ソンラートラムワニッチ ウィラット

In the present borderless information society, we need a lot of fundamental linguistic tools as well as the standard reference resources to facilitate our daily communications across the languages and cultures for better understanding or smoothing the communications. Online collaborative works are efficiently conducted among expert groups via many existing services such as Sourceforge, Wiki or Weblog. However, in the process of multilingual resource development and intercultural communication we still need to fulfill the requirements in well-structured design of the database, and communication tools that provide necessary linkages between records of intention to particular assertions, and functions to realize selectional preference in case that there are more than one assertion. In this paper, we propose a new platform, called Knowledge Unifying Initiator (KUI). We conducted a study on multilingual medical text collaborative translation and the initiative in Asian WordNet development to evaluate our proposed platform.

This paper discusses the problem of language, script and ecoding(LSE) identification.

A lexicon is an important linguistic resource needed for both shallow and deep language processing. Currently, there are few machine-readable Thai dictionaries available, and most of them do not satisfy the computational requirements. This paper presents the design of a Thai lexicon named the TCL's Computational Lexicon (TCLLEX) and proposes a method to construct a large-scale Thai lexicon by re-using two existing dictionaries and a large number of texts on the Internet. In addition to morphological, syntactic, semantic case role and logical information in the existing dictionaries, a sort of semantic constraint called selectional preference is automatically acquired by analyzing Thai texts on the web and then added into the lexicon. In the acquisition process of the selectional preferences, the so-called Bayesian Information Criterion (BIC) is applied as the measure in a tree cut model. The experiments are done to verify the feasibility and effectiveness of obtained selection preferences. Copyright © 2006 The Institute of Electronics, Information and Communication Engineers.

The growing of multilingual information processing technology has created the need of linguistic resources, especially lexical database. Many attempts were put to alter the traditional dictionary to computational dictionary or widely named as computational lexicon. TCL's Computational Lexicon (TCLLEX) is a recent development of a large-scale Thai Lexicon which aims to serve as a fundamental linguistic resource for natural language processing research. We design either terminology or ontology for structuring the lexicon based on the idea of computability and reusability.

This paper compares the effectiveness of two different Thai search engines by using a blind evaluation. The probabilistic-based dictionary-less search engine is evaluated against the traditional word-based indexing method. The web documents from 12 Thai newspaper web sites consisting of 83,453 documents are used as the test collection. The relevance judgment is conducted on the first five returned results from each system. The evaluation process is completely blind. That is, the retrieved documents from both systems are shown to the judges without any information about the search techniques. Statistical testing shows that the dictionary-less approach is better than the word-based indexing approach in terms of the number of found documents and the number of relevance documents.

This paper presents a framework for Thai morphological analysis based on the theoretical background of conditional random fields. We formulate morphological analysis of an unsegmented language as the sequential supervised learning problem. Given a sequence of characters, all possibilities of word/tag segmentation are generated, and then the optimal path is selected with some criterion. We examine two different techniques, including the Viterbi score and the confidence estimation. Preliminary results are given to show the feasibility of our proposed framework.

This paper compares the retrieval performance of seven public search engines on Thai queries by using a blind evaluation. Two of the engines used in this study are Thai-focused search services, while the remaining engines are large, commercial search engines that have wider collections of web data and support several languages. The results are compared by Mean average precision (MAP) and Mean reciprocal rank of the first correct answer (MRR). These results are calculated from binary relevance judgments of the first 20 returned results, using 56 topics. Statistical testing shows that there are significant differences among engines.

The motivation of this work comes from the need of a Thai web corpus for testing our information retrieval algorithm. Two collections of news web documents are gathered from two different Thai newspaper web sites. Our goal is to find a simple yet effective method to extract news articles from these web collections. We explore the use of machine learning methods to distinguish article pages from non-article pages, e.g. table of contents, advertisements. Then, the selected web articles are compared in a fine-grained manner in order to find informative structures. Both steps of information extraction utilize the structural features of web documents rather than the extracted keywords or terms. Thus, the inherent errors of word segmentation, one of the major problems in Thai text processing, do not affect to this method. © Springer-Verlag Berlin Heidelberg 2006.

オントロジー構築の効率化手法を検討した。

This paper presents a novel algorithm for document clustering based on a combinatorial framework of the Principal Direction Divisive Partitioning (PDDP) algorithm and a simplified version of the EM algorithm called the spherical Gaussian EM (sGEM) algorithm. The idea of the PDDP algorithm is to recursively split data samples into two sub-clusters using the hyperplane normal to the principal direction derived from the covariance matrix. However, the PDDP algorithm can yield poor results, especially when clusters are not well-separated from one another. To improve the quality of the clustering results, we deal with this problem by re-allocating new cluster membership using the sGEM algorithm with different settings. Furthermore, based on the theoretical background of the sGEM algorithm, we can naturally extend the framework to cover the problem of estimating the number of clusters using the Bayesian Information Criterion. Experimental results on two different corpora are given to show the effectiveness of our algorithm. © Springer-Verlag Berlin Heidelberg 2005.

This paper analyzes the results of automatic concept alignment between two ontologies. We use an iterative algorithm to perform concept alignment. The algorithm uses the similarity of shared terms in order to find the most appropriate target concept for a particular source concept. The results show that the proposed algorithm not only finds the relation between the target concepts and the source concepts, but the algorithm also shows some flaws in the ontologies. These results can be used to improve the correctness of the ontologies. © Springer-Verlag Berlin Heidelberg 2005.

In this paper, we propose a novel approach for automatically identifying the language of a given text based on the concept of string kernels. Our approach can identify the language from the text directly, regardless of its coding system. In particular, we view the text in a more fine-grained encoding as the string of bytes. The similarity between two strings can be implicitly computed through an efficient dynamic alignment using suffix trees. We provide empirical evidence that applying the string kernels to the language identification problem yields an impressive performance using two different kernel classifiers: the kernelized version of the centroid-based method and the support vector machines. Our experiments are based on a reasonable scale of the data set in terms of the number of languages to be identified, including 17 different languages. © 2005 IEEE.

辞書の開発のため、テキストから単語の持つ意味情報を自動的に獲得する手法を開発した。

大量の電子化テキストに基づく辞書開発のための枠組みを提案する。辞書開発用のツールを開発し､実際にタイ語に対し適用した結果を報告する。

NICTが作成した英語学習者コーパスを再利用するために、Semantic Webの技術を利用し、有機的に共用利用する枠組みを提案した。

大量の電子化テキストに基づく辞書開発のための枠組みを提案する。

新しい教育の実現のためのE-learning について述べた。

大量の電子化テキストに基づく辞書開発のための枠組みを提案する。大規模データからの語義の自動獲得を効率的に行う手法を開発した。

大量の電子化テキストに基づく辞書開発のための枠組みを提案する。

Language Resources are recognized as an essential component in linguistic infrastructure and a starting point of Natural Language Processing systems and applications. In this paper, we describe the achievement of the development and the use of Thai Language Resources germinated with an open collaboration platform, under the collaboration between research institutes. The resources include either text or speech. Text resources are divided into lexicon database and annotated corpus. We started developing a corpus-based Thai-English lexicon database (LEXiTRON) since 1994. It was originated from a dictionary designed for using in developing a machine translation system. Since then the Thai POS was designed and evaluated in several applications (word segmentation, machine translation, grapheme-to-phoneme, etc.) Extending the lexicon database, POS tagged corpus (ORCHID), and speech corpora for both synthesis and recognition are developed and functioned as an important part of research and development on NLP or HLT. These language resources are available for academic experiment.

A statistical corpus-based approach for acquiring selectional preferences of verbs is proposed. By parsing through text corpora, we obtain examples of context nouns that are considered to be the selectional preferences of a given verb. The approach is to generalize initial noun classes to the most appropriate levels on a semantic hierarchy. We present an iterative algorithm for generalization by combining an agglomerative merging and a model selection technique called the Bayesian Information Criterion (BIC). In our experiments, we consider the Web as the large corpora. We also propose approaches for extracting examples from the Web. Preliminarily experimental results are given to show the feasibility and effectiveness of our approach.

デジタルライブラリ等を例に、ネットワークを介した支援環境の動向について述べる｡

タイ語言語資源の強調的開発の現状について報告する｡

The Principal Direction Divisive Partitioning (PDDP) algorithm is a fast and scalable clustering algorithm [3]. The basic idea is to recursively split the data set into sub-clusters based on principal direction vectors. However, the PDDP algorithm can yield poor results, especially when cluster structures are not well-separated from one another. Its stopping criterion is based on a heuristic that often tends to over-estimate the number of clusters. In this paper, we propose simple and efficient solutions to the problems by refining results from the splitting process, and applying the Bayesian Information Criterion (BIC) to estimate the true number of clusters. This motivates a novel algorithm for unsupervised clustering, which its experimental results on different data sets are very encouraging.

One of the crucial problems in developing high quality Thai text-to-speech synthesis is to detect phrase break from Thai texts. Unlike English, Thai has no word boundary delimiter and no punctuation mark at the end of a sentence. It makes the problem more serious. Because when we detect phrase break incorrectly, it is not only producing unnatural speech but also creating the wrong meaning. In this paper, we apply machine learning algorithms namely C4.5 and RIPPER in detecting phrase break. These algorithms can learn useful features for locating a phrase break position. The features which are investigated in our experiments are collocations in different window sizes and the number of syllables before and after a word in question to a phrase break position. We compare the results from C4.5 and RIPPER with a based-line method (Part-of-Speech sequence model). The experiment shows that C4.5 and RIPPER appear to outperform the basedline method and RIPPER performs better accuracy results than C4.5.

The Greenstone Digital Library (GSDL) system, developed by the New Zealand Digital Library (NZDL) Consortium at the University of Waikato is a suite of open-source software for building and distributing digital library collections. At the Thai Computational Linguistic (TCL) Laboratory of CRL Asia Research Center, we plan to implement and host digital libraries in several major Asian languages. In this paper, we describe our experiences in implementing Thai and Japanese digital libraries using Greenstone. © Springer-Verlag Berlin Heidelberg 2003.

ORCHID (Open Linguistic Resources Chanelled toward InterDisciplinary research) is a project aimed at building linguistic resources to support research in, but not limited to, natural language processing. Based on the concept of an open architecture design, the resources must be fully compatible with those which already exist, and software tools must also be made available. This paper describes the construction of a Thai part-of-speech (POS) tagged corpus, a preliminary stage in the construction of a Thai speech corpus. The paper also details the development of a POS tagger to be used in the construction of the POS-tagged corpus. Additionally, we describe a proposal for a new tagset, based on the results of a prior multilingual machine translation project. The corpus is annotated on three levels: paragraph, sentence, and word. Text information is maintained in the form of the text information lines and the number lines, which are both utilized in data retrieval. Finally, we describe a POS neuro tagger, which consists of a three-layer perceptron with elastic input. Computer experiments show that the neuro tagger has an accuracy of 94.4 per cent for tagging ambiguous words when tested on a small Thai training corpus containing 22,311 ambiguous words. A series of comparative experiments further show that the neuro tagger is superior to the statistical models including the frequency model (a baseline model), a local n-gram model, and HMM (Hidden Markov Model). © 2000 Oxford University Press.

Copyright 2000 ACM. Some languages including Thai, Japanese and Chinese do not have explicit word boundary. This causes the problem of word boundary ambiguity that results in decreasing the accuracy of information retrieval. This paper proposes a new technique so-called character clustering to reduce the ambiguity of word boundary in Thai documents and hence improve searching efficiency. To investigate the efficiency, a set of experiments using Thai newspapers is conducted in both non-indexing and indexing searching approaches. The experimental results show our method outperform the traditional methods in both nonindexing and indexing approaches in all measures.

This paper presents a method to improve the naturalness of Thai Text-to-speech synthesis, in 4 main parts. In the pausing module, its main function is to determine the break location when synthesizing a Thai text which has no explicit sentence/phrase/word boundary. In the syllable duration and tone generation, a set of rules is provided to generate proper prosodic parameters for synthesizing more natural speech. The syllable duration rule is applied using the Klatt's method to handle the task in syllabic frame. The tonal rule considers the effect of tonal coarticulation and F0 downdrift in generating the F0 contour parameter. In the demisyllable concatenation, the TD-PSOLA technique is applied to modify the waveform for obtaining the required prosody. The LSP-based concatenated boundary smoothing is also included to imitate the crosssyllable coarticulation effect. The result of comparative quality test shows a significant improvement in our proposed method.

In this research we proposes a multilingual learning system for Technical Japanese courses. The new system is based on the concept of interlingua, which is a language-independent representation of all the information that a sentence in any language expresses. Many of the foreign students in Japan are from Asian Countries. In most cases, these students are forced to access Japanese texts and educational materials through both dictionaries and grammatical explanations written in English. Students should be able to study more efficiently if they could learn Japanese directly from their mother tongue. Our project will develop a multilingual dictionary based on an interlingua. This paper describes the background research required for the development of a multilingual dictionary, which includes analysis and discussion of a survey of 1,000 concepts which appear in mainly computer science-related texts, with the corresponding terms in Japanese, Thai, and Chinese languages.

ORCHID (Open linguistic Resources CHanelled toward InterDisciplinary research) is an initiative project aimed at building linguistic resources to support research in, but not limited to, natural language processing. Based on the concept of an open architecture design, the resources must be fully compatible with similar resources, and software tools must also be made available. This paper presents one result of the project, the construction of a Thai part-of-speech (POS) tagged corpus, which is a preliminary stage in the construction of a Thai speech corpus. The POS-tagged corpus is the result of collaborative research between the Communications Research Laboratory (CRL) in Japan and the National Electronics and Computer Technology Center (NECTEC) in Thailand, with technical support from the Electrotechnical Laboratory (ETL) in Japan. In this paper, we propose a new tagset, based on the results of a prior multilingual machine translation project. The corpus is annotated on three levels: the paragraph, sentence, and word levels. Text information is maintained in the form of the text information lines and the number lines, which are both utilized in data retrieval. Both word segmentation and POS tagging were carried out by way of a probabilistic trigram model. Rules for syllable demarkation were additionally used to reduce the number of candidates in computing tagging probabilities. Some typical problems in POS assignment are also formalized to resolve ambiguity.

This paper presents an algorithm for selecting an appropriate classifier word for a noun. In Thai language, it frequently happens that there is fluctuation in the choice of classifier for a given concrete noun, both from the point of view of the whole spe ech community and individual speakers. Basically, there is no exect rule for classifier selection. As far as we can do in the rule-based approach is to give a default rule to pick up a corresponding classifier of each noun. Registration of classifier for each noun is limited to the type of unit classifier because other types are open due to the meaning of representation. We propose a corpus-based method (Biber, 1993; Nagao, 1993; Smadja, 1993) which generates Noun Classifier Associations (NCA) to overcome the problems in classifier assignment and semantic construction of noun phrase. The NCA is created statistically from a large corpus and recomposed under concept hierarchy constraints and frequency of occurrences.