Báo cáo khoa học: "Combining POMDPs trained with User Simulations and Rule-based Dialogue Management in a Spoken Dialogue System" docx

... been trained yet, and several systems that use various pre -trained policies. These policies are obtained by dialogue simulations with user models based on data obtained from human- machine dialogues ... uniform way. This facilitates dialog eval- uation, data mining and online learning because data is available for querying as soon as it has been stored. There is no need for se...

Ngày tải lên: 23/03/2014, 17:20

... used average accuracy to evaluate their per- formance. Specifically, with χ 2 based feature ranking, F SGMM attained 55.4% average accu- racy, while the best average accuracy of CGD term and CGD SV ... feature space should be stable and robust against random sampling. After deter- mination of important contextual words, we use a Gaussian mixture model (GMM) based clustering algorithm (...

Ngày tải lên: 20/02/2014, 16:20

... between associations and counts. The phrase trading and what has a higher count than the phrase trading and transacting, but trading and transacting are more highly associated. In this paper, we use point-wise mutual ... counts involving both ate and fork. We use associations rather than raw counts. The phrases trading and transacting versus trading and what provide an examp...

Ngày tải lên: 23/03/2014, 14:20

... it: a similarity degree measure, a string editing distance (Wagner and Fischer, 1974) and a tree editing distance 1 (Zhang and Shasha, 1989). Section 4 discusses and evaluates these methods and ... the following example (Kan)): Rear: a paragraph in which a link just stopped occurring the paragraph before. No link: any remaining paragraphs. Different methods, different re...

Ngày tải lên: 24/03/2014, 03:20

... Proceedings of EACL '99 Complementing WordNet with Roget's and Corpus-based Thesauri for Information Retrieval Rila Mandala, Takenobu Tokunaga and Hozumi Tanaka Abstract This paper ... development and application of spaceborne ocean re- mote sensing in oceanography, seabed prospecting and mining, or any marine- science activity are relevant. Documents that discuss...

Ngày tải lên: 31/03/2014, 21:20

... Winograd's terminology for functional gran~nar (Winograd, 1983). In general, case schemata will be mapped into CLAUSE-RS and concept schemata are mapped into NP-R~. A CLAUSE-RS has a ... to ATLAS/II's analysis. In stage i, we first examine those semantic symbols which have an attached case schema and instantiate them according to their trans- formation rules. I...

Ngày tải lên: 08/03/2014, 18:20

... Pro- gram, Grant -in- Aid for Scientific Research (Nos. 16510164, 17310130, and 17035069), and Cooperation of Innovative Technology and Advanced Research in Evolutional Area (CITY AREA) -in the Central ... Shi- zuoka Area-, from the Ministry of Education, Culture, Sports, Science and Technology, Japan, and by Health and Labour Sciences Research Grant from the Minis- try of Healt...

Ngày tải lên: 16/03/2014, 14:20

... New England Biolabs (Pickering, Ontario, Canada). All other chemicals were of analytical grade and were obtained from Sigma-Aldrich and Fisher Scientific (Nepean, Ontario, Canada). Bacterial strains and ... research was supported by a grant from the National Science and Engineering Research Council of Canada. S. Seah thanks the Canadian Foundation for Innovation and Ontario Innovat...

Ngày tải lên: 30/03/2014, 15:20

... straints alone and again using both syntactic and semantic constraints. As can be seen, Gemini is quite accurate at detecting a repair, although somewhat less accurate at detecting a false ... which pattern matching, syntactic and semantic analy- sis, and acoustic analysis can be helpful in detect- ing and correcting these repairs. We use pattern matching to determine...

Ngày tải lên: 31/03/2014, 06:20

... training data are comparatively smaller than WAQ and WAQA, they however yield comparable results. The linear combination of datasets (WAQ+WAQA+LSR Lin ) yields statistically significant performance ... the availability of suitable training data for the translation probabilities. Berger and Laf- ferty (1999) initially built synthetic training data consisting of queries automatically generat...

Ngày tải lên: 20/02/2014, 07:20