Joint Learning of Sentence Embeddings for Relevance and Entailment

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F16%3A00301135" target="_blank" >RIV/68407700:21230/16:00301135 - isvavai.cz</a>

Výsledek na webu

<a href="http://aclweb.org/anthology/W16-1602" target="_blank" >http://aclweb.org/anthology/W16-1602</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/W16-1602" target="_blank" >10.18653/v1/W16-1602</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Joint Learning of Sentence Embeddings for Relevance and Entailment

Popis výsledku v původním jazyce

We consider the problem of Recognizing Textual Entailment within an Information Retrieval context, where we must simultaneously determine the relevancy as well as degree of entailment for individual pieces of evidence to determine a yes/no answer to a binary natural language question. We compare several variants of neural networks for sentence embeddings in a setting of decision-making based on evidence of varying relevance. We propose a basic model to integrate evidence for entailment, show that joint training of the sentence embeddings to model relevance and entailment is feasible even with no explicit perevidence supervision, and show the importance of evaluating strong baselines. We also demonstrate the benefit of carrying over text comprehension model trained on an unrelated task for our small datasets. Our research is motivated primarily by a new open dataset we introduce, consisting of binary questions and news-based evidence snippets. We also apply the proposed relevance-entailment model on a similar task of ranking multiple-choice test answers, evaluating it on a preliminary dataset of school test questions as well as the standard MCTest dataset, where we improve the neural model state-of-art.

Název v anglickém jazyce

Joint Learning of Sentence Embeddings for Relevance and Entailment

Popis výsledku anglicky

We consider the problem of Recognizing Textual Entailment within an Information Retrieval context, where we must simultaneously determine the relevancy as well as degree of entailment for individual pieces of evidence to determine a yes/no answer to a binary natural language question. We compare several variants of neural networks for sentence embeddings in a setting of decision-making based on evidence of varying relevance. We propose a basic model to integrate evidence for entailment, show that joint training of the sentence embeddings to model relevance and entailment is feasible even with no explicit perevidence supervision, and show the importance of evaluating strong baselines. We also demonstrate the benefit of carrying over text comprehension model trained on an unrelated task for our small datasets. Our research is motivated primarily by a new open dataset we introduce, consisting of binary questions and news-based evidence snippets. We also apply the proposed relevance-entailment model on a similar task of ranking multiple-choice test answers, evaluating it on a preliminary dataset of school test questions as well as the standard MCTest dataset, where we improve the neural model state-of-art.

Druh

D - Stať ve sborníku

CEP obor

IN - Informatika

OECD FORD obor

—

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2016

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the 1st Workshop on Representation Learning for NLP

ISBN

978-1-945626-04-3

ISSN

—

e-ISSN

—

Počet stran výsledku

10

Strana od-do

8-17

Název nakladatele

Association for Computational Linguistics (ACL)

Místo vydání

Stroudsburg

Místo konání akce

Berlin

Datum konání akce

11. 8. 2016

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—