Question analysis: How Watson reads a clue

Lally, Adam; Prager, John; McCord, Michael C.; Boguraev, Branimir; Patwardhan, Siddharth V.; Fan, Jialu; Fodor, Paul; Chu-Carroll, Jennifer

doi:10.1147/jrd.2012.2184637

Cited by 116 publications

(84 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use predicate-argument structure (PAS) (McCord et al, 2012) derived from dependency parses produced by the English Slot Grammar parser (McCord, 1990). In addition to normalizing across different tree structures expressing essentially the same meaning, PAS provides a simplified view over 'raw' syntactic trees, gathering all arguments to a predicate from local, and distant, parse tree nodes (see (McCord et al, 2012) for details).…”

Section: E1mentioning

confidence: 99%

“…In addition to normalizing across different tree structures expressing essentially the same meaning, PAS provides a simplified view over 'raw' syntactic trees, gathering all arguments to a predicate from local, and distant, parse tree nodes (see (McCord et al, 2012) for details). Figures 1 and 2 show the PASes for examples e2 and e3.…”

Section: E1mentioning

confidence: 99%

“…However, they assume that the questions are referring to some concept and the problem is to uncover which one, whereas the implicit entity mention problem requires us to first check whether a particular sentence/phrase has a mention of an entity at all. Furthermore, question answering systems benefit from the presence of pronouns, nouns, and noun phrases in the questions and the candidate answers to derive helpful syntactic and semantic features (Lally et al, 2012) (Wang, 2006), while phrases with implicit entity mentions may not contain such features.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Proceedings of the Fourth Joint Conference on Lexical and Computational Semantics

2015

View full text Add to dashboard Cite

will take place in Denver on June 4 and 5 and is colocated with SemEval and NAACL. As in 2014 at COLING, also on this occasion *SEM and SemEval chose to coordinate their programs by featuring a joint invited talk. In this way, *SEM aims to bring together the ACL SIGLEX and ACL SIGSEM communities.The acceptance rate of *SEM 2015 was quite competitive: out of 98 submissions, we accepted 36 papers for an overall acceptance of 37%. The acceptance rate of long paper that were accepted for oral presentation (18 out of 62) is 29%. The papers cover a wide range of topics including distributional semantics; lexical semantics and lexical acquisition; formal and linguistic semantics; discourse semantics; lexical resources, linked data and ontologies; semantics for applications; and extra-propositional semantics: sentiment and figurative meaning.The *SEM 2015 program consists of oral presentations for selected long papers and a poster session for long and short papers.Day One, June 4th:• Joint *SEM SemEval keynote talk by Marco Baroni;• Oral presentation sessions on distributional semantics, lexical semantics, and extra-propositional semantics;• Poster session.Day Two, June 5th:• Keynote talk by Preslav Natkov;• Oral presentation sessions on semantics for applications, lexical resources and ontologies, formal semantics, and discourse semantics;• *SEM Best Paper Award.We cannot finish without saying that *SEM 2015 would not have been possible without the considerable efforts of our area chairs, their reviewers, and the computational semantics community in general.We hope you will enjoy *SEM 2015, Distributional semantic methods have some a priori appeal as models of human meaning acquisition, because they induce word representations from contextual distributions naturally occurring in corpus data without need for supervision. However, learning the meaning of a (concrete) word also involves establishing a link between the word and its typical visual referents, which is beyond the scope of classic, text-based distributional semantics. Since recently several proposals have been put forward about how to induce multimodal word representations from linguistic and visual contexts, it is natural to ask if this line of work, besides its practical implications, can help us to develop more realistic, grounded models of human word learning within the distributional semantics framework.In my talk, I will report about two studies in which we used multimodal distributional semantics (MDS) to simulate human word learning. In one study, we first measured the ability of subjects to link a nonce word to relevant linguistic and visual associates when prompted only by exposure to minimal corpus evidence about it. We then simulated the same task with an MDS model, finding its behavior remarkably similar to that of subjects. In the second study, we constructed a corpus in which child-directed speech is aligned with real-life pictures of the objects mentioned by care-givers. We then trained our MDS model on these data, and inspected the generaliza...

show abstract

Section: E1mentioning

confidence: 99%

Section: E1mentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Proceedings of the Fourth Joint Conference on Lexical and Computational Semantics

2015

View full text Add to dashboard Cite

show abstract

“…Questions may be filtered through an upfront classifier based on machine-learning techniques, parsing (Hermjakob, 2001), regular expressions and syntactic rules, or hybrid methods (Lally et al, 2012). To achieve that, a question analysis process might precede, which may involve detecting lexical answer types, question targets or the question focus.…”

Section: Introductionmentioning

confidence: 99%

BioNLP 2017

Cohen¹,

Demner‐Fushman²,

Ananiadou³

et al. 2017

View full text Add to dashboard Cite

According to the Association for Computational Linguistics guidelines on special interest groups (SIGs), The function of a SIG is to encourage interest and activity in specific areas within the ACL's field [1]. Is the SIGBioMed special interest group "within the ACL's field"? The titles of this year's papers suggest that it is, in that the current interest in deep learning in its many and varied manifestations is mirrored in those titles. Do those papers cover a specific area? They do, and in doing so, they demonstrate one of the great satisfactions of working in biomedical natural language processing.One of the joys of involvement in the biomedical natural language processing community is seeing the development of research with clinical applications. As examples of such work being presented at BioNLP 2017, we would like to point out the two papers that discuss the application of natural language processing to the diagnosis of neurological disorders. Bhatia et al. [2] describe an approach to using speech processing in the assessment of patients with amyotrophic lateral sclerosis (also known as Lou Gehrig's disease), one of the more horrific motor neuron diseases. Good assessment of amyotrophic lateral sclerosis patients is important for a number of reasons, including the fact that accurate tracking of the inevitable deterioration that is a hallmark of this disease gives patients and their families the possibility of purposeful planning for the attendant disability and death. However, current methodologies for evaluating the status of amyotrophic lateral sclerosis patients necessarily involve expensive equipment and highly trained personnel; when further developed, this methodology could make such evaluation much more, and more frequently, available to ALS patients. The fact that the work reported here involves a speech modality is especially exciting, as speech-related indicators of future ALS can be present long before diagnosis. The paper uses measurements of phonological features of speech and their divergence from a baseline, and demonstrates correlation with physiological measures.Adams et al. [3] describe work on detecting and categorizing word production errors associated with anomia, a particular kind of inability to find words. Screening for anomia is important because anomia is a symptom of stroke, but it is difficult and time-consuming to do, and therefore is not done as often as it should be. Automatic detection of anomia could be a nice enabler of improved care for stroke victims, but it is made difficult due to the subtlety of the phonological and semantic judgments that have to be made when assessing the phenomenon. The paper uses a combination of language modeling and phonologically-based edit distance calculation to approach the task, applying these techniques to data from the AphasiaBank collection of transcribed aphasic and healthy speech.Although we have summarized only these two examples that address neurological disorders, there are several other papers on the use of natural language proc...

show abstract

“…There have been a series of studies investigating the expected answer type of a question in different contexts such as Li and Roth (2002), Lally et al (2012), and Balog and Neumayer (2012). Most of these approaches classify the questions into a small set of types.…”

mentioning

confidence: 99%

Improving Semantic Parsing via Answer Type Inference

Yavuz¹,

Gür²,

Su³

et al. 2016

Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

In this work, we show the possibility of inferring the answer type before solving a factoid question and leveraging the type information to improve semantic parsing. By replacing the topic entity in a question with its type, we are able to generate an abstract form of the question, whose answer corresponds to the answer type of the original question. A bidirectional LSTM model is built to train over the abstract form of questions and infer their answer types. It is also observed that if we convert a question into a statement form, our LSTM model achieves better accuracy. Using the predicted type information to rerank the logical forms returned by AgendaIL, one of the leading semantic parsers, we are able to improve the F1-score from 49.7% to 52.6% on the WE-BQUESTIONS data.

show abstract

Question analysis: How Watson reads a clue

Cited by 116 publications

References 23 publications

Proceedings of the Fourth Joint Conference on Lexical and Computational Semantics

Proceedings of the Fourth Joint Conference on Lexical and Computational Semantics

BioNLP 2017

Improving Semantic Parsing via Answer Type Inference

Contact Info

Product

Resources

About