bilotti

Elsas

Carbonell

et al. 2010

Proceedings of the 19th ACM International Conference on Information and Knowledge Management

Self Cite

View full text Add to dashboard Cite

This work presents a general rank-learning framework for passage ranking within Question Answering (QA) systems using linguistic and semantic features. The framework enables query-time checking of complex linguistic and semantic constraints over keywords. Constraints are composed of a mixture of keyword and named entity features, as well as features derived from semantic role labeling. The framework supports the checking of constraints of arbitrary length relating any number of keywords. We show that a trained ranking model using this rich feature set achieves greater than a 20% improvement in Mean Average Precision over baseline keyword retrieval models. We also show that constraints based on semantic role labeling features are particularly effective for passage retrieval; when they can be leveraged, an 40% improvement in MAP over the baseline can be realized.

show abstract

Structured retrieval for question answering

Cited by 73 publications

References 18 publications

79. Electronic dictionaries for question answering and natural language database access

79. Electronic dictionaries for question answering and natural language database access

Linguistic and semantic passage retrieval strategies for question answering

Rank learning for factoid question answering with linguistic and semantic constraints

Contact Info

Product

Resources

About