Báo cáo khoa học: "Semantic Discourse Segmentation and Labeling for Route Instructions" potx

... Workshop, pages 31–36,Sydney, July 2006.c2006 Association for Computational LinguisticsSemantic Discourse Segmentation and Labeling for Route InstructionsNobuyuki ShimizuDepartment of Computer ... detec-tion and tracking, and one segment of the text.In order to annotate unambiguously, we need todetect and track both landmarks and actions. Alandmark is a hallway or a door, and an actionis ... ”room”, ”doorway” and their plural forms map to DOOR, and the or-dinal number 1 will be represented by ”ﬁrst” and ”1st”, and so on.5 DatasetAs noted, we have 427 route instructions, and theaverage...

Báo cáo khoa học: "SenseClusters: Unsupervised Clustering and Labeling of Similar Contexts" potx

... Interactive Poster and Demonstration Sessions,pages 105–108, Ann Arbor, June 2005.c2005 Association for Computational LinguisticsSenseClusters: Unsupervised Clustering and Labeling of Similar ... features and unsupervised clustering.Our initial work (Purandare and Pedersen, 2004)focused on word sense discrimination, which takesas input contexts that each contain a given targetword, and ... represented using ﬁrst order (Pedersen and Bruce, 1997) or second order (Sch¨utze, 1998) rep-resentations. SenseClusters provides extensive sup-port for both, and allows for them to be applied in awider...

Tài liệu Báo cáo khoa học: "Unsupervised Discourse Segmentation of Documents with Inherently Parallel Structure" pdf

... interfaces and im-prove the performance of summarization and in-formation retrieval systems. Discourse segmentation of the documents com-posed of parallel parts is a novel and challeng-ing problem, ... 2010.c2010 Association for Computational LinguisticsUnsupervised Discourse Segmentation of Documents with Inherently Parallel StructureMinwoo Jeong and Ivan TitovSaarland UniversitySaarbr¨ucken, ... of ane-learning system and a podcast search engine for ESL learners.2 Related Work Discourse segmentation has been an active areaof research (Hearst, 1994; Utiyama and Isahara,2001; Galley...

Tài liệu Báo cáo khoa học: "Joint Word Segmentation and POS Tagging using a Single Perceptron" docx

... Daumé III and Marcu,2005; Finkel et al., 2006) and for specific problemssuch as language modeling and utterance classifica-tion (Saraclar and Roark, 2005) and labeling and chunking (Shimizu and Haas, ... tagger, and the best output is selected using the overall POS- segmentation probability score. In this system, thedecoding for word segmentation and POS taggingare still performed separately, and ... sentence, and T is the size of the tagset (T = 1 for pure word segmentation) . It workedwell for word segmentation alone (Zhang and Clark,2007), even with an agenda size as small as 8, and a simple...

Báo cáo khoa học: "Joint Hebrew Segmentation and Parsing using a PCFG-LA Lattice Parser" docx

... extendedinformation available in them.Gold Segmentation and Tagging To assess theadequacy of the Berkeley parser for Hebrew, we per-formed baseline experiments in which either gold segmentation and ... an F-measure of about 88.8% for thegold segmentation and tagging, and about 82.8% for gold segmentation only. This shows the adequacyof the PCFG-LA methodology for parsing the He-brew treebank, ... Nouns(which should agree on Gender and Number and deﬁniteness), and between Subjects and Verbs(which should agree on Gender and Number).3 PCFG-LA Grammar EstimationKlein and Manning (2003) demonstrated...

Báo cáo khoa học: "Semantic Class Induction and Coreference Resolution" pptx

... over the baseline for both data sets, whereasME and NB perform signiﬁcantly worse.9Addi-tional experiments are needed to determine the rea-son for ME and NB’s poor performance.In an attempt ... matching operations to be performedon the given pair of NPs, NPx and NPy10, including(1) exact string match for pronouns, proper nouns, and non-pronominal NPs (both before and after de-terminers ... city,administration, society, island, communityLOCATIO N dry land, region, landmass, body of water,geographical area, geological formationTable 1: List of keywords used in WordNet search for generating WN...

Báo cáo khoa học: "Semantic Caseframe Parsing and Syntactic Generality" ppt

... be handled by pretending they are a form of prepos=t=onal phrase The regular and predictable relattonsn~p between s~mple statements. ¢~uestions and relalwe clauses and between act=ve and ... The sOlutions, therefore, rested on separating the syntact=c and semanttc reformat=on. Plume already incorporates just me separation of syntax and semantics necessary for syntactic generahly ... queried. On the other hand. apart from var)ations in the order ancl placement of marked cases, there is only one standard constituent order for =mperatives and only two for declaratives (corresponding...

Tài liệu Báo cáo khoa học: "Unsupervized Word Segmentation: the case for Mandarin Chinese" doc

... Association for Computational Linguistics, pages 383–387,Jeju, Republic of Korea, 8-14 July 2012.c2012 Association for Computational LinguisticsUnsupervized Word Segmentation: the case for Mandarin ... fromthe Segmentation bake-off II (Emerson, 2005) and define a more precise topline for the taskusing cross-trained supervized system availableoff-the-shelf (Zhang and Clark, 2010; Zhao and Kit, ... rely on punctuation and char-acter encoding information (such as distinguishingLatin and Chinese characters). However, they opti-mize their parameter for each setting. We thereforeconsider that...

Tài liệu Báo cáo khoa học: "A New Dataset and Method for Automatically Grading ESOL Texts" pdf

... Discovery and Data Mining(KDD), pages 133–142. ACM.T.K. Landauer and P.W. Foltz. 1998. An introduction tolatent semantic analysis. Discourse processes, pages259–284.T.K. Landauer, D. Laham, and ... a ceiling for the perfor-mance of our system, we calculate the average corre-lation between the CLC and the examiners’ scores, and ﬁnd an upper bound of 0.796 and 0.792 Pear-son’s and Spearman’s ... analyses of the perfor-180mance of individual systems, as yet there is no pub-lically available shared dataset for training and test-ing such systems and comparing their performance.As it is...

Tài liệu Báo cáo khoa học: "Nonlinear Evidence Fusion and Propagation for Hyponymy Relation Mining" pdf

... pages 538–549, Auckland, New Zealand. B. Van Durme and M. Pasca. 2008. Finding cars, god-desses and enzymes: Parametrizable acquisition of labeled instances for open-domain information ex-traction. ... Borkovsky, A M. Popescu and V. Vyas. 2009. Web-Scale Distributional Similarity and Entity Set Expansion. EMNLP’2009. Singapore. P. Pantel and D. Ravichandran. 2004. Automatically Labeling Semantic ... similar terms. For example, Mount Vesuvius and Kilimanjaro are volcanoes and Lhotse is similar to them, but Lhotse is not a volcano. Therefore we should be very conservative and careful in...

Xem thêm