Along with the development of Chinese NLP,the universal Chinese question parser is in *** this paper we explored a new approach to design it,that is: the dependency relations in question were annotated based on Hownet...
详细信息
ISBN:
(纸本)0780379020
Along with the development of Chinese NLP,the universal Chinese question parser is in *** this paper we explored a new approach to design it,that is: the dependency relations in question were annotated based on Hownet in first,question semantic rules, semantic models and syntactic lists were extracted from question dependency tree,a vector was used to represent the output result after matching of question focus chunks,semantic chunks,question syntactic lists and question models based on these rules base.A universal Chinese question parser has been constructed by the guideline that the knowledge is apart from computational *** has achieved some success in finance question answering system that about 80%questions could be parsed correctly.
Grammar-based speech recognition systems exhibit performance degradation as their vocabulary sizes *** clustering is deemed to reduce the proportionality of this *** introduce an approach to data clustering for automa...
详细信息
Grammar-based speech recognition systems exhibit performance degradation as their vocabulary sizes *** clustering is deemed to reduce the proportionality of this *** introduce an approach to data clustering for automatic speech recognition systems using Kohonen Self-Organized *** results are used further to build a language model for each of the clusters using CMU-Cambridge *** approach was implemented as a prototype for a large vocabulary and continuous speech recognition system and about 8% performance improvement was achieved in comparison with the performance achieved using the language model and dictionary provided by Sphinx3. In this paper we present the experimental results along with discussions,analysis and potential future directions.
The availability of bandwidth resources fluctuates much more severely in mobile communication networks compared to wired networks. There is growing interest in developing adaptive multimedia services in mobile communi...
详细信息
This paper examines a method to automatically detect lexical stress of English,which is a part of a Computer-Assisted language Learning(CALL) system for Chinese to learn English pronunciation. Acoustic differences bet...
详细信息
ISBN:
(纸本)0780379020
This paper examines a method to automatically detect lexical stress of English,which is a part of a Computer-Assisted language Learning(CALL) system for Chinese to learn English pronunciation. Acoustic differences between vowels in syllables that do or do not carry lexical stress are *** use coefficients based on fundamental frequency, energy and duration to differentiate stressed vowels from unstressed ones by linear classification. Although the distributions of the coefficients overlap to some extent,the new normalized pitch-based coefficient obtains correct classification results of about 84%.Compared with previous detection methods,the results suggest that proper normalization of acoustic feature is much more valuable than the combination of different features.
This paper introduces a new text categorization metliod utilizing Machine Learning based on Extension *** dependent degree based on the Extension Theory represents the extent to which the element belongs to the predef...
详细信息
ISBN:
(纸本)0780379020
This paper introduces a new text categorization metliod utilizing Machine Learning based on Extension *** dependent degree based on the Extension Theory represents the extent to which the element belongs to the predefined *** "closeness degree" between the input document vector and standard range of each predefined category can be *** new method is conceptually simple;it can be used with relatively low complexity and high flexibility:The algorithm is highly *** can be effectively applied to text categorization,of which various features are consecutive ***,this algorithm can be widely applied to computational linguistics.
This paper focuses on integrated prosodic word prediction for Chinese *** avoid the problem of inconsistency between lexical words and prosodic words in Chinese,lexical word segmentation and prosodic word prediction a...
详细信息
This paper focuses on integrated prosodic word prediction for Chinese *** avoid the problem of inconsistency between lexical words and prosodic words in Chinese,lexical word segmentation and prosodic word prediction are taken as one process instead of two independent ***,two word-based approaches are proposed to drive this integrated prosodic word prediction:The first one follows the notion of lexicalized hidden Markov models,and the second one is borrowed from unknown word identification for *** results of our primary experiment show these integrated approaches are effective.
As a study of computational semantics and iiifomiation science we describe a method of logical translation that seems to constitute a kind of semantic analysis,but it may also be considered a kind of information extra...
详细信息
As a study of computational semantics and iiifomiation science we describe a method of logical translation that seems to constitute a kind of semantic analysis,but it may also be considered a kind of information extraction. We discuss the translation from Dataflow Structures partly to parser programs and partly to informational *** methods are independent of any particular semantic theory and seem to fit nicely with a variety of available semantic *** illustrate the principles we discuss the detailed analysis of some carefully selected prototypical database queries fomiulated in simple English.
We have presented a new method called Super-Function Based Machine Translation(SFBMT).SFBMT uses Super-Function(SF) to translate without syntactic and semantic analysis as most conventional MT systems *** represents c...
详细信息
We have presented a new method called Super-Function Based Machine Translation(SFBMT).SFBMT uses Super-Function(SF) to translate without syntactic and semantic analysis as most conventional MT systems *** represents correspondence with a source language sentence structure and a target language sentence structure *** to this feature,the system realizes very fast *** we do not feel the sense of incompatibility in the translation result because the SFs are automatically generate from a large *** this paper,we dscribe(1) SF based machine translation,(2) extracting SF,(3) the evaluation *** the experimental results,we obtain 84%of effective rates of translation.
An algorithm of additive audio watermarking based on SNR to determine a scaling parameterαis proposed in this *** visually recognizable binary watermark image could be embedded in the original signal wavelet domain i...
详细信息
ISBN:
(纸本)0780379020
An algorithm of additive audio watermarking based on SNR to determine a scaling parameterαis proposed in this *** visually recognizable binary watermark image could be embedded in the original signal wavelet domain in the *** intensity of embedded watermarks on the different audio segments can be modified by adaptively adjusting the scaling parameterα.Comparing to the algorithm without using SNR,the experimental results demonstrate that the watermark is inaudible and the algorithm is robust to many operations of digital audio signal,such as resampling,cropping, low pass filtering,additive noise and so on.
In Text Classification systems,the quality and quantity of the training document is one of the most import factors which affect *** gathering,filtering and classifying the training documents is very *** paper employs ...
详细信息
ISBN:
(纸本)0780379020
In Text Classification systems,the quality and quantity of the training document is one of the most import factors which affect *** gathering,filtering and classifying the training documents is very *** paper employs feedback mechanism for text classification systems and reduces the need for labeled training documents by unifying the strengths of k-NN and linear *** collects various types of example documents provided by the users and set up the user *** the documents from stream are matched with the profiles,the relevant documents are used to improve the user *** online approach offers the advantage of continuous learning in the batch-adaptive text considerations,on the batch-adaptive filtering task.
暂无评论