Filters








70,719 Hits in 3.4 sec

Explaining Answers with Entailment Trees [article]

Bhavana Dalvi, Peter Jansen, Oyvind Tafjord, Zhengnan Xie, Hannah Smith, Leighanna Pipatanangkura, Peter Clark
2021 arXiv   pre-print
To train a model with this skill, we created ENTAILMENTBANK, the first dataset to contain multistep entailment trees.  ...  Our goal, in the context of open-domain textual question-answering (QA), is to explain answers by showing the line of reasoning from what is known to the answer, rather than simply showing a fragment of  ...  These methods are primarily designed for answers to "lookup" questions, to explain where/how an answer was found in a corpus.  ... 
arXiv:2104.08661v2 fatcat:kzmhrhor7ncl5jlwk4geuravom

Explaining Answers with Entailment Trees

Bhavana Dalvi, Peter Jansen, Oyvind Tafjord, Zhengnan Xie, Hannah Smith, Leighanna Pipatanangkura, Peter Clark
2021 Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing   unpublished
To train a model with this skill, we created ENTAILMENTBANK 1 , the first dataset to contain multistep entailment trees.  ...  Our goal, in the context of open-domain textual question-answering (QA), is to explain answers by showing the line of reasoning from what is known to the answer, rather than simply showing a fragment of  ...  These methods are primarily designed for answers to "lookup" questions, to explain where/how an answer was found in a corpus.  ... 
doi:10.18653/v1/2021.emnlp-main.585 fatcat:5lazh5a5u5b4fo4ww7esjsmpyi

Predicting the Understandability of OWL Inferences [chapter]

Tu Anh T. Nguyen, Richard Power, Paul Piwek, Sandra Williams
2013 Lecture Notes in Computer Science  
In this paper, we describe a method for predicting the understandability level of inferences with OWL.  ...  We also present an evaluation study which confirms that our model works relatively well for two-step inferences with OWL.  ...  A justification provides a set of premises for an entailment, so is helpful for diagnosing an erroneous entailment; however, unlike a proof, it does not explain how the premises combine with each other  ... 
doi:10.1007/978-3-642-38288-8_8 fatcat:jzrgwj7d3bbm7c5ploxczbsz6i

XTE: Explainable Text Entailment [article]

Vivian S. Silva, André Freitas, Siegfried Handschuh
2020 arXiv   pre-print
Furthermore, for entailments requiring world knowledge, most systems still work as a "black box", providing a yes/no answer that does not explain the underlying reasoning process.  ...  In this work, we introduce XTE - Explainable Text Entailment - a novel composite approach for recognizing text entailment which analyzes the entailment pair to decide whether it must be resolved syntactically  ...  With respect to the entailment recognition methods adopted in this work, although tree edit distance has already been used in text entailment [13] , graph traversal approaches have been explored more  ... 
arXiv:2009.12431v1 fatcat:cqlnab54yvdyjevzij7yzaxsaq

METGEN: A Module-Based Entailment Tree Generation Framework for Answer Explanation [article]

Ruixin Hong, Hongming Zhang, Xintong Yu, Changshui Zhang
2022 arXiv   pre-print
Advances on QA explanation propose to explain the answers with entailment trees composed of multiple entailment steps.  ...  Knowing the reasoning chains from knowledge to the predicted answers can help construct an explainable question answering (QA) system.  ...  Conclusion We propose METGEN, a module-based framework to generate the entailment trees for explaining answers. METGEN reasons with single-step entailment modules and the reasoning controller.  ... 
arXiv:2205.02593v1 fatcat:eco3edoegvavhgnihhecu2q6bm

Cross-Policy Compliance Detection via Question Answering [article]

Marzieh Saeidi, Majid Yazdani, Andreas Vlachos
2021 arXiv   pre-print
an expression tree combines the answers to obtain the label.  ...  We conduct our experiments using a recent dataset consisting of government policies, which we augment with expert annotations and find that the cost of annotating question answering decomposition is largely  ...  The procedure for converting the data is explained in the next section.  ... 
arXiv:2109.03731v1 fatcat:rc643hekzndebj4bswpus6si44

Towards Entailment-Based Question Answering: ITC-irst at CLEF 2006 [chapter]

Milen Kouylekov, Matteo Negri, Bernardo Magnini, Bonaventura Coppola
2007 Lecture Notes in Computer Science  
The participation in the AVE task, with an answer validation module based on textual entailment recognition, is motivated by our objectives of (i) creating a modular framework for an entailment-based approach  ...  to QA, and (ii) developing, in compliance with this framework, a stand-alone component for answer validation which implements different approaches to the problem.  ...  The difficulty of the rte task explains the relatively poor performance of the systems participating in the PASCAL Challenge.  ... 
doi:10.1007/978-3-540-74999-8_64 fatcat:pjmrgp7ccfezrb4gtbmw7q5d64

Diary of a parliamentary intern

Anthony Ridge
2016 The Veterinary Record  
I usually meet twice a day with Lord Trees to exchange information and discuss future plans.  ...  The administrative work entails organising events and meetings and helping to manage Lord Trees' diary.  ... 
doi:10.1136/vr.i3062 fatcat:alszxefbv5f33mcrybmzpcyhpu

Using Neural Networks to Generate Inferential Roles for Natural Language

Peter Blouw, Chris Eliasmith
2018 Frontiers in Psychology  
We then show that it is possible to train a tree-structured neural network model to generate very simple examples of such inferential roles using the recently released Stanford Natural Language Inference  ...  On an empirical front, we evaluate the performance of this model by reporting entailment prediction accuracies on a set of test sentences not present in the training data.  ...  It is therefore theoretically possible to compute ranked collections of entailments with each tree.  ... 
doi:10.3389/fpsyg.2017.02335 pmid:29387031 pmcid:PMC5776445 fatcat:x7kzq6hyufcgnhebvneli4jkwu

How similar is similar enough?

Michela Ippolito
2017 Semantics and Pragmatics  
behind Non-Triviality is that a counterfactual statement answers a conditional question under discussion and, therefore, is required to make a non-trivial assertion.  ...  Assuming a model of the discourse along the lines proposed by Roberts (2012) and Büring (2003) , according to which conversational moves are answers to often implicit questions under discussion, the idea  ...  What explains these judgments?  ... 
doi:10.3765/sp.9.6 fatcat:5s4r5624xfd7ll3rnqtexfpq2e

A Theoretical Study of (Full) Tabled Constraint Logic Programming [article]

Joaquín Arias, Manuel Carro
2020 arXiv   pre-print
., TCLP featuring entailment checking in the calls and in the answers. We present a top-down semantics for TCLP and show that it is equivalent to a fixpoint semantics.  ...  Logic programming with tabling and constraints (TCLP, tabled constraint logic programming) has been shown to be more expressive and, in some cases, more efficient than LP, CLP, or LP with tabling.  ...  In Section 4 we explain the benefits of using entailment checking with more relaxed notions projections.  ... 
arXiv:2009.14430v1 fatcat:akof6wma6zcr5jsyj7z5ji7kwm

Exploring Knowledge Graphs in an Interpretable Composite Approach for Text Entailment

Vivian S. Silva, André Freitas, Siegfried Handschuh
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Recognizing textual entailment is a key task for many semantic applications, such as Question Answering, Text Summarization, and Information Extraction, among others.  ...  , explaining the semantic relationship holding between the pieces of text.  ...  confirm and explain the entailment.  ... 
doi:10.1609/aaai.v33i01.33017023 fatcat:j5gfq4oydzal3fsqpfn4y5dz5y

Information structure as parallel tree building

Jon Stevens
2011 Proceedings of the International Conference on Head-Driven Phrase Structure Grammar  
This paper presents a Synchronous Tree Adjoining Grammar (STAG) account of Information Structure, whereby Givenness-marking requires a link between nodes on a syntactic tree and LF nodes whose interpretation  ...  This, combined with some established insights from Game-theoretic pragmatics, yields empirical advantages over more traditional semantic/pragmatic analyses of equal simplicity.  ...  The interaction of two such principles explains why answers to QUDs are not G-marked, even when Given: effort minimization and forward induction.  ... 
doi:10.21248/hpsg.2011.21 fatcat:lfi4k3r3gjglzaxba337q7pt5u

Query Answering in Description Logics: The Knots Approach [chapter]

Thomas Eiter, Carsten Lutz, Magdalena Ortiz, Mantas Šimkus
2009 Lecture Notes in Computer Science  
When annotated with suitable query information, knots are a flexible tool for query answering that allows for solving the problem in a simple and intuitive way.  ...  In the recent years, query answering over Description Logic (DL) knowledge bases has been receiving increasing attention, and various methods and techniques have been presented for this problem.  ...  This observation suggests that entailment of q can be verified by replacing q with a set of tree-shaped rewritings and checking directed entailment. Definition 8.  ... 
doi:10.1007/978-3-642-02261-6_3 fatcat:6dmw3kd4mnfi5anf3idv2gjcj4

Exploring lexical, syntactic, and semantic features for Chinese textual entailment in NTCIR RITE evaluation tasks

Wei-Jie Huang, Chao-Lin Liu
2015 Soft Computing - A Fusion of Foundations, Methodologies and Applications  
We conducted more experiments with the test data of NTCIR-9 RITE, with good results.  ...  syntactic parsing, named-entity recognition, and near synonym recognition were employed, and features like counts of common words, statement lengths, negation words, and antonyms were considered to judge the entailment  ...  The first was to compare the similarity between the parse trees of T and H with the same method (the DICE coefficient) that we used to compare the definition trees of different senses as explained in Sect  ... 
doi:10.1007/s00500-015-1629-1 fatcat:2jhermxnmzedtd5wyll7notjm4
« Previous Showing results 1 — 15 out of 70,719 results