DOCSLIB.ORG

Methods for Using Textual Entailment in Open-Domain Question Answering

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Methods for Using Textual Entailment in Open-Domain Question Answering Methods for Using Textual Entailment in Open-Domain Question Answering Sanda Harabagiu and Andrew Hickl Language Computer Corporation 1701 North Collins Boulevard Richardson, Texas 75080 USA [email protected] Abstract better constrain the answers to individual ques- tions (cf. (Prager et al., 2004)). While different in Work on the semantics of questions has their approach, each of these frameworks seeks to argued that the relation between a ques- approximate the forms of semantic inference that tion and its answer(s) can be cast in terms will allow them to identify valid textual answers to of logical entailment. In this paper, we natural language questions. demonstrate how computational systems Recently, the task of automatically recog- designed to recognize textual entailment nizing one form of semantic inference – tex- can be used to enhance the accuracy of tual entailment – has received much attention current open-domain automatic question from groups participating in the 2005 and 2006 answering (Q/A) systems. In our experi- PASCAL Recognizing Textual Entailment (RTE) ments, we show that when textual entail- Challenges (Dagan et al., 2005). 1 ment information is used to either filter or As currently defined, the RTE task requires sys- rank answers returned by a Q/A system, tems to determine whether, given two text frag- accuracy can be increased by as much as ments, the meaning of one text could be reason- 20% overall. ably inferred, or textually entailed, from the mean- ing of the other text. We believe that systems de- 1 Introduction veloped specifically for this task can provide cur- Open-Domain Question Answering (Q/A) sys- rent question-answering systems with valuable se- tems return a textual expression, identified from mantic information that can be leveraged to iden- a vast document collection, as a response to a tify exact answers from ranked lists of candidate question asked in natural language. In the quest answers. By replacing the pairs of texts evaluated for producing accurate answers, the open-domain in the RTE Challenge with combinations of ques- Q/A problem has been cast as: (1) a pipeline of tions and candidate answers, we expect that textual linguistic processes pertaining to the processing entailment could provide yet another mechanism of questions, relevant passages and candidate an- for approximating the types of inference needed swers, interconnected by several types of lexico- in order answer questions accurately. semantic feedback (cf. (Harabagiu et al., 2001; In this paper, we present three different methods Moldovan et al., 2002)); (2) a combination of for incorporating systems for textual entailment language processes that transform questions and into the traditional Q/A architecture employed by candidate answers in logic representations such many current systems. Our experimental results that reasoning systems can select the correct an- indicate that (even at their current level of per- swer based on their proofs (cf. (Moldovan et al., formance) textual entailment systems can substan- 2003)); (3) a noisy-channel model which selects tially improve the accuracy of Q/A, even when no the most likely answer to a question (cf. (Echi- other form of semantic inference is employed. habi and Marcu, 2003)); or (4) a constraint sat- The remainder of the paper is organized as fol- isfaction problem, where sets of auxiliary ques- tions are used to provide more information and 1http://www.pascal-network.org/Challenges/RTE 905 Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the ACL, pages 905–912, Sydney, July 2006. c 2006 Association for Computational Linguistics QUESTION ANSWERING SYSTEM Answers Expected Keywords Ranked List of Paragraphs Answer Type AUTO−QUAB Question Passage Generation Answer Question Processing Retrieval Processing Module Module List of Questions Module (QP) (PR) (AP) Documents Entailed Paragraphs List of Entailed Paragraphs TEXTUAL ENTAILMENT Ranked List of Answers Entailed Questions Method 1 TEXTUAL TEXTUAL ENTAILMENT Answers−M1 ENTAILMENT Answers−M2 Method 2 Method 3 Answers−M3 Figure 1: Integrating Textual Entailment in Q/A. lows. Section 2 describes the three methods of techniques. using textual entailment in open-domain question As illustrated in Figure 1, most open-domain answering that we have identified, while Section 3 Q/A systems generally consist of a sequence of presents the textual entailment system we have three modules: (1) a question processing (QP) used. Section 4 details our experimental methods module; (2) a passage retrieval (PR) module; and and our evaluation results. Finally, Section 5 pro- (3) an answer processing (AP) module. Questions vides a discussion of our findings, and Section 6 are first submitted to a QP module, which extracts summarizes our conclusions. a set of relevant keywords from the text of the question and identifies the question's expected an- 2 Integrating Textual Entailment in swer type (EAT). Keywords – along with the ques- Question Answering tion's EAT – are then used by a PR module to re- In this section, we describe three different meth- trieve a ranked list of paragraphs which may con- ods for integrating a textual entailment (TE) sys- tain answers to the question. These paragraphs are tem into the architecture of an open-domain Q/A then sent to an AP module, which extracts an ex- system. act candidate answer from each passage and then ranks each candidate answer according to the like- Work on the semantics of questions (Groe- lihood that it is a correct answer to the original nendijk, 1999; Lewis, 1988) has argued that the question. formal answerhood relation found between a ques- tion and a set of (correct) answers can be cast Method 1. In Method 1, each of a ranked list of in terms of logical entailment. Under these ap- answers that do not meet the minimum conditions proaches (referred to as licensing by (Groenendijk, for TE are removed from consideration and then 1999) and aboutness by (Lewis, 1988)), p is con- re-ranked based on the entailment confidence (a sidered to be an answer to a question ?q iff ?q logi- real-valued number ranging from 0 to 1) assigned cally entails the set of worlds in which p is true(i.e. by the TE system to each remaining example. The ?p). While the notion of textual entailment has system then outputs a new set of ranked answers been defined far less rigorously than logical en- which do not contain any answers that are not en- tailment, we believe that the recognition of textual tailed by the user's question. entailment between a question and a set of candi- Table 1 provides an example where Method 1 date answers – or between a question and ques- could be used to make the right prediction for a set tions generated from answers – can enable Q/A of answers. Even though A1 was ranked in sixth systems to identify correct answers with greater position, the identification of a high-confidence precision than current keyword- or pattern-based positive entailment enabled it to be returned as the 906 top answer. In contrast, the recognition of a neg- tify a set of answer passages that contain correct ative entailment for A2 caused this answer to be answers to the original question. For example, in dropped from consideration altogether. Table 2, we find that entailment between questions Q1: “What did Peter Minuit buy for the equivalent of $24.00?” indicates the correctness of a candidate answer: Rank1 TE Rank2 Answer Text th st Q2 1 2 A1 6 YES 1 Everyone knows that, back in 1626, Peter Mi- here, establishing that entails AGQ and AGQ (0.89) nuit bought Manhattan from the Indians for $24 (but not AGQ3 or AGQ4) enables the system to se- worth of trinkets. st A2 1 NO – In 1626, an enterprising Peter Minuit flagged lect A2 as the correct answer. (0.81) down some passing locals, plied them with beads, cloth and trinkets worth an estimated When at least one of the AGQs generated by $24, and walked away with the whole island. the AutoQUAB module is entailed by the original Table 1: Re-ranking of answers by Method 1. question, all AGQs that do not reach TE are fil- Method 2. Since AP is often a resource- tered from consideration; remaining passages are intensive process for most Q/A systems, we ex- assigned an entailment confidence score and are pect that TE information can be used to limit the sent to the AP module in order to provide an ex- number of passages considered during AP. As il- act answer to the question. Following this pro- lustrated in Method 2 in Figure 1, lists of passages cess, candidate answers extracted from the AP retrieved by a PR module can either be ranked (or module were then re-associated with their AGQs filtered) using TE information. Once ranking is and resubmitted to the TE system (as in Method complete, answer extraction takes place only on 1). Question-answer pairs deemed to be posi- the set of entailed passages that the system consid- tive instances of entailment were then stored in a ers likely to contain a correct answer to the user's database and used as additional training data for question. the AutoQUAB module. When no AGQs were Method 3. In previous work (Harabagiu et al., found to be entailed by the original question, how- 2005b), we have described techniques that can be ever, passages were ranked according to their en- used to automatically generate well-formed natu- tailment confidence and sent to AP for further pro- ral language questions from the text of paragraphs cessing and validation. retrieved by a PR module. In our current system, sets of automatically-generated questions (AGQ) 3 The Textual Entailment System are created using a stand-alone AutoQUAB gen- Processing textual entailment, or recognizing eration module, which assembles question-answer whether the information expressed in a text can be pairs (known as QUABs) from the top-ranked pas- inferred from the information expressed in another sages returned in response to a question.
Load more

Recommended publications
  • Are Annotators' Word-Sense-Disambiguation
    J. Hlavácovᡠ(Ed.): ITAT 2017 Proceedings, pp. 5–14 CEUR Workshop Proceedings Vol. 1885, ISSN 1613-0073, c 2017 S. Cinková, A. Vernerová Are Annotators’ Word-Sense-Disambiguation Decisions Affected by Textual Entailment between Lexicon Glosses? Silvie Cinková, Anna Vernerová Charles University, Faculty of Mathematics and Physics Institute of Formal and Applied Linguistics Malostranské námestíˇ 25 118 00 Praha 1 Czech Republic ufal.mff.cuni.cz {cinkova,vernerova}@ufal.mff.cuni.cz Abstract: We describe an annotation experiment com- verb on the one hand, and those related to the lexicograph- bining topics from lexicography and Word Sense Disam- ical design of PDEV, such as semantic relations between biguation. It involves a lexicon (Pattern Dictionary of En- implicatures within a lemma or denotative similarity of the glish Verbs, PDEV), an existing data set (VPS-GradeUp), verb arguments, on the other hand (see Section 3 for defi- and an unpublished data set (RTE in PDEV Implicatures). nitions and examples). The aim of the experiment was twofold: a pilot annota- This paper focuses on a feature related to PDEV’s de- tion of Recognizing Textual Entailment (RTE) on PDEV sign (see Section 2.3), namely on textual entailment be- implicatures (lexicon glosses) on the one hand, and, on tween implicatures in pairs of patterns of the same lemma the other hand, an analysis of the effect of Textual Entail- entry (henceforth colempats, see Section 3.1 for definition ment between lexicon glosses on annotators’ Word-Sense- and more detail). Disambiguation decisions, compared to other predictors, We pairwise compare all colempats, examining their such as finiteness of the target verb, the explicit presence scores in the graded decision annotation with respect to of its relevant arguments, and the semantic distance be- how much they compete to become the most appropriate tween corresponding syntactic arguments in two different pattern, as well as the scores of presence of textual entail- patterns (dictionary senses).
    [Show full text]
  • Textual Inference for Machine Comprehension Martin Gleize
    Textual Inference for Machine Comprehension Martin Gleize To cite this version: Martin Gleize. Textual Inference for Machine Comprehension. Computation and Language [cs.CL]. Université Paris Saclay (COmUE), 2016. English. NNT : 2016SACLS004. tel-01317577 HAL Id: tel-01317577 https://tel.archives-ouvertes.fr/tel-01317577 Submitted on 18 May 2016 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Le cas échéant, logo de l’établissement co-délivrant le doctorat en cotutelle internationale de thèse , sinon mettre le logo de l’établissement de préparation de la thèse (UPSud, HEC, UVSQ, UEVE, ENS Cachan, Polytechnique, IOGS, …) NNT : 2016SACLS004 THÈSE DE DOCTORAT DE L’UNIVERSITÉ PARIS-SACLAY PRÉPARÉE À L'UNIVERSITÉ PARIS-SUD ECOLE DOCTORALE N° 580 Sciences et technologies de l'information et de la communication (STIC) Spécialité de doctorat : Informatique Par M. Martin Gleize Textual Inference for Machine Comprehension Thèse présentée et soutenue à Orsay, le 7 janvier 2016 Composition du Jury : M. Yvon François Professeur, Université Paris-Sud Président, Examinateur Mme Gardent Claire Directeur de recherche CNRS, LORIA Rapporteur M. Magnini Bernardo Senior Researcher, FBK Rapporteur M. Piwowarski Benjamin Chargé de recherche CNRS, LIP6 Examinateur Mme Grau Brigitte Professeur, ENSIIE Directeur de thèse Abstract With the ever-growing mass of published text, natural language understanding stands as one of the most sought-after goal of artificial intelligence.
    [Show full text]
  • Repurposing Entailment for Multi-Hop Question Answering Tasks
    Repurposing Entailment for Multi-Hop Question Answering Tasks Harsh Trivedi|, Heeyoung Kwon|, Tushar Khot♠, Ashish Sabharwal♠, Niranjan Balasubramanian| | Stony Brook University, Stony Brook, U.S.A. fhjtrivedi,heekwon,[email protected] ♠ Allen Institute for Artificial Intelligence, Seattle, U.S.A. ftushark,[email protected] Abstract Question Answering (QA) naturally reduces to an entailment problem, namely, verifying whether some text entails the answer to a ques- tion. However, for multi-hop QA tasks, which require reasoning with multiple sentences, it remains unclear how best to utilize entailment models pre-trained on large scale datasets such as SNLI, which are based on sentence pairs. We introduce Multee, a general architecture Figure 1: An example illustrating the challenges in us- that can effectively use entailment models for ing sentence-level entailment model for multi-sentence multi-hop QA tasks. Multee uses (i) a local reasoning needed for QA, and the high-level approach module that helps locate important sentences, used in Multee. thereby avoiding distracting information, and level, whereas question answering requires verify- (ii) a global module that aggregates informa- ing whether multiple sentences, taken together as tion by effectively incorporating importance weights. Importantly, we show that both mod- a premise, entail a hypothesis. ules can use entailment functions pre-trained There are two straightforward ways to address on a large scale NLI datasets. We evaluate per- this mismatch: (1) aggregate independent entail- formance on MultiRC and OpenBookQA, two ment decisions over each premise sentence, or (2) multihop QA datasets. When using an entail- make a single entailment decision after concate- ment function pre-trained on NLI datasets, nating all premise sentences.
    [Show full text]
  • Techniques for Recognizing Textual Entailment and Semantic Equivalence
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by EPrints Complutense Techniques for Recognizing Textual Entailment and Semantic Equivalence Jes´us Herrera, Anselmo Pe˜nas, Felisa Verdejo Departamento de Lenguajes y Sistemas Inform´aticos Universidad Nacional de Educaci´ona Distancia Madrid, Spain {jesus.herrera, anselmo, felisa}@lsi.uned.es Abstract. After defining what is understood by textual entailment and semantic equivalence, the present state and the desirable future of the systems aimed at recognizing them is shown. A compilation of the currently implemented techniques in the main Recognizing Textual Entailment and Semantic Equivalence systems is given. 1 Introduction The concept “textual entailment” is used to indicate the state in which the semantics of a natural language written text can be inferred from the semantics of another one. More specifically, if the truth of an enunciation entails the truth of another enunciation. For example, given the texts: 1. The three-day G8 meeting will take place in Scotland. 2. The Group of Eight summit will last three days. it is clear that the semantics of the second one can be inferred from the semantics of the first one; then, it is said that textual entailment exists between both texts. Textual entailment is a directional relationship: in the example above, the first statement entails the second one, but this entailment is not given in the opposite direction. The recognition of textual entailment requires a processing at the lexical level (for example, synonymy between meeting and summit or between G8 and Group of Eight), as well as at the syntactic level and the sentence semantic level.
    [Show full text]
  • Automatic Evaluation of Summary Using Textual Entailment
    Automatic Evaluation of Summary Using Textual Entailment Pinaki Bhaskar Partha Pakray Department of Computer Science & Department of Computer Science & Engineering, Jadavpur University, Engineering, Jadavpur University, Kolkata – 700032, India Kolkata – 700032, India [email protected] [email protected] summary is very much needed when a large Abstract number of summaries have to be evaluated, spe- cially for multi-document summaries. For sum- This paper describes about an automatic tech- mary evaluation we have developed an automat- nique of evaluating summary. The standard ed evaluation technique based on textual entail- and popular summary evaluation techniques or ment. tools are not fully automatic; they all need Recognizing Textual Entailment (RTE) is one some manual process. Using textual entail- of the recent research areas of Natural Language ment (TE) the generated summary can be evaluated automatically without any manual Processing (NLP). Textual Entailment is defined evaluation/process. The TE system is the as a directional relationship between pairs of text composition of lexical entailment module, lex- expressions, denoted by the entailing “Text” (T) ical distance module, Chunk module, Named and the entailed “Hypothesis” (H). T entails H if Entity module and syntactic text entailment the meaning of H can be inferred from the mean- (TE) module. The syntactic TE system is ing of T. Textual Entailment has many applica- based on the Support Vector Machine (SVM) tions in NLP tasks, such as Summarization, In- that uses twenty five features for lexical simi- formation Extraction, Question Answering, In- larity, the output tag from a rule based syntac- formation Retrieval. tic two-way TE system as a feature and the outputs from a rule based Chunk Module and 2 Related Work Named Entity Module as the other features.
    [Show full text]
  • Dependency-Based Textual Entailment
    Dependency-based Textual Entailment Vasile Rus Department of Computer Science Institute for Intelligent Systems The University of Memphis Memphis, TN 38120, USA [email protected] Abstract and Hypothesis are mapped into two sets of dependen- cies: the T-set and H-set, respectively. A score is then This paper studies the role of dependency information computed that measures the degree to which the depen- for the task of textual entailment. Both the Text and Hypothesis of an entailment pair are mapped into sets dencies in the H-set are present in the T-set. Based on of dependencies and a score is computed that measures the score, an entailment decision is made. Dependen- the similarity of the two sets. Based on the score an cies are labelled relations among words in a sentence. entailment decision is made. Two experiments are con- For example, there is a subj dependency between cap- ducted to measure the impact of dependencies on the ital and Besancon in the previous example which we entailment task. In one experiment we compare the represent as the triplet (subj, capital, Besancon). The dependency-based approach with a baseline approach first term in the triplet is the name or label of the re- on a standard data set. In a second experiment, we lation, the second term represents the head word, and measure the performance on a subset of the standard the last term is the dependent word. We use MINIPAR data set. The subset is so selected to minimize the (Lin 1998), an automated dependency parser, as our effect of other factors, such as word-level information, in the performance measurement process.
    [Show full text]
  • Parser Evaluation Using Textual Entailments
    Language Resources and Evaluation manuscript No. (will be inserted by the editor) Parser Evaluation Using Textual Entailments Deniz Yuret · Laura Rimell · Aydın Han Received: date / Accepted: date Abstract Parser Evaluation using Textual Entailments (PETE) is a shared task in the SemEval-2010 Evaluation Exercises on Semantic Evaluation. The task involves recognizing textual entailments based on syntactic information alone. PETE introduces a new parser evaluation scheme that is formalism independent, less prone to annotation error, and focused on semantically rele- vant distinctions. This paper describes the PETE task, gives an error analysis of the top-performing Cambridge system, and introduces a standard entail- ment module that can be used with any parser that outputs Stanford typed dependencies. Keywords Parsing · Textual Entailments 1 Introduction Parser Evaluation using Textual Entailments (PETE) is a shared task in the SemEval-2010 Evaluation Exercises on Semantic Evaluation that involves rec- ognizing textual entailments to evaluate parser performance. Given two text Deniz Yuret and Aydın Han Ko¸cUniversity Rumelifeneri Yolu 34450 Sarıyer, Istanbul,˙ Turkey Tel.: +90-212-338-1724 Fax: +90-212-338-1548 E-mail: dyuret,[email protected] Laura Rimell Computer Laboratory William Gates Building 15 JJ Thomson Avenue Cambridge CB3 0FD, UK Tel.: +44 (0)1223 334696 E-mail: [email protected] 2 Deniz Yuret et al. fragments called “text” (T) and “hypothesis” (H), recognizing textual entail- ment (RTE) is the task of determining whether the meaning of the hypothesis is entailed (can be inferred) from the text. In contrast to general RTE tasks (Dagan et al 2009) PETE is a targeted textual entailment task that focuses on syntactic entailments: Text: The man with the hat was tired.
    [Show full text]
  • Textual Entailment Recognition Based on Dependency Analysis and Wordnet
    Textual Entailment Recognition Based on Dependency Analysis and WordNet Jesus´ Herrera, Anselmo Penas,˜ Felisa Verdejo Departamento de Lenguajes y Sistemas Informaticos´ Universidad Nacional de Educacion´ a Distancia Madrid, Spain jesus.herrera, anselmo, felisa @lsi.uned.es f g Abstract. The Recognizing Textual Entailment System shown here is based on the use of a broad-coverage parser to extract dependency relationships; in addi- tion, WordNet relations are used to recognize entailment at the lexical level. The work investigates whether the mapping of dependency trees from text and hy- pothesis give better evidence of entailment than the matching of plain text alone. While the use of WordNet seems to improve system's performance, the notion of mapping between trees here explored (inclusion) shows no improvement, sug- gesting that other notions of tree mappings should be explored such as tree edit distances or tree alignment distances. 1 Introduction Textual Entailment Recognition (RTE) aims at deciding whether the truth of a text en- tails the truth of another text called hypothesis. This concept has been the basis for the PASCAL1 RTE Challenge [3]. The system presented here is aimed at validating the hypothesis that (i) a certain amount of semantic information could be extracted from texts by means of the syntactic structure given by a dependency analysis, and that (ii) lexico-semantic information such as WordNet relations can improve RTE. In short, the techniques involved in this system are the following: – Dependency analysis of texts and hypothesises. – Lexical entailment between dependency tree nodes using WordNet. – Mapping between dependency trees based on the notion of inclusion.
    [Show full text]
  • A Probabilistic Classification Approach for Lexical Textual Entailment
    A Probabilistic Classification Approach for Lexical Textual Entailment Oren Glickman and Ido Dagan and Moshe Koppel Computer Science Department , Bar Ilan University Ramat-Gan, 52900, Iarael {glikmao, dagan, koppel}@cs.biu.ac.il Abstract The uncertain nature of textual entailment calls for its The textual entailment task – determining if a given text explicit modeling in probabilistic terms. In this paper we entails a given hypothesis – provides an abstraction of first propose a general generative probabilistic setting for applied semantic inference. This paper describes first a textual entailment. This probabilistic framework may be general generative probabilistic setting for textual considered analogous to (though different than) the entailment. We then focus on the sub-task of recognizing probabilistic setting defined for other phenomena, like whether the lexical concepts present in the hypothesis are language modeling and statistical machine translation. entailed from the text. This problem is recast as one of text Obviously, such frameworks are needed to allow for categorization in which the classes are the vocabulary principled (fuzzy) modeling of the relevant language words. We make novel use of Naïve Bayes to model the problem in an entirely unsupervised fashion. Empirical tests phenomenon, compared to utilizing ad-hoc ranking scores suggest that the method is effective and compares favorably which have no clear probabilistic interpretation. We with state-of-the-art heuristic scoring approaches. suggest that the proposed setting may provide a unifying framework for modeling uncertain semantic inferences from texts. Introduction An important sub task of textual entailment, which we term lexical entailment , is recognizing if the concepts in a Many Natural Language Processing (NLP) applications hypothesis h are entailed from a given text t, even if the need to recognize when the meaning of one text can be relations between these concepts may not be entailed from expressed by, or inferred from, another text.
    [Show full text]
  • Frame Semantics Annotation Made Easy with Dbpedia
    Frame Semantics Annotation Made Easy with DBpedia Marco Fossati, Sara Tonelli, and Claudio Giuliano ffossati, satonelli, [email protected] Fondazione Bruno Kessler - via Sommarive, 18 - 38123 Trento, Italy Abstract. Crowdsourcing techniques applied to natural language pro- cessing have recently experienced a steady growth and represent a cheap and fast, albeit valid, solution to create benchmarks and training data. Nevertheless, some particularly complex tasks such as semantic role an- notation have been rarely conducted in a crowdsourcing environment, due to their intrinsic difficulty. In this paper, we present a novel ap- proach to accomplish this task by leveraging information automatically extracted from DBpedia. We show that replacing role definitions, typi- cally meant for expert annotators, with a list of DBpedia types, makes the identification and assignment of role labels more intuitive also for non-expert workers. Results prove that such strategy improves on the standard annotation workflow, both in terms of accuracy and of time consumption. Keywords: Natural Language Processing, Frame Semantics, En- tity Linking, DBpedia, Crowdsourcing, Task Modeling 1 Introduction Frame semantics [6] is one of the theories that originate from the long strand of linguistic research in artificial intelligence. A frame can be informally defined as an event triggered by some term in a text and embedding a set of partic- ipants. For instance, the sentence Goofy has murdered Mickey Mouse evokes the Killing frame (triggered by murdered) together with the Killer and Vic- tim participants (respectively Goofy and Mickey Mouse). Such theory has led to the creation of FrameNet [2], namely an English lexical database containing manually annotated textual examples of frame usage.
    [Show full text]
  • Constructing a Textual Semantic Relation Corpus Using a Discourse Treebank
    Constructing a Textual Semantic Relation Corpus Using a Discourse Treebank Rui Wang and Caroline Sporleder Computational Linguistics Saarland University frwang,[email protected] Abstract In this paper, we present our work on constructing a textual semantic relation corpus by making use of an existing treebank annotated with discourse relations. We extract adjacent text span pairs and group them into six categories according to the different discourse relations between them. After that, we present the details of our annotation scheme, which includes six textual semantic relations, backward entailment, forward entailment, equality, contradiction, overlapping, and independent. We also discuss some ambiguous examples to show the difficulty of such annotation task, which cannot be easily done by an automatic mapping between discourse relations and semantic relations. We have two annotators and each of them performs the task twice. The basic statistics on the constructed corpus looks promising: we achieve 81.17% of agreement on the six semantic relation annotation with a .718 kappa score, and it increases to 91.21% if we collapse the last two labels with a .775 kappa score. 1. Introduction TRADICTION, and UNKNOWN. ENTAILMENT is the same Traditionally, the term ‘semantic relation’ refers to rela- as the previous YES; but NO is divided into CONTRADIC- tions that hold between the meanings of two words, e.g., TION and UNKNOWN, to differentiate cases where T and H synonymy, hypernymy, etc. These relations are usually are contradictory to each other from all the other cases. The situation-independent. Discourse relations, on the other RTE data are acquired from other natural language process- hand, tend to depend on the wider context of an utterance.
    [Show full text]
  • PATTY: a Taxonomy of Relational Patterns with Semantic Types
    PATTY: A Taxonomy of Relational Patterns with Semantic Types Ndapandula Nakashole, Gerhard Weikum, Fabian Suchanek Max Planck Institute for Informatics Saarbrucken,¨ Germany {nnakasho,weikum,suchanek}@mpi-inf.mpg.de Abstract Yet, existing large-scale knowledge bases are mostly limited to abstract binary relationships be- This paper presents PATTY: a large resource for textual patterns that denote binary relations tween entities, such as “bornIn” (Auer 2007; Bol- between entities. The patterns are semanti- lacker 2008; Nastase 2010; Suchanek 2007). These cally typed and organized into a subsumption do not correspond to real text phrases. Only the Re- taxonomy. The PATTY system is based on ef- Verb system (Fader 2011) yields a larger number of ficient algorithms for frequent itemset mining relational textual patterns. However, no attempt is and can process Web-scale corpora. It har- made to organize these patterns into synonymous nesses the rich type system and entity popu- patterns, let alone into a taxonomy. Thus, the pat- lation of large knowledge bases. The PATTY taxonomy comprises 350,569 pattern synsets. terns themselves do not exhibit semantics. Random-sampling-based evaluation shows a Goal. Our goal in this paper is to systematically pattern accuracy of 84.7%. PATTY has 8,162 compile relational patterns from a corpus, and to im- subsumptions, with a random-sampling-based pose a semantically typed structure on them. The precision of 75%. The PATTY resource is freely available for interactive access and result we aim at is a WordNet-style taxonomy of download. binary relations. In particular, we aim at patterns that contain semantic types, such as hsingeri sings 1 Introduction hsongi.
    [Show full text]