Studies on acoustic space have strengthened the view that vowels are acoustically and perceptually defined in terms of their relative positioning in vowel space. Every speaker identifies an optimal vowel space within which perceptual, phonological contrast is maintained. This is an interdisciplinary study involving speech pathology, physics of speech and neurology of speech. Two case studies of dysarthria presented in this paper are -- one Parkinson's disease and one case of acute ischemic stroke with age-gender-language matched controls. A detailed acoustic analysis shows how acoustic space gets considerably reduced, in both PD and stroke, and in these two very different kinds of dysarthrias the acoustic space is also modified very differently. The study also examines the third formant to show that the higher formants are consistently lowered in both PD and stroke. Hypokinetic speech production in these cases is reflected in lower intensity. The results have significant applications in clinical acoustics and in the theoretical fields of neurology of speech, linguistics and phonology.
{"title":"Acoustic Space in Motor Disorders of Speech: Two Case Studies","authors":"Vaishna Narang, Deepshikha Misra, Garima Dalal","doi":"10.1109/IALP.2011.25","DOIUrl":"https://doi.org/10.1109/IALP.2011.25","url":null,"abstract":"Studies on acoustic space have strengthened the view that vowels are acoustically and perceptually defined in terms of their relative positioning in vowel space. Every speaker identifies an optimal vowel space within which perceptual, phonological contrast is maintained. This is an interdisciplinary study involving speech pathology, physics of speech and neurology of speech. Two case studies of dysarthria presented in this paper are -- one Parkinson's disease and one case of acute ischemic stroke with age-gender-language matched controls. A detailed acoustic analysis shows how acoustic space gets considerably reduced, in both PD and stroke, and in these two very different kinds of dysarthrias the acoustic space is also modified very differently. The study also examines the third formant to show that the higher formants are consistently lowered in both PD and stroke. Hypokinetic speech production in these cases is reflected in lower intensity. The results have significant applications in clinical acoustics and in the theoretical fields of neurology of speech, linguistics and phonology.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"150 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116357609","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The paper abandons a strict two-way sub-classification of intransitive verbs into unaccuasative and unergative for Hindi and proposes a distribution plotting of the same in a diffusion chart. The diagnostics tests that Bhatt (2003) applied on Hindi data are ranked for their efficiency of attributing correct sub-class to verbs. The diffusion chart shows that a tripartite classification handles the issue of classification of intransitive verbs in a better manner than the classical binary approach. The tripartite classification is as follows: (1) Verbs that take animate subject and are compatible with adverb of volitionality; (2) Verbs that take animate subject but are not compatible with adverb of volitionality; and (3) Verbs that take inanimate subject. The classification is of immense advantage for various NLP tasks such as machine translation, natural language generation.
{"title":"Issues with the Unergative/Unaccusative Classification of the Intransitive Verbs","authors":"Nitesh Surtani, Khushboo Jha, Soma Paul","doi":"10.1109/IALP.2011.54","DOIUrl":"https://doi.org/10.1109/IALP.2011.54","url":null,"abstract":"The paper abandons a strict two-way sub-classification of intransitive verbs into unaccuasative and unergative for Hindi and proposes a distribution plotting of the same in a diffusion chart. The diagnostics tests that Bhatt (2003) applied on Hindi data are ranked for their efficiency of attributing correct sub-class to verbs. The diffusion chart shows that a tripartite classification handles the issue of classification of intransitive verbs in a better manner than the classical binary approach. The tripartite classification is as follows: (1) Verbs that take animate subject and are compatible with adverb of volitionality; (2) Verbs that take animate subject but are not compatible with adverb of volitionality; and (3) Verbs that take inanimate subject. The classification is of immense advantage for various NLP tasks such as machine translation, natural language generation.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115897276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
For Chinese word segmentation and POS tagging problem, both character-based and word-based discriminative approaches can be used. Experiments show that these two approaches bring different errors and can complement each other. In this paper, we propose a joint decoding model based on both character-based and word-based models using multi-beam search algorithm. Experimental results show that the joint decoding model outperforms character-based and word-based baseline models.
{"title":"Joint Decoding for Chinese Word Segmentation and POS Tagging Using Character-Based and Word-Based Discriminative Models","authors":"Xinxin Li, Xuan Wang, Lin Yao","doi":"10.1109/IALP.2011.24","DOIUrl":"https://doi.org/10.1109/IALP.2011.24","url":null,"abstract":"For Chinese word segmentation and POS tagging problem, both character-based and word-based discriminative approaches can be used. Experiments show that these two approaches bring different errors and can complement each other. In this paper, we propose a joint decoding model based on both character-based and word-based models using multi-beam search algorithm. Experimental results show that the joint decoding model outperforms character-based and word-based baseline models.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127020677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Speech visualization can be extended to a task of pronunciation animation for language learners. In this paper, a three dimensional English articulation database is recorded using Carstens Electro-Magnetic Articulograph (EMA AG500). An HMM-based visual synthesis method for continuous speech is implemented to recover 3D articulatory information. The synthesized articulations are then compared to the EMA recordings for objective evaluation. Using a data-driven 3D talking head, the distinctions between the confusable phonemes can be depicted through both external and internal articulatory movements. The experiments have demonstrated that the HMM-based synthesis with limited training data can achieve the minimum RMS error of less than 2mm. The synthesized articulatory movements can be used for computer assisted pronunciation training.
{"title":"The Phoneme-Level Articulator Dynamics for Pronunciation Animation","authors":"Sheng Li, Lan Wang, En Qi","doi":"10.1109/IALP.2011.13","DOIUrl":"https://doi.org/10.1109/IALP.2011.13","url":null,"abstract":"Speech visualization can be extended to a task of pronunciation animation for language learners. In this paper, a three dimensional English articulation database is recorded using Carstens Electro-Magnetic Articulograph (EMA AG500). An HMM-based visual synthesis method for continuous speech is implemented to recover 3D articulatory information. The synthesized articulations are then compared to the EMA recordings for objective evaluation. Using a data-driven 3D talking head, the distinctions between the confusable phonemes can be depicted through both external and internal articulatory movements. The experiments have demonstrated that the HMM-based synthesis with limited training data can achieve the minimum RMS error of less than 2mm. The synthesized articulatory movements can be used for computer assisted pronunciation training.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129690948","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we proposed an approach to model the pronunciation of non-native accented speech for automatic speech recognition system. The proposed method consists of two phases: phones adaptation and pronunciation generalization. In phones adaptation, we identify the phones used by non-native speakers compared to the standard phones, and then remove the mismatch, as a result of the influence from mother tongue. In pronunciation adaptation, we predict the pronunciations of words by non-native speakers. The results shown the proposed approach reduce the WER from 44.8% to 41.9%.
{"title":"Non-native Accent Pronunciation Modeling in Automatic Speech Recognition","authors":"Basem H. A. Ahmed, T. Tan","doi":"10.1109/IALP.2011.65","DOIUrl":"https://doi.org/10.1109/IALP.2011.65","url":null,"abstract":"In this paper, we proposed an approach to model the pronunciation of non-native accented speech for automatic speech recognition system. The proposed method consists of two phases: phones adaptation and pronunciation generalization. In phones adaptation, we identify the phones used by non-native speakers compared to the standard phones, and then remove the mismatch, as a result of the influence from mother tongue. In pronunciation adaptation, we predict the pronunciations of words by non-native speakers. The results shown the proposed approach reduce the WER from 44.8% to 41.9%.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"345 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132024778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The performance of Chinese Pinyin-to-Character conversion is severely affected when the characteristics of the training and conversion data differ. As natural language is highly variable and uncertain, it is impossible to build a complete and general language model to suit all the tasks. The traditional adaptive MAP models mix the task independent data with task dependent data using a mixture coefficient but we never can predict what style of language users have and what new domain will appear. This paper presents a statistical error-driven adaptive language modeling approach to Chinese Pinyin input system. This model can be incrementally adapted when an error occurs during Pinyin-to-Character converting time. It significantly improves Pinyin-to-Character conversion rate.
{"title":"Error-Driven Adaptive Language Modeling for Chinese Pinyin-to-Character Conversion","authors":"J. Huang, D. Powers","doi":"10.1109/IALP.2011.46","DOIUrl":"https://doi.org/10.1109/IALP.2011.46","url":null,"abstract":"The performance of Chinese Pinyin-to-Character conversion is severely affected when the characteristics of the training and conversion data differ. As natural language is highly variable and uncertain, it is impossible to build a complete and general language model to suit all the tasks. The traditional adaptive MAP models mix the task independent data with task dependent data using a mixture coefficient but we never can predict what style of language users have and what new domain will appear. This paper presents a statistical error-driven adaptive language modeling approach to Chinese Pinyin input system. This model can be incrementally adapted when an error occurs during Pinyin-to-Character converting time. It significantly improves Pinyin-to-Character conversion rate.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130184442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sophia Yat-Mei Lee, Daming Dai, Shoushan Li, K. Ahrens
Sentiment and emotion analysis have been traditionally established as independent research topics in NLP. Although they are two important aspects of subjective information and are closely related, there have been few attempts to combine the two analyses. As a preliminary attempt, we integrate emotion information into sentiment analysis by employing emotion keywords to help automatically extract pseudo-labeled samples. The extracted pseudo-labeled samples are then used as the initial training data to perform semi-supervised learning for sentiment classification. Experimental results across four domains show that our approach using emotion keywords is capable of extracting pseudo-labeled samples with high precision (about 90%). Moreover, the pseudo-labeled samples along with the semi-supervised learning approach further improve the classification performance.
{"title":"Extracting Pseudo-Labeled Samples for Sentiment Classification Using Emotion Keywords","authors":"Sophia Yat-Mei Lee, Daming Dai, Shoushan Li, K. Ahrens","doi":"10.1109/IALP.2011.61","DOIUrl":"https://doi.org/10.1109/IALP.2011.61","url":null,"abstract":"Sentiment and emotion analysis have been traditionally established as independent research topics in NLP. Although they are two important aspects of subjective information and are closely related, there have been few attempts to combine the two analyses. As a preliminary attempt, we integrate emotion information into sentiment analysis by employing emotion keywords to help automatically extract pseudo-labeled samples. The extracted pseudo-labeled samples are then used as the initial training data to perform semi-supervised learning for sentiment classification. Experimental results across four domains show that our approach using emotion keywords is capable of extracting pseudo-labeled samples with high precision (about 90%). Moreover, the pseudo-labeled samples along with the semi-supervised learning approach further improve the classification performance.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"54 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126641534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Information retrieval model is still can not achieve satisfactory performance after decades of development. One of the reasons is the queries can not express information need precisely. Researches have shown that query reformulation can improve the performance of retrieval model. In this paper, we propose a query reformulation model, which use Markov network to represent term relationship to obtain useful information from corpus to reformulate query. Experimental results show that our model can avoid topic drift and then improve the retrieval performance.
{"title":"A Query Reformulation Model Using Markov Graphic Method","authors":"Jiali Zuo, Mingwen Wang","doi":"10.1109/IALP.2011.62","DOIUrl":"https://doi.org/10.1109/IALP.2011.62","url":null,"abstract":"Information retrieval model is still can not achieve satisfactory performance after decades of development. One of the reasons is the queries can not express information need precisely. Researches have shown that query reformulation can improve the performance of retrieval model. In this paper, we propose a query reformulation model, which use Markov network to represent term relationship to obtain useful information from corpus to reformulate query. Experimental results show that our model can avoid topic drift and then improve the retrieval performance.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115014669","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper introduces a model to describe the dynamic evolution of network information, identifying and analyzing the document collection on the same topic in different stages. In order to characterize the dynamic relationship of evolutionary content differences, this paper presents a dynamic multi-document summarization model, which is called the Dynamic Manifold-Ranking Model (DMRM). Some experiments were conducted on the Update Task test data from TAC2008, and results of new model were compared with results from the TAC2008 evaluation. This comparison demonstrated the effectiveness of the model.
{"title":"Research on Multi-document Summarization Model Based on Dynamic Manifold-Ranking","authors":"Meiling Liu, Honge Ren, Dequan Zheng, T. Zhao","doi":"10.1109/IALP.2011.55","DOIUrl":"https://doi.org/10.1109/IALP.2011.55","url":null,"abstract":"This paper introduces a model to describe the dynamic evolution of network information, identifying and analyzing the document collection on the same topic in different stages. In order to characterize the dynamic relationship of evolutionary content differences, this paper presents a dynamic multi-document summarization model, which is called the Dynamic Manifold-Ranking Model (DMRM). Some experiments were conducted on the Update Task test data from TAC2008, and results of new model were compared with results from the TAC2008 evaluation. This comparison demonstrated the effectiveness of the model.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114745902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hoang-Quynh Le, Mai-Vu Tran, Nhat-Nam Bui, N. Phan, Quang-Thuy Ha
Personal names are among one of the most frequently searched items in web search engines and a person entity is always associated with numerous properties. In this paper, we propose an integrated model to recognize person entity and extract relevant values of a pre-defined set of properties related to this person simultaneously for Vietnamese. We also design a rich feature set by using various kind of knowledge resources and a apply famous machine learning method CRFs to improve the results. The obtained results show that our method is suitable for Vietnamese with the average result is 84 % of precision, 82.56% of recall and 83.39 % of F-measure. Moreover, performance time is pretty good, and the results also show the effectiveness of our feature set.
{"title":"An Integrated Approach Using Conditional Random Fields for Named Entity Recognition and Person Property Extraction in Vietnamese Text","authors":"Hoang-Quynh Le, Mai-Vu Tran, Nhat-Nam Bui, N. Phan, Quang-Thuy Ha","doi":"10.1109/IALP.2011.37","DOIUrl":"https://doi.org/10.1109/IALP.2011.37","url":null,"abstract":"Personal names are among one of the most frequently searched items in web search engines and a person entity is always associated with numerous properties. In this paper, we propose an integrated model to recognize person entity and extract relevant values of a pre-defined set of properties related to this person simultaneously for Vietnamese. We also design a rich feature set by using various kind of knowledge resources and a apply famous machine learning method CRFs to improve the results. The obtained results show that our method is suitable for Vietnamese with the average result is 84 % of precision, 82.56% of recall and 83.39 % of F-measure. Moreover, performance time is pretty good, and the results also show the effectiveness of our feature set.","PeriodicalId":297167,"journal":{"name":"2011 International Conference on Asian Language Processing","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117083362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：