DOCSLIB.ORG

Word Sense Disambiguation and Text Segmentation Based on Lexical

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Word Sense Disambiguation and Text Segmentation Based on Lexical Word Sense i)ismnl}iguati<)n and Text Set mentation Bas(q on I,c×ical (+ohcslOll ()KUMUI{A Manabu, IIONI)A Takeo School of [nforma,tion Science, Japan Advanced Institute of Science a.nd Technology ('l'al.sunokuchi, lshikawa 923-12 Japan) c-nmil: { oku,honda}¢~jaist.ac.ji~ Abstract cohesion is far easier to idenlAfy than reference be- cause 1)oth words in lexical cohesion relation ap- In this paper, we describe ihow word sense am= pear in a text while one word in reference relation biguity can be resolw'.d with the aid of lexical eo- is a pr<mom, or elided and has less information to hesion. By checking ]exical coheshm between the infer the other word in the relation automatically. current word and lexical chains in the order of Based on this observation, we use lexical cohe- the salience, in tandem with getmration of lexica] sion as a linguistic device for discourse analysis. chains~ we realize incretnental word sense disam We call a sequence of words which are in lexieal biguation based on contextual infl)rmation that cohesion relation with each other a Icxical chain lexical chains,reveah Next;, we <le~<:ribe how set like [10]. l,exical chains tend to indicate portions men< boundaries of a text can be determined with of a text; that form a semantic uttit. And so vari.- the aid of lexical cohesion. Wc can measure the ous lexical chains tend to appear in a text corre. plausibility of each point in the text as a segment spou(ling to the change of the topic. Therefore, boundary by computing a degree of agreement of I. lexical chains provide a local context to aid the start and end points of lexical chaihs. in the resolution of word sense ambiguity; 2. lexical <'hains provide a <'lue for the determi- 1 Introduction nation of segnlent boundaries of the text[10]. A text is not a mere set of unrelated sentences. ]n this paper, we first describe how word sense Rather, sentences in a text are about the same ambiguity can t)e resolved with the aid of lexical thing and connected to each other[l()]. Cohesion cohesion. During the process of generating lex- and cohere'nee are said to contribute to such con- i<'al chains incrementally, they are recorded in a nection of the sentences. While coherence is a register in the order of the salience. The salie'ncc semantic relationship and needs computationally of lexical chains is based on their recency and expensive processing for identification, cohesion length. Since the more salient lexical chain rep is a surface relationship among words iu a text resents the nearby local context, by checking lexi: and more accessible than coherence. Cohesion ca[ cohesion between the current word and lexieal is roughly classitled into reference t, co'r@tnction, chains in the order of tile salience, in tandem with and lezical coh,esion 2. generatiou of lexical chains, we realize incremen. Except conjmwtion that explicitly indicates l;he tal word sense disambiguation based on contex- relationship between sentences, l;he other two tual information that lexical chains reveal. <:lasses are considered to t>e similar in that the re- Next;, we describe how segment boundaries of lationship hetweer~ sentences is in<licated by two a text can be determined with the aid of lexical semantically same(or related) words. But lexical cohesion. Since the start and end points of lexical chains it, the text tend to indicate the start and 1Reference by pronouns and ellipsis in Halliday and end points of the segment, we can measure the Hasan's classification[3] are included here. plausibility o[' each point in the text as a segment 2Reference by flfll NPs, substitution mtd lcxical cohe-. sion in Ilalllday and Hasan's classillcation a.re included boundary by computing a degree of agreement of here. the sta.rt and end points of lexical chains. 755 Morris and Itirst[10] pointed out the above and has a hierarchical structure above this level. two importance of lexical cohesion for discourse For each word, a list of category numbers which analysis and presented a way of computing corresponds to its multiple word senses is given. lexical chains by using Roger's International We count a sequence of words which are included Thesaurus[15]. IIowever, in spite of their mention in the same category as a lexical chain. It might to the importance, they did not present the way be (:lear that this task is computationally trivial. of word sense disambiguation based on lexical co- Note that we regard only a sequence of words in hesion and they only showed the correspondences the same category as a lexical chain, rather than between lexical chains and segment boundaries by using the complete Morris and Hirst's framework their intuitive analysis. with five types of thesaural relations. McRoy's work[8] can be considered as the one The word sense of a word can be determined that uses the information of lexical cohesion for in its context. For example, in the context word sense disambiguation, but her method does {universe, star, universe, star, galaxy, sun}, the not; take into account the necessity to arrange word 'earth' has a 'planet' sense, not a 'ground' lexical chains dynamically. Moreover, her word one. As clear from this example, lexical chains sense disambignation method based on lexical co- ('an be used as a contextual aid to resolve word hesion is not evaluated fully. sense ambiguity[10]. In the generation process In section two we outline what lexical cohe- of lexical chains, by choosing the lexical chain sion is. In section three we explain the way of that the current word is added to, its word sense incremental generation of lexical chains in tan- is determined. Thus, we regard word sense dis- dem with word sense disambiguation and describe ambiguation as selecting the most likely category the result of the evaluation of our disambiguation number of the thesaurus, as similar to [16]. method. In section four we explain the measure l';arlier we proposed incremental disambigua- of the plausibility of segment boundaries and de- tion method that uses intrasentential informa- scribe the result of the evaluation of our measure. tion, such as selectional restrictions and case frames[l 2]. In the next section, we describe incre- 2 Lexical Cohesion mental disambiguation method that uses lexical chains as intersentential(contextual) information. Consider the following example, which is the English translation of the fragment of one of Japanese texts that we use for the experiment 3 Generation of Lexical Chains later. In the last section, we showed that lexical chains In the universe that continues expancb carl play a role of local context, t]owever, multi- ing, a number of stars have appeared ple lexical chains might cooccur in portions of a aml disappeared again and again. And text and they might vary in their plausibility as about ten billion years after tile birth local context. For this reason, for lexical chains of the universe, in the same way as to function truly as local context, it is necessary the other stars, a primitive galaxy was to arrange them in the order of the salience that formed with the primitive sun as the indicates the degree of tile plausibility. We base center. the salience on the following two factors: the re- Words {nniverse, star, universe, star, galaxy, cency and the length. The more recently updated sun} seem to be semantically same or related to chains are considered to be the more activated each other and they are included in the same cat- context in the neighborhood and are given more egory in Roget's International Thesaurus. Like salience. The longer chains are considered to be Morris and tIirst, we compute such sequences of more about the topic in the neighborhood and related words(lexical chains) by using a thesaurus are given more salience. as the knowledge base to take into account not By checking lexical cohesion between the cu> only the repetition of the same word but the use rent word and lexical chains in the order of the of superordinates, subordinates, and synonyms. salience, the lexical chain that is selected to add We. use a Japanese thesaurus 'Bnnrui- the current word determines its word sense and goihyo'[1]. Bunrui-goihyo has a similar organi- plays a role of local context. zation to Roger's: it consists of 798 categories Based on this idea, incremental generation of 756 lexical chains realizes incremental word sense dis- If a candidate word can not be added to the ambiguation using contextual information that existing lexical chain, new lexieal chains for each lexical chains reveal. During the generation word sense are recorded in the register. of lexical chains, their salience is also in As clear fl'om tile algorithm, rather than the crementally updated. We think incremental truly incremental method where the register of disambiguation[9] is a better strategy, because lexical chains is updated word by word in a sen- a combinatorial explosion of the number of to tenee, we adopt the incremental method where tal ambiguities rnight occur if ambiguity in not updates are performed at the end of each sentence resolved as early as possible during the analyt because we regard intrasentential information as ical process. Moreover, incremental word sense more iml)ortant. disarnbiguation is indist)ensable during the gem The process of word sense disambiguation us- eration of lexical chains if lexical chains are used ing lexical chains is illustrated in Figure 1. The for incremental analysis, because tile word sense most salient lexical chain is located at the top in ambiguity might cause many undesirable lexical the register.
Load more

Recommended publications
  • Sentence Boundary Detection for Handwritten Text Recognition Matthias Zimmermann
    Sentence Boundary Detection for Handwritten Text Recognition Matthias Zimmermann To cite this version: Matthias Zimmermann. Sentence Boundary Detection for Handwritten Text Recognition. Tenth International Workshop on Frontiers in Handwriting Recognition, Université de Rennes 1, Oct 2006, La Baule (France). inria-00103835 HAL Id: inria-00103835 https://hal.inria.fr/inria-00103835 Submitted on 5 Oct 2006 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Sentence Boundary Detection for Handwritten Text Recognition Matthias Zimmermann International Computer Science Institute Berkeley, CA 94704, USA [email protected] Abstract 1) The summonses say they are ” likely to persevere in such In the larger context of handwritten text recognition sys- unlawful conduct . ” <s> They ... tems many natural language processing techniques can 2) ” It comes at a bad time , ” said Ormston . <s> ”A singularly bad time ... potentially be applied to the output of such systems. How- ever, these techniques often assume that the input is seg- mented into meaningful units, such as sentences. This pa- Figure 1. Typical ambiguity for the position of a sen- per investigates the use of hidden-event language mod- tence boundary token <s> in the context of a period els and a maximum entropy based method for sentence followed by quotes.
    [Show full text]
  • Multiple Segmentations of Thai Sentences for Neural Machine Translation
    Proceedings of the 1st Joint SLTU and CCURL Workshop (SLTU-CCURL 2020), pages 240–244 Language Resources and Evaluation Conference (LREC 2020), Marseille, 11–16 May 2020 c European Language Resources Association (ELRA), licensed under CC-BY-NC Multiple Segmentations of Thai Sentences for Neural Machine Translation Alberto Poncelas1, Wichaya Pidchamook2, Chao-Hong Liu3, James Hadley4, Andy Way1 1ADAPT Centre, School of Computing, Dublin City University, Ireland 2SALIS, Dublin City University, Ireland 3Iconic Translation Machines 4Trinity Centre for Literary and Cultural Translation, Trinity College Dublin, Ireland {alberto.poncelas, andy.way}@adaptcentre.ie [email protected], [email protected], [email protected] Abstract Thai is a low-resource language, so it is often the case that data is not available in sufficient quantities to train an Neural Machine Translation (NMT) model which perform to a high level of quality. In addition, the Thai script does not use white spaces to delimit the boundaries between words, which adds more complexity when building sequence to sequence models. In this work, we explore how to augment a set of English–Thai parallel data by replicating sentence-pairs with different word segmentation methods on Thai, as training data for NMT model training. Using different merge operations of Byte Pair Encoding, different segmentations of Thai sentences can be obtained. The experiments show that combining these datasets, performance is improved for NMT models trained with a dataset that has been split using a supervised splitting tool. Keywords: Machine Translation, Word Segmentation, Thai Language In Machine Translation (MT), low-resource languages are 1. Combination of Segmented Texts especially challenging as the amount of parallel data avail- As the encoder-decoder framework deals with a sequence able to train models may not be enough to achieve high of tokens, a way to address the Thai language is to split the translation quality.
    [Show full text]
  • A Clustering-Based Algorithm for Automatic Document Separation
    A Clustering-Based Algorithm for Automatic Document Separation Kevyn Collins-Thompson Radoslav Nickolov School of Computer Science Microsoft Corporation Carnegie Mellon University 1 Microsoft Way 5000 Forbes Avenue Redmond, WA USA Pittsburgh, PA USA [email protected] [email protected] ABSTRACT For text, audio, video, and still images, a number of projects have addressed the problem of estimating inter-object similarity and the related problem of finding transition, or ‘segmentation’ points in a stream of objects of the same media type. There has been relatively little work in this area for document images, which are typically text-intensive and contain a mixture of layout, text-based, and image features. Beyond simple partitioning, the problem of clustering related page images is also important, especially for information retrieval problems such as document image searching and browsing. Motivated by this, we describe a model for estimating inter-page similarity in ordered collections of document images, based on a combination of text and layout features. The features are used as input to a discriminative classifier, whose output is used in a constrained clustering criterion. We do a task-based evaluation of our method by applying it the problem of automatic document separation during batch scanning. Using layout and page numbering features, our algorithm achieved a separation accuracy of 95.6% on the test collection. Keywords Document Separation, Image Similarity, Image Classification, Optical Character Recognition contiguous set of pages, but be scattered into several 1. INTRODUCTION disconnected, ordered subsets that we would like to recombine. Such scenarios are not uncommon when The problem of accurately determining similarity between scanning large volumes of paper: for example, one pages or documents arises in a number of settings when document may be accidentally inserted in the middle of building systems for managing document image another in the queue.
    [Show full text]
  • An Incremental Text Segmentation by Clustering Cohesion
    An Incremental Text Segmentation by Clustering Cohesion Raúl Abella Pérez and José Eladio Medina Pagola Advanced Technologies Application Centre (CENATAV), 7a #21812 e/ 218 y 222, Rpto. Siboney, Playa, C.P. 12200, Ciudad de la Habana, Cuba {rabella, jmedina} @cenatav.co.cu Abstract. This paper describes a new method, called IClustSeg, for linear text segmentation by topic using an incremental overlapped clustering algorithm. Incremental algorithms are able to process new objects as they are added to the collection and, according to the changes, to update the results using previous information. In our approach, we maintain a structure to get an incremental overlapped clustering. The results of the clustering algorithm, when processing a stream, are used any time text segmentation is required, using the clustering cohesion as the criteria for segmenting by topic. We compare our proposal against the best known methods, outperforming significantly these algorithms. 1 Introduction Topic segmentation intends to identify the boundaries in a document with goal of capturing the latent topical structure. The automatic detection of appropriate subtopic boundaries in a document is a very useful task in text processing. For example, in information retrieval and in passages retrieval, to return documents, segments or passages closer to the user’s queries. Another application of topic segmentation is in summarization, where it can be used to select segments of texts containing the main ideas for the summary requested [6]. Many text segmentation methods by topics have been proposed recently. Usually, they obtain linear segmentations, where the output is a document divided into sequences of adjacent segments [7], [9].
    [Show full text]
  • A Text Denormalization Algorithm Producing Training Data for Text Segmentation
    A Text Denormalization Algorithm Producing Training Data for Text Segmentation Kilian Evang Valerio Basile University of Groningen University of Groningen [email protected] [email protected] Johan Bos University of Groningen [email protected] As a first step of processing, text often has to be split into sentences and tokens. We call this process segmentation. It is often desirable to replace rule-based segmentation tools with statistical ones that can learn from examples provided by human annotators who fix the machine's mistakes. Such a statistical segmentation system is presented in Evang et al. (2013). As training data, the system requires the original raw text as well as information about the boundaries between tokens and sentences within this raw text. Although raw as well as segmented versions of text corpora are available for many languages, this required information is often not trivial to obtain because the segmented version differs from the raw one also in other respects. For example, punctuation marks and diacritics have been normalized to canonical forms by human annotators or by rule-based segmentation and normalization tools. This is the case with e.g. the Penn Treebank, the Dutch Twente News Corpus and the Italian PAISA` corpus. This problem of missing alignment between raw and segmented text is also noted by Dridan and Oepen (2012). We present a heuristic algorithm that recovers the alignment and thereby produces standoff annotations marking token and sentence boundaries in the raw test. The algorithm is based on the Levenshtein algorithm and is general in that it does not assume any language-specific normalization conventions.
    [Show full text]
  • Topic Segmentation: Algorithms and Applications
    University of Pennsylvania ScholarlyCommons IRCS Technical Reports Series Institute for Research in Cognitive Science 8-1-1998 Topic Segmentation: Algorithms And Applications Jeffrey C. Reynar University of Pennsylvania Follow this and additional works at: https://repository.upenn.edu/ircs_reports Part of the Databases and Information Systems Commons Reynar, Jeffrey C., "Topic Segmentation: Algorithms And Applications" (1998). IRCS Technical Reports Series. 66. https://repository.upenn.edu/ircs_reports/66 University of Pennsylvania Institute for Research in Cognitive Science Technical Report No. IRCS-98-21. This paper is posted at ScholarlyCommons. https://repository.upenn.edu/ircs_reports/66 For more information, please contact [email protected]. Topic Segmentation: Algorithms And Applications Abstract Most documents are aboutmore than one subject, but the majority of natural language processing algorithms and information retrieval techniques implicitly assume that every document has just one topic. The work described herein is about clues which mark shifts to new topics, algorithms for identifying topic boundaries and the uses of such boundaries once identified. A number of topic shift indicators have been proposed in the literature. We review these features, suggest several new ones and test most of them in implemented topic segmentation algorithms. Hints about topic boundaries include repetitions of character sequences, patterns of word and word n-gram repetition, word frequency, the presence of cue words and phrases and the use of synonyms. The algorithms we present use cues singly or in combination to identify topic shifts in several kinds of documents. One algorithm tracks compression performance, which is an indicator of topic shift because self-similarity within topic segments should be greater than between-segment similarity.
    [Show full text]
  • A Generic Neural Text Segmentation Model with Pointer Network
    Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI-18) SEGBOT: A Generic Neural Text Segmentation Model with Pointer Network Jing Li, Aixin Sun and Shafiq Joty School of Computer Science and Engineering, Nanyang Technological University, Singapore [email protected], {axsun,srjoty}@ntu.edu.sg Abstract [A person]EDU [who never made a mistake]EDU [never tried any- thing new]EDU Text segmentation is a fundamental task in natu- ral language processing that comes in two levels of Figure 1: A sentence with three elementary discourse units (EDUs). granularity: (i) segmenting a document into a se- and Thompson, 1988]. quence of topical segments (topic segmentation), Both topic and EDU segmentation tasks have received a and (ii) segmenting a sentence into a sequence of lot of attention in the past due to their utility in many NLP elementary discourse units (EDU segmentation). tasks. Although related, these two tasks have been addressed Traditional solutions to the two tasks heavily rely separately with different sets of approaches. Both supervised on carefully designed features. The recently propo- and unsupervised methods have been proposed for topic seg- sed neural models do not need manual feature en- mentation. Unsupervised topic segmentation models exploit gineering, but they either suffer from sparse boun- the strong correlation between topic and lexical usage, and dary tags or they cannot well handle the issue of can be broadly categorized into two classes: similarity-based variable size output vocabulary. We propose a ge- models and probabilistic generative models. The similarity- neric end-to-end segmentation model called SEG- based models are based on the key intuition that sentences BOT.SEGBOT uses a bidirectional recurrent neural in a segment are more similar to each other than to senten- network to encode input text sequence.
    [Show full text]
  • Text Segmentation Techniques: a Critical Review
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Sunway Institutional Repository Text Segmentation Techniques: A Critical Review Irina Pak and Phoey Lee Teh Department of Computing and Information Systems, Sunway University, Bandar Sunway, Malaysia [email protected], [email protected] Abstract Text segmentation is widely used for processing text. It is a method of splitting a document into smaller parts, which is usually called segments. Each segment has its relevant meaning. Those segments categorized as word, sentence, topic, phrase or any information unit depending on the task of the text analysis. This study presents various reasons of usage of text segmentation for different analyzing approaches. We categorized the types of documents and languages used. The main contribution of this study includes a summarization of 50 research papers and an illustration of past decade (January 2007- January 2017)’s of research that applied text segmentation as their main approach for analysing text. Results revealed the popularity of using text segmentation in different languages. Besides that, the “word” seems to be the most practical and usable segment, as it is the smaller unit than the phrase, sentence or line. 1 Introduction Text segmentation is process of extracting coherent blocks of text [1]. The segment referred as “segment boundary” [2] or passage [3]. Another two studies referred segment as subtopic [4] and region of interest [5]. There are many reasons why the splitting document can be useful for text analysis. One of the main reasons is because they are smaller and more coherent than whole documents [3].
    [Show full text]
  • Text Segmentation Based on Semantic Word Embeddings
    Text Segmentation based on Semantic Word Embeddings Alexander A Alemi Paul Ginsparg Dept of Physics Depts of Physics and Information Science Cornell University Cornell University [email protected] [email protected] ABSTRACT early algorithm in this class was Choi's C99 algorithm [3] We explore the use of semantic word embeddings [14, 16, in 2000, which also introduced a benchmark segmentation 12] in text segmentation algorithms, including the C99 seg- dataset used by subsequent work. Instead of looking only at nearest neighbor coherence, the C99 algorithm computes mentation algorithm [3, 4] and new algorithms inspired by 1 the distributed word vector representation. By developing a coherence score between all pairs of elements of text, a general framework for discussing a class of segmentation and searches for a text segmentation that optimizes an ob- objectives, we study the effectiveness of greedy versus ex- jective based on that scoring by greedily making a succes- act optimization approaches and suggest a new iterative re- sion of best cuts. Later work by Choi and collaborators finement technique for improving the performance of greedy [4] used distributed representations of words rather than a strategies. We compare our results to known benchmarks bag of words approach, with the representations generated [18, 15, 3, 4], using known metrics [2, 17]. We demonstrate by LSA [8]. In 2001, Utiyama and Ishahara introduced a state-of-the-art performance for an untrained method with statistical model for segmentation and optimized a poste- our Content Vector Segmentation (CVS) on the Choi test rior for the segment boundaries. Moving beyond the greedy set.
    [Show full text]
  • Steps Involved in Text Recognition and Recent Research in OCR; a Study
    International Journal of Recent Technology and Engineering (IJRTE) ISSN: 2277-3878, Volume-8, Issue-1, May 2019 Steps Involved in Text Recognition and Recent Research in OCR; A Study K.Karthick, K.B.Ravindrakumar, R.Francis, S.Ilankannan Abstract: The Optical Character Recognition (OCR) is one of Preprocessing the automatic identification techniques that fulfill the automation Feature Extraction needs in various applications. A machine can read the Recognition information present in natural scenes or other materials in any Post processing form with OCR. The typed and printed character recognition is uncomplicated due to its well-defined size and shape. The In addition to the above steps segmentation and handwriting of individuals differs in the above aspects. So, the morphological processing also involved in the recognition handwritten OCR system faces complexity to learn this difference process. These steps may be added before the feature to recognize a character. In this paper, we discussed the various extraction process. stages in text recognition, handwritten OCR systems classification according to the text type, study on Chinese and II. PREPROCESSING Arabic text recognition as well as application oriented recent research in OCR. The preprocessing is a fundamental stage that is proceeding Index Terms: Edge detection, Optical Character Recognition, OCR, Preprocessing stages, Text Recognition to the stage of feature extraction; it regulates the appropriateness of the outcomes for the consecutive stages. I. INTRODUCTION The OCR success rate is contingent on the success percentage of each stage. The technologyadventfoundan amazing and noble development curve for the last two centuries. For the last A. Factors Affecting the Text Recognition Quality few decades, it is easy in using mouse and keyboard to assist as interfacing device between us and computer.
    [Show full text]
  • Multiple Segmentations of Thai Sentences for Neural Machine
    Multiple Segmentations of Thai Sentences for Neural Machine Translation Alberto Poncelas1, Wichaya Pidchamook2, Chao-Hong Liu3, James Hadley4, Andy Way1 1ADAPT Centre, School of Computing, Dublin City University, Ireland 2SALIS, Dublin City University, Ireland 3Iconic Translation Machines 4Trinity Centre for Literary and Cultural Translation, Trinity College Dublin, Ireland {alberto.poncelas, andy.way}@adaptcentre.ie [email protected], [email protected], [email protected] Abstract Thai is a low-resource language, so it is often the case that data is not available in sufficient quantities to train an Neural Machine Translation (NMT) model which perform to a high level of quality. In addition, the Thai script does not use white spaces to delimit the boundaries between words, which adds more complexity when building sequence to sequence models. In this work, we explore how to augment a set of English–Thai parallel data by replicating sentence-pairs with different word segmentation methods on Thai, as training data for NMT model training. Using different merge operations of Byte Pair Encoding, different segmentations of Thai sentences can be obtained. The experiments show that combining these datasets, performance is improved for NMT models trained with a dataset that has been split using a supervised splitting tool. Keywords: Machine Translation, Word Segmentation, Thai Language In Machine Translation (MT), low-resource languages are sentence into words (or tokens). We investigate three split- especially challenging as the amount of parallel data avail- ting strategies: (i) Character-based, using each character as able to train models may not be enough to achieve high token. This is the simplest approach as there is no need to translation quality.
    [Show full text]
  • A Journey Through Existing Procedures on Ocr Recognition and Text
    A JOURNEY THROUGH EXISTING PROCEDURES ON OCR RECOGNITION AND TEXT TO SPEECH CONVERSION PROCESS OF BILINGUAL LANGUAGES K.Shanmugam1, Dr.B.vanathi2 1Assistant Professor,2Professor, 1,2Department of Computer Science and Engineering Valliammai Engineering College, Chennai,Tamilnadu ABSTRACT Microcontroller to hear the spoken words of the In this paper we try to explore various text through the earphone or speaker plugged Character recognition and Text-to-Speech into it through its audio jack. synthesis techniques developed and OCR is one of the challenging areas of character implemented by several researchers and recognition and it has various practical research groups over the world during the last applications. decades. Various existing proposals focus on A Text-To-Speech (TTS) synthesizer is a the implementation of either OCR or Text to computer-based system that should be able to Speech synthesis. Text segmentation and read any text aloud, whether it was directly recognition of Tamil characters poses a new introduced in the computer by an operator or challenge owing to its large number of scanned and submitted to an Optical Character characters, curves, indefinite shapes; Recognition (OCR) system. techniques that recognise such text are The recognised text is converted into discussed. This paper aims to give an overview speech by using text to speech synthesizer. of OCR and text synthesis for Bilingual languages (English and Tamil languages), summarizes and compares the characteristics of various synthesis techniques used. KEYWORDS: Optical character recognition, Text-To- Speech (TTS) synthesizer, Win32 SAPI. INTRODUCTION: Machine replication of human functions like reading, is an ancient dream. However, over the last five decades, machine reading has grown from dream to reality.
    [Show full text]