Báo cáo khoa học: "A Machine Learning Approach to Extract Temporal Information from Texts in Swedish and Generate Animated 3D Scenes" docx

... A Machine Learning Approach to Extract Temporal Information from Texts in Swedish and Generate Animated 3D Scenes Anders Berglund Richard Johansson Pierre ... Annotating Texts with Temporal Information Several schemes have been proposed to anno- tate temporal information in texts, see Setzer and Gaizauskas (2002), inter alia. Many of them were in...

Ngày tải lên: 31/03/2014, 20:20

... between ignoring pronominal antecedents (and risking to discard all the correct antecedents within the window) and re- solving them (and risking multiplication of errors). In case all of the instances ... rule-based and the machine learning approach. For the latter, good results have been achieved with large feature sets (including syntactic, se- mantic, grammatical and morp...

Ngày tải lên: 17/03/2014, 06:20

... A Machine Learning Approach to Pronoun Resolution in Spoken Dialogue Michael Strube and Christoph M ¨ uller European Media Laboratory GmbH Villa Bosch Schloß-Wolfsbrunnenweg ... resolution in written text. 3.2 Data Generation Training and test data instances were generated from our corpus as follows. All markables were sorted in document order, and markables for first an...

Ngày tải lên: 23/03/2014, 19:20

... paper consists in the de- velopment of a machine- learning approach to the identification of WH gaps. This approach reduces the lexical prerequisites for this task, while main- taining a high degree ... necessary for in- terpretation of the sentence, and could contribute to a natural language understanding or machine translation application. Since WH dependencies also tend...

Ngày tải lên: 24/03/2014, 03:20

... a machine learning approach to evaluating the well- formedness of output of a machine translation system, using classifiers that learn to distinguish human reference translations from machine ... transferred to the target language. Acknowledgements Our thanks go to Eric Ringger and Max Chickering for programming assistance with the tools used in building and evaluati...

Ngày tải lên: 31/03/2014, 04:20

... similarity set- ting, in contrast, the contributions of context types are uniformly fixed. In order to elucidate what is happening in this situation, the investigations on ma- chine learning settings, as ... well- performing contextual information in the past stud- ies (Ruge, 1997; Lin, 1998). In this paper the sophis- ticated parser RASP Toolkit 2 (Briscoe et al., 2006) was ut...

Ngày tải lên: 23/03/2014, 17:20

... training and test data. We hold out 10% of the data for the test. We applied 20% under-sampling to the training instances that do not have any errors to alleviate data imbalance in the training ... containing ~3,300 mistakes in article usage were extracted from the entire CLC-FCE corpus, and (3) about 30k noun phrases containing ~2,500 mistakes were extracted from the J...

Ngày tải lên: 30/03/2014, 17:20

... c i , and w j is the corresponding feature weight. 3 Ranking Model Training To learn ranking function in Equation (1), we need to determine the feature set θ and learn weight vector w from reorder ... and integrated decoding settings. In the rest of the paper, we will first formally present our ranking-based word reordering model, then followed by detailed steps of modeling tra...

Ngày tải lên: 19/02/2014, 19:20

... clustering for discriminative learning. In Proceedings of the 47th Annual Meeting of the Association for Computational Linguistics (ACL). Yang Liu, Qun Liu, and Shouxun Lin. 2006. Tree -to- string alignment ... feature informing the labeling pro- cess seems more desirable. We thus propose to represent each phrase pair in- stance (including its bilingual one-word contexts) as feature...

Ngày tải lên: 23/03/2014, 16:20

... nominal chunks to the begin- ning and to the end of the sentence  number of subordinate-clause marks to the beginning and to the end of the sen- tence  distance (in tokens) to the beginning and ... means we took into account the pre- vious 5 tokens and the following 5. We also used the C4.5 algorithm initially, since this algorithm gets very good results in other sim...

Ngày tải lên: 31/03/2014, 01:20