Báo cáo khoa học: "Identifying Word Translations in Non-Parallel Texts" potx

... co-occurrences of German word pairs in the German corpus. As a starting point, word order in the two matrices was chosen such that word n in the German matrix was the translation of word n in the English ... German words axe in corresponding order. Word n in the English matrix is then the translation of word n in the German matrix. 3 Simulation A simulation experiment was conducted in order ... for checking the consistency of terminological usage in translations, and for assisting the terminological work of translators and interpreters. However, despite serious efforts in the compilation...

Báo cáo khoa học: "Identifying Word Translations from Comparable Corpora Using Latent Topic Models" potx

... whichdetermine how words in documents might be gener-ated. Fitting a generative model means ﬁnding thebest set of those latent variables in order to explainthe observed data. Within that setting, ... knowledge from word- topic distributions outperform methodsbased on similarity measures in the original word- document space. The best results, ob-tained by combining knowledge from word- topic ... additional translations of the context words. In contrast, ourmethod does not bootstrap on language pairs thatshare morphology, cognates or similar words.Some attempts of obtaining translations usingcross-lingual...

Tài liệu Báo cáo khoa học: "Identifying Linguistic Structure in a Quantitative Analysis of Dialect Pronunciation" docx

... different vowels found in the data204 vowel positions are investigated, where a vowelposition is, e.g., the ﬁrst vowel in the word ’Wash-ington’ or the second vowel in the word ’thirty’.Factor ... of linguistic structure in the aggregate analysis is based on the analysis ofthe pronunciation of the vowels found in the data set. In work presented in this paper the identiﬁcationof linguistic ... explaining 35% of the total amount of vari-ance. The main drawback of applying this technique in dialectometry is that it is not directly related to theaggregate analysis, but is rather an independent...

Báo cáo khoa học: "Identifying Repair Targets in Action Control Dialogue" pdf

... which we adopted in ourframework. We will extend the grounding actmodel by introducing degree of groundedness thathave a quaternary distinction instead of the orig-inal binary distinction. The ... focused on the di-alogue involving only utterances. In this paper,we discuss misunderstanding problem in the di-alogue involving participant’s actions as well asutterances. In particular, we focus ... re-sponder succeeded in interpretation.E.g., turning to the right as the speakerintended in response to the request “turnto the right.”3.3 Partial and mid-DU Grounding In Traum’s grounding model, the...

Tài liệu Báo cáo khoa học: "User Participation Prediction in Online Forums" potx

... using the bag-of -word model sincethat would require the exact words to appear in thetraining set. In order to take advantage of the topic level in- formation while not losing the “ﬁne-grained” ... capture word relationship semantically. To illustrate thewords inside latent topics in the LDA model in- ferred from online forums, we show in Table 2 thetop words in 3 out of 20 latent topics inferred ... mutual information between users afterdoing an average-link clustering on their pairwisemutual information. In a clean clustering, intra-cluster mutual information should be high, whileinter-cluster...

Báo cáo khoa học: "Unsupervised Word Alignment with Arbitrary Features" potx

... NeurAlign:combining word alignments using neural networks. In Proc. of HLT-EMNLP.T. Berg-Kirkpatrick, A. Bouchard-Cˆot´e, J. DeNero, andD. Klein. 2010. Painless unsupervised learning withfeatures. In ... Lin. 2010. Discriminative word alignment by linear modeling. Computational Lin-guistics, 36(3):303–339.A. Lopez. 2008. Tera-scale translation models via pat-tern matching. In Proc. of COLING.R. ... Discrimina-tive word alignment with a function word reorderingmodel. In Proc. of EMNLP.N. A. Smith and J. Eisner. 2005. Contrastive estimation:training log-linear models on unlabeled data. In...

Báo cáo khoa học: "Classifying Semantic Relations in Bioscience Texts" pot

... sur-rounding forms of the stem bind which signifyentities that can enter into molecular binding re-lationships. In Srinivasan and Rindﬂesch (2002)MeSH term co-occurrences within MEDLINE ar-ticles ... containing it.We used a large domain-speciﬁc lexical hi-erarchy (MeSH, Medical Subject Headings3) tomap words into semantic categories. There areabout 19,000 unique terms in MeSH and 15 mainsub-hierarchies, ... ofsome linguistic expression. These linguistic pat-terns could in principle distinguish between differ-ent relations, but instead are usually used to iden-tify examples of one relation. In the...

Báo cáo khoa học: "Using Linguistic Knowledge in Automatic Abstracting" potx

... topics according to the reader's interest by motivating the topics, describing en- tities and defining concepts. We have defined our method of automatic abstracting by study- ing a corpus ... topics according to the reader's inter- est by motivating the topics, describing entities, defining concepts and so on. This kind of ab- stract could be used in tasks such as accessing the ... uments containing the information reported in the abstract. In a previous study (Saggion and Lapalme, 1998), we have shown that 72% of the information in professional abstracts lies in ti- tles,...

Báo cáo khoa học: "Hedge classiﬁcation in biomedical texts with a weakly supervised selection of keywords" doc

... keyword) selectionTo handle the inherent noise in the training datasetthat originates from its weakly supervised construc-tion, we applied the following feature selection pro-cedure. The main ... very few useful keywords were eliminated andthis indicated that our feature selection procedurewas capable of distinguishing useful keywords fromnoise (i.e. keywords having a very high specula-tive ... that,as we said earlier, hedging is expressed via key-words in natural language texts; and often severalkeywords are present in a single sentence. Theseed set Sspeccontained either suggest or...

Báo cáo khoa học: "Dynamic Strategy Selection in Flexible Parsing" potx

... interface, including the invocation of parsing strategies, dictionanes and concepts, rather than requiring any domain adaptations by the interface system itself. With these goals in mind, we ... clustering domain concepts into functionally useful categories for user interaction. Semantic grammars, like case systems, can bring domain knowledge to bear in dissmbiguatmg word meaningS. ... string PATTERN-MATCH to the initial segment of the input using only the patterns previously indexed as corresponding to command words/phrases in imperative constructions. Patterns contain...

Xem thêm