Báo cáo khoa học: "Learning Synchronous Grammars for Semantic Parsing with Lambda Calculus" docx

... 960–967, Prague, Czech Republic, June 2007. c 2007 Association for Computational Linguistics Learning Synchronous Grammars for Semantic Parsing with Lambda Calculus Yuk Wah Wong and Raymond J. Mooney Department ... sub-parse. For example, see the rule 962 FORM → state , λx 1 .state(x 1 ) FORM → by area , λx 1 .λx 2 .area(x 1 ,x 2 ) FORM → smallest FORM 1 FORM 2 , λx 1...

Ngày tải lên: 17/03/2014, 04:20

... 208–215, Prague, Czech Republic, June 2007. c 2007 Association for Computational Linguistics Learning Predictive Structures for Semantic Role Labeling of NomBank Chang Liu and Hwee Tou Ng Department ... the meanings of each label. Unlike in PropBank, arguments in NomBank can overlap with each other and with the predicate. For example: [ location U.S.] [ pred,subj,obj steelmakers]...

Ngày tải lên: 23/03/2014, 18:20

... also show more genre and content effects. For exam- ple, the sentiment enriched vectors for ghastly are truly semantic alternatives to that word, whereas the vectors without sentiment also contain some ... ver- sions of our model with only the unsupervised se- mantic component, and the full model (semantic and sentiment). We also include results for a version of our full model trai...

Ngày tải lên: 20/02/2014, 04:20

... very encour- aging: even without being particularly customized for this monolingual task, the ILP approach signif- icantly outperforms the majority sense baseline and performs as well as the state-of-the-art ... state-of-the-art system re- porting results for the same set of verbs. As with the multilingual task, the models produced contain a small number of rules (from 6, for verbs...

Ngày tải lên: 08/03/2014, 02:21

... nent provides very rich syntactic and semantic information which has not been explored before for intonation modeling. This includes, for ex- ample, the semantic role played by each seman- tic ... model without context information with p=0.0135 for break index, p=0 for both phrase accent and pitch accent. 4.3 The Rules Learned In this section we describe some typical ru...

Ngày tải lên: 08/03/2014, 05:21

... the rank of an SCFG would result in more efficient parsing algorithms, for most common parsing strategies. In the general case, normal forms with bounded rank are not admitted by SCFGs, as shown ... gram- mar (SCFG) is defined as a CFG, with the dif- ference that it uses synchronous rules of the form [A 1 → α 1 , A 2 → α 2 ], with A 1 , A 2 nonterminals and α 1 , α 2 synchronous...

Ngày tải lên: 23/03/2014, 18:20

... 398 phrases (40%) with a verb-attached PP and 606 phrases (60%) with a noun-attached PP. Additionally, as held-out corpus for testing the efficacy of PP attachment information for prosodic phrasing, ... five intended phrase bound- 144 Table 4: Peiformance on phrasing complemented with PP attachment information from RIPPER and IB 1 with and without bigram features. phrasing a...

Ngày tải lên: 24/03/2014, 03:20

... to train with the same POS tagger available in test- ing, even if its performance is somewhat lower. Table 6 compares the performances of several of the state of the art dependency parsing systems with ... allows that use of future information in ad- dition to past information. The pipeline model nat- urally allows access to all the past information. Algorithm 3 Pseudo code for the lo...

Ngày tải lên: 20/02/2014, 12:20

... natural language use- ful for training a robust semantic role labeler. A possible relief for this problem is the utility of semi-supervised learning methods along with the existence of huge ... covering 38 semantic roles out of 52 roles present in the total training set. We use one unlabeled training set (U) for in- domain and another for out-of-domain experi- ments. The fo...

Ngày tải lên: 07/03/2014, 22:20

... and information-based methods, e.g. pointwise mutual information (Church & Hanks 1990). Several studies of the performance of lexical association metrics suggest significant room for improvement, ... frequency outperformed mutual information though not the t- test, while in Evert and Krenn 2001, log-likelihood and the t-test gave the best results, and mutual information again pe...

Ngày tải lên: 08/03/2014, 04:22