Detecting Expected Answer Relations through Textual Entailment
نویسندگان
چکیده
This paper presents a novel approach to Question Answering over structured data, which is based on Textual Entailment recognition. The main idea is that the QA problem can be recast as an entailment problem, where the text (T ) is the question and the hypothesis (H ) is a relational pattern, which is associated to “instructions” for retrieving the answer to the question. In this framework, given a question Q and a set of answer patterns P, the basic operation is to select those patterns in P that are entailed by Q. We report on a number of experiments which show the great potentialities of the proposed approach.
منابع مشابه
Semantic Answer Validation using Universal Networking Language
we present a rule-based answer validation (AV) system based on textual entailment (TE) recognition mechanism that uses semantic features expressed in the Universal Networking Language (UNL). We consider the question as the TE hypothesis (H) and the supporting text as TE text (T). Our proposed TE system compares the UNL relations in both T and H in order to identify the entailment relation as ei...
متن کاملAnnotating Entailment Relations for Shortanswer Questions
This paper presents an annotation project that explores the relationship between textual entailment and short answer scoring (SAS). We annotate entailment relations between learner and target answers in the Corpus of Reading Comprehension Exercises for German (CREG) with a finegrained label inventory and compare them in various ways to correctness scores assigned by teachers. Our main finding i...
متن کاملFinding Contradictions in Text
In this paper, I seek to understand the ways contradictions occur across texts and I describe a system for automatically detecting such constructions. Finding conflicting statements is foundational for text understanding, a problem which recently received a surge of interest in the computational linguistics community. Condoravdi et al. (2003) first recognized the importance of handling both ent...
متن کاملOn-demand Injection of Lexical Knowledge for Recognising Textual Entailment
We approach the recognition of textual entailment using logical semantic representations and a theorem prover. In this setup, lexical divergences that preserve semantic entailment between the source and target texts need to be explicitly stated. However, recognising subsentential semantic relations is not trivial. We address this problem by monitoring the proof of the theorem and detecting unpr...
متن کاملDiscovering Entailment Relations Using "Textual Entailment Patterns"
In this work we investigate methods to enable the detection of a specific type of textual entailment (strict entailment), starting from the preliminary assumption that these relations are often clearly expressed in texts. Our method is a statistical approach based on what we call textual entailment patterns, prototypical sentences hiding entailment relations among two activities. We experimente...
متن کامل