Báo cáo khoa học: "Combining Trigram and Winnow in Thai OCR Error Correction" potx

... Combining Trigram and Winnow in Thai OCR Error Correction Surapant Meknavin National Electronics and Computer Technology Center 73/1 Rama VI Road, Rajthevi, Bangkok, Thailand surapan@nectec.or.th ... are intro- duced in the text. The existing approach for correcting the spelling error in the languages that have no word boundary assumes that all substrings in input sentence are error strings, ... string representing a dubious area, and in- form at and on are words. In this 838 case, the unknown word and its sur- rounding known words are combined together, resulting in " ;in/ ormatjon"...

Báo cáo khoa học: "Combining Trigram-based and Feature-based Methods for Context-Sensitive Spelling Correction" docx

... in the back of Random House (Flexner, 1983). The confusion sets were selected on the basis of being frequently-occurring in Brown, and representing a variety of types of errors, including ... Spelling errors that result in valid, though unin- tended words, have been found to be very common in the production of text. Such errors were thought to be too difficult to handle and remain ... Performance of the baseline method for 18 confusion sets. "Train" and "Test" give the number of occurrences of any word in the confusion set in the training and test corpora. "Most...

Tài liệu Báo cáo khoa học: "Combining Functionality and Object Orientedness for Natural Language Processing" ppt

... for manipulating data. Objects can be classified into classes and instances. A class defines a procedure [called a method) for handling incoming messages of its instances. A class inherits methods ... flexible means of abstracting modules and sharing common knowledge. 1. Introduction The goal of this paper is to elaborate a domain-independent way of organizing linguistic knowledge, as a ... more principled and constrained way modules of the linguistic component interact, the less complicated will be the system and therefore the better perspective we can obtain for writing a large...

Báo cáo khoa học: "Combining Statistical and Knowledge-based Spoken Language Understanding in Conditional Models" pptx

... joint expertise in natural language processing and speech recognition, and best practices in language engineering for every new domain. On the other hand, a statistical learning approach needs ... in natural language processing and speech recognition, and best practices in language engineering for every new domain. In the past decade many statistical learning approaches have been proposed, ... directly porting the model to HCRFs, and finally introduce the CRFs and 882the features that obtain the best SLU result on ATIS test data. We compare the CRF and perceptron training performances...

Báo cáo khoa học: "Combining Acoustic and Pragmatic Features to Predict Recognition Performance in Spoken Dialogue Systems" pdf

... Randomly split the remaining data into an80% training and 20% test set.3. Run TiMBL with all possible parameter set-tings on the generated training and test sets and store the best performing ... – an interactive (i.e. mouse click-able) map – and speciﬁes mission goals using nat-ural language commands spoken into a headset, orby using combinations of GUI actions and spokencommands. ... (IS) and the update procedures for pro-cessing user input and generating system responses.Here, we brieﬂy introduce parts of the IS which areneeded to understand the system’s basic workings,and...

Báo cáo khoa học: "Combining Stochastic and Rule-Based Methods for Disambiguation in Agglutinative Languages" pptx

...

Báo cáo khoa học: "Combining data and mathematical models of language change" ppt

... languages: The effects of bilingualism and social structure. Lingua, 118(1):19–45.D. Minkova. 1997. Constraint ranking in Middle En-glish stress-shifting. English Language and Linguis-tics, 1(1):135–175.W.G. ... coupled. In Model 1 there is no coupling (ˆαt and ˆβtlearnedindependently), in Models 2–3 coupling takes theform of a hard constraint corresponding to Ross’generalization, and in Models ... Lan-guage and Behavior, pages 178–203. Chicago Lin-guistic Society, Chicago.L. Pearl and A. Weinberg. 2007. Input ﬁltering in syn-tactic acquisition: Answers from language changemodeling. Language...

Báo cáo khoa học: "Combining Deep and Shallow Approaches in Parsing German" pptx

... selected certainreadings become undistinguishable. Rather, in or-der to distinguish a maximum number of readings,pre-modifiers must attach to the last conjunct and post-modifiers and coordinating conjunctions ... verb,preposition and head noun into a triple, and thusonly count content words in the evaluation. Forlearning, the matter can be resolved empirically:2Even in this setting some readings cannot be distinguished(see ... produce and have in com-mon with the gold standard (measured in precision and recall). Another figure (crossing brackets) onlycounts those incorrect nodes that change the partialorder induced...

Báo cáo khoa học: "Combining Linguistic and Gaze Features to Resolve Second-Person References in Dialogue" docx

... referential and generic uses ofyou. Then, within the referential uses, we dis-tinguish between singular and plural, and ﬁnally,we resolve the singular referential instances byidentifying the intended ... when addressing the group,thus drawing the listeners’ gaze in this direction.Future work will involve expanding our data-set, and investigating new potentially predictivefeatures. In the slightly ... during the next ﬁve utterances) - fewerspeaker changes correlates with plural addressing.Information about dialogue acts also plays arole in distinguishing between singular and plu-ral interpretations....

Báo cáo khoa học: "Combining Source and Target Language Information for Name Tagging of Machine Translation Output" ppt

... performance on machine translated text, because of the poor syntax of the MT output and other errors in the translation. As some tagging distinctions are clearer in the source, and some in the target, ... ORGANIZATION, GPE and LOCATION, and so a total of 9 states. 4.4 Feature Sets for MEMM In our experiment, we are interested not only in training a module, but also in measuring the different ... training data, and so divided the NIST05 data into 5 subsets, each containing 20 documents. We ran a cross validation experiment on this small corpus, with 4 subsets as training data and...

Xem thêm