This document describes the development of a text-to-speech synthesizer for the Pali language. It discusses previous work on speech synthesis systems for Indian languages. It then outlines the methodology used, including developing a phone set and speech database for Pali, and using a unit selection approach for speech synthesis. The system was evaluated based on the naturalness of the synthesized speech output. Results showed smooth spectral changes at concatenation points and uniform spectral changes across syllable boundaries, indicating the system produces intelligible synthetic Pali speech.
Improving Sentiment Analysis of Short Informal Indonesian Product Reviews usi...TELKOMNIKA JOURNAL
Sentiment analysis in short informal texts like product reviews is more challenging. Short texts are
sparse, noisy, and lack of context information. Traditional text classification methods may not be suitable
for analyzing sentiment of short texts given all those difficulties. A common approach to overcome these
problems is to enrich the original texts with additional semantics to make it appear like a large document of
text. Then, traditional classification methods can be applied to it. In this study, we developed an automatic
sentiment analysis system of short informal Indonesian texts using Naïve Bayes and Synonym Based
Feature Expansion. The system consists of three main stages, preprocessing and normalization, features
expansion and classification. After preprocessing and normalization, we utilize Kateglo to find some
synonyms of every words in original texts and append them. Finally, the text is classified using Naïve
Bayes. The experiment shows that the proposed method can improve the performance of sentiment
analysis of short informal Indonesian product reviews. The best sentiment classification performance using
proposed feature expansion is obtained by accuracy of 98%.The experiment also show that feature
expansion will give higher improvement in small number of training data than in the large number of them.
Sentiment analysis is an important current research area. The demand for sentiment analysis and classification is growing day by day; this paper presents a novel method to classify Urdu documents as previously no work recorded on sentiment classification for Urdu text. We consider the problem by determining whether the review or sentence is positive, negative or neutral. For the purpose we use two machine learning methods Naïve Bayes and Support Vector Machines (SVM) . Firstly the documents are preprocessed and the sentiments features are extracted, then the polarity has been calculated, judged and classify through Machine learning methods.
Myanmar news summarization using different word representations IJECEIAES
There is enormous amount information available in different forms of sources and genres. In order to extract useful information from a massive amount of data, automatic mechanism is required. The text summarization systems assist with content reduction keeping the important information and filtering the non-important parts of the text. Good document representation is really important in text summarization to get relevant information. Bag-ofwords cannot give word similarity on syntactic and semantic relationship. Word embedding can give good document representation to capture and encode the semantic relation between words. Therefore, centroid based on word embedding representation is employed in this paper. Myanmar news summarization based on different word embedding is proposed. In this paper, Myanmar local and international news are summarized using centroid-based word embedding summarizer using the effectiveness of word representation approach, word embedding. Experiments were done on Myanmar local and international news dataset using different word embedding models and the results are compared with performance of bag-of-words summarization. Centroid summarization using word embedding performs comprehensively better than centroid summarization using bag-of-words.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
Mining Opinion Features in Customer ReviewsIJCERT JOURNAL
Now days, E-commerce systems have become extremely important. Large numbers of customers are choosing online shopping because of its convenience, reliability, and cost. Client generated information and especially item reviews are significant sources of data for consumers to make informed buy choices and for makers to keep track of customer’s opinions. It is difficult for customers to make purchasing decisions based on only pictures and short product descriptions. On the other hand, mining product reviews has become a hot research topic and prior researches are mostly based on pre-specified product features to analyse the opinions. Natural Language Processing (NLP) techniques such as NLTK for Python can be applied to raw customer reviews and keywords can be extracted. This paper presents a survey on the techniques used for designing software to mine opinion features in reviews. Elven IEEE papers are selected and a comparison is made between them. These papers are representative of the significant improvements in opinion mining in the past decade.
Improving Sentiment Analysis of Short Informal Indonesian Product Reviews usi...TELKOMNIKA JOURNAL
Sentiment analysis in short informal texts like product reviews is more challenging. Short texts are
sparse, noisy, and lack of context information. Traditional text classification methods may not be suitable
for analyzing sentiment of short texts given all those difficulties. A common approach to overcome these
problems is to enrich the original texts with additional semantics to make it appear like a large document of
text. Then, traditional classification methods can be applied to it. In this study, we developed an automatic
sentiment analysis system of short informal Indonesian texts using Naïve Bayes and Synonym Based
Feature Expansion. The system consists of three main stages, preprocessing and normalization, features
expansion and classification. After preprocessing and normalization, we utilize Kateglo to find some
synonyms of every words in original texts and append them. Finally, the text is classified using Naïve
Bayes. The experiment shows that the proposed method can improve the performance of sentiment
analysis of short informal Indonesian product reviews. The best sentiment classification performance using
proposed feature expansion is obtained by accuracy of 98%.The experiment also show that feature
expansion will give higher improvement in small number of training data than in the large number of them.
Sentiment analysis is an important current research area. The demand for sentiment analysis and classification is growing day by day; this paper presents a novel method to classify Urdu documents as previously no work recorded on sentiment classification for Urdu text. We consider the problem by determining whether the review or sentence is positive, negative or neutral. For the purpose we use two machine learning methods Naïve Bayes and Support Vector Machines (SVM) . Firstly the documents are preprocessed and the sentiments features are extracted, then the polarity has been calculated, judged and classify through Machine learning methods.
Myanmar news summarization using different word representations IJECEIAES
There is enormous amount information available in different forms of sources and genres. In order to extract useful information from a massive amount of data, automatic mechanism is required. The text summarization systems assist with content reduction keeping the important information and filtering the non-important parts of the text. Good document representation is really important in text summarization to get relevant information. Bag-ofwords cannot give word similarity on syntactic and semantic relationship. Word embedding can give good document representation to capture and encode the semantic relation between words. Therefore, centroid based on word embedding representation is employed in this paper. Myanmar news summarization based on different word embedding is proposed. In this paper, Myanmar local and international news are summarized using centroid-based word embedding summarizer using the effectiveness of word representation approach, word embedding. Experiments were done on Myanmar local and international news dataset using different word embedding models and the results are compared with performance of bag-of-words summarization. Centroid summarization using word embedding performs comprehensively better than centroid summarization using bag-of-words.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
Mining Opinion Features in Customer ReviewsIJCERT JOURNAL
Now days, E-commerce systems have become extremely important. Large numbers of customers are choosing online shopping because of its convenience, reliability, and cost. Client generated information and especially item reviews are significant sources of data for consumers to make informed buy choices and for makers to keep track of customer’s opinions. It is difficult for customers to make purchasing decisions based on only pictures and short product descriptions. On the other hand, mining product reviews has become a hot research topic and prior researches are mostly based on pre-specified product features to analyse the opinions. Natural Language Processing (NLP) techniques such as NLTK for Python can be applied to raw customer reviews and keywords can be extracted. This paper presents a survey on the techniques used for designing software to mine opinion features in reviews. Elven IEEE papers are selected and a comparison is made between them. These papers are representative of the significant improvements in opinion mining in the past decade.
Parameters Optimization for Improving ASR Performance in Adverse Real World N...Waqas Tariq
From the existing research it has been observed that many techniques and methodologies are available for performing every step of Automatic Speech Recognition (ASR) system, but the performance (Minimization of Word Error Recognition-WER and Maximization of Word Accuracy Rate- WAR) of the methodology is not dependent on the only technique applied in that method. The research work indicates that, performance mainly depends on the category of the noise, the level of the noise and the variable size of the window, frame, frame overlap etc is considered in the existing methods. The main aim of the work presented in this paper is to use variable size of parameters like window size, frame size and frame overlap percentage to observe the performance of algorithms for various categories of noise with different levels and also train the system for all size of parameters and category of real world noisy environment to improve the performance of the speech recognition system. This paper presents the results of Signal-to-Noise Ratio (SNR) and Accuracy test by applying variable size of parameters. It is observed that, it is really very hard to evaluate test results and decide parameter size for ASR performance improvement for its resultant optimization. Hence, this study further suggests the feasible and optimum parameter size using Fuzzy Inference System (FIS) for enhancing resultant accuracy in adverse real world noisy environmental conditions. This work will be helpful to give discriminative training of ubiquitous ASR system for better Human Computer Interaction (HCI). Keywords: ASR Performance, ASR Parameters Optimization, Multi-Environmental Training, Fuzzy Inference System for ASR, ubiquitous ASR system, Human Computer Interaction (HCI)
A Novel Approach for Rule Based Translation of English to Marathiaciijournal
This paper presents a design for rule-based machine translation system for English to Marathi language pair. The machine translation system will take input script as English sentence and parse with the help of Stanford parser. The Stanford parser will be used for main purposes on the source side processing, in the machine translation system. English to Marathi Bilingual dictionary is going to be created. The system will take the parsed output and separate the source text word by word and searches for their corresponding target words in the bilingual dictionary. The hand coded rules are written for Marathi inflections and also reordering rules are there. After applying the reordering rules, English sentence will be syntactically reordered to suit Marathi language
ADVANCEMENTS ON NLP APPLICATIONS FOR MANIPURI LANGUAGEkevig
Manipuri is both a minority and morphologically rich language with genetic features similar to Tibeto Burman languages. It has Subject-Object-Verb (SOV) order, agglutinative verb morphology and ismonosyllabic. Morphology and syntax are not clearly distinguished in this language. Natural Language Processing (NLP) is a useful research field of computer science that deals with processing of a large amount of natural language corpus. The NLP applications encompass E-Dictionary, Morphological
Analyzer, Reduplicated Multi-Word Expression (RMWE), Named Entity Recognition (NER), Part of Speech (POS) Tagging, Machine Translation (MT), Word Net, Word Sense Disambiguation (WSD) etc. In this paper, we present a study on the advancements in NLP applications for Manipuri language, at the same time presenting a comparison table of the approaches and techniques adopted and the results obtained of each of the applications followed by a detail discussion of each work.
Abstract
Part of speech tagging plays an important role in developing natural language processing software. Part of speech tagging means assigning part of speech tag to each word of the sentence. The part of speech tagger takes a sentence as input and it assigns respective/appropriate part of speech tag to each word of that sentence. In this article I surveys the different work have done about odia POS tagging.
________________________________________________
Design of A Spell Corrector For Hausa LanguageWaqas Tariq
In this article, a spell corrector has been designed for the Hausa language which is the second most spoken language in Africa and do not yet have processing tools. This study is a contribution to the automatic processing of the Hausa language. We used existing techniques for other languages and adapted them to the special case of the Hausa language. The corrector designed operates essentially on Mijinguini’s dictionary and characteristics of the Hausa alphabet. After a brief review on spell checking and spell correcting techniques and the state of art in the Hausa language processing, we opted for the data structures trie and hash table to represent the dictionary. The edit distance and the specificities of the Hausa alphabet have been used to detect and correct spelling errors. The implementation of the spell corrector has been made on a special editor developed for that purpose (LyTexEditor) but also as an extension (add-on) for OpenOffice.org. A comparison was made on the performance of the two data structures used.
A New Approach to Parts of Speech Tagging in Malayalamijcsit
Parts-of-speech tagging is the process of labeling each word in a sentence. A tag mentions the word’s
usage in the sentence. Usually, these tags indicate syntactic classification like noun or verb, and sometimes
include additional information, with case markers (number, gender etc) and tense markers. A large number
of current language processing systems use a parts-of-speech tagger for pre-processing.
There are mainly two approaches usually followed in Parts of Speech Tagging. Those are Rule based
Approach and Stochastic Approach. Rule based Approach use predefined handwritten rules. This is the
oldest approach and it use lexicon or dictionary for reference. Stochastic Approach use probabilistic and
statistical information to assign tag to words. It use large corpus, so that Time complexity and Space
complexity is high whereas Rule base approach has less complexity for both Time and Space. Stochastic
Approach is the widely used one nowadays because of its accuracy.
Malayalam is a Dravidian family of languages, inflectional with suffixes with the root word forms. The
currently used Algorithms are efficient Machine Learning Algorithms but these are not built for
Malayalam. So it affects the accuracy of the result of Malayalam POS Tagging.
My proposed Approach use Dictionary entries along with adjacent tag information. This algorithm use
Multithreaded Technology. Here tagging done with the probability of the occurrence of the sentence
structure along with the dictionary entry.
Survey on Indian CLIR and MT systems in Marathi LanguageEditor IJCATR
Cross Language Information Retrieval (CLIR) deals with retrieving relevant information stored in a language different from
the language of user’s query. This helps users to express the information need in their native languages. Machine translation based (MTbased)
approach of CLIR uses existing machine translation techniques to provide automatic translation of queries. This paper covers the
research work done in CLIR and MT systems for Marathi language in India.
Punjabi to Hindi Transliteration System for Proper Nouns Using Hybrid ApproachIJERA Editor
The language is an effective medium for the communication that conveys the ideas and expression of the human
mind. There are more than 5000 languages in the world for the communication. To know all these languages is
not a solution for problems due to the language barrier in communication. In this multilingual world with the
huge amount of information exchanged between various regions and in different languages in digitized format,
it has become necessary to find an automated process to convert from one language to another. Natural
Language Processing (NLP) is one of the hot areas of research that explores how computers can be utilizing to
understand and manipulate natural language text or speech. In the Proposed system a Hybrid approach to
transliterate the proper nouns from Punjabi to Hindi is developed. Hybrid approach in the proposed system is a
combination of Direct Mapping, Rule based approach and Statistical Machine Translation approach (SMT).
Proposed system is tested on various proper nouns from different domains and accuracy of the proposed system
is very good.
ON THE UTILITY OF A SYLLABLE-LIKE SEGMENTATION FOR LEARNING A TRANSLITERATION...cscpconf
Source and target word segmentation and alignment is a primary step in the statistical learning of a Transliteration. Here, we analyze the benefit of a syllable-like segmentation approach for learning a transliteration from English to an Indic language, which aligns the training set word pairs in terms of sub-syllable-like units instead of individual character units. While this has been found useful in the case of dealing with Out-of-vocabulary words in English-Chinese in the presence of multiple target dialects, we asked if this would be true for Indic languages which are simpler in their phonetic representation and pronunciation. We expected this syllable-like method to perform marginally better, but we found instead that even though our proposed approach improved the Top-1 accuracy, the individual-character-unit alignment model
somewhat outperformed our approach when the Top-10 results of the system were re-ranked using language modeling approaches. Our experiments were conducted for English to Telugu transliteration (our method will apply equally well to most written Indic languages); our training consisted of a syllable-like segmentation and alignment of a large training set, on which we built a statistical model by modifying a previous character-level maximum entropy based Transliteration learning system due to Kumaran and Kellner; our testing consisted of using the same segmentation of a test English word, followed by applying the model, and reranking the resulting top 10 Telugu words. We also report the dataset creation and selection since standard datasets are not available.
Speech synthesis we can, in theory, mean any kind of synthetization of speech. For example, it can be the process in which a speech decoder generates the speech signal based on the parameters it has received through the transmission line, or it can be a procedure performed by a computer to estimate some kind of a presentation of the speech signal given a text input. Since there is a special course about the codecs (Puheen koodaus, Speech Coding), this chapter will concentrate on text-to-speech synthesis, or shortly TTS, which will be often referred to as speech synthesis to simplify the notation. Anyway, it is good to keep in mind that irrespective of what kind of synthesis we are dealing with, there are similar criteria in regard to the speech quality. We will return to this topic after a brief TTS motivation, and the rest of this chapter will be dedicated to the implementation point of view in TTS systems. Text-to-speech synthesis is a research field that has received a lot of attention and resources during the last couple of decades – for excellent reasons. One of the most interesting ideas (rather futuristic, though) is the fact that a workable TTS system, combined with a workable speech recognition device, would actually be an extremely efficient method for speech coding). It would provide incomparable compression ratio and flexible possibilities to choose the type of speech (e.g., breathless or hoarse), the fundamental frequency along with its range, the rhythm of speech, and several other effects. Furthermore, if the content of a message needs to be changed, it is much easier to retype the text than to record the signal again. Unfortunately this kind of a system does not yet exist for large vocabularies. Of course there are also numerous speech synthesis applications that are closer to being available than the one discussed above. For instance, a telephone inquiry system where the information is frequently updated, can use TTS to deliver answers to the customers. Speech synthesizers are also important to the visually impaired and to those who have lost their ability to speak. Several other examples can be found in everyday life, such as listening to the messages and news instead of reading them, and using hands-free functions through a voice interface in a car, and so on.
A decision tree based word sense disambiguation system in manipuri languageacijjournal
This paper manifests a primary attempt on building a word sense disambiguation system in Manipuri
language. The paper discusses related attempts made in the Manipuri language followed by the proposed
plan. A database, consisting of 650 sentences, is collected in Manipuri language in the course of the study.
Conventional positional and context based features are suggested to capture the sense of the words, which
have ambiguous and multiple senses. The proposed work is expected to predict the senses of the
polysemous words with high accuracy with the help of the suitable knowledge acquisition techniques. The
system produces an accuracy of 71.75 %.
Parameters Optimization for Improving ASR Performance in Adverse Real World N...Waqas Tariq
From the existing research it has been observed that many techniques and methodologies are available for performing every step of Automatic Speech Recognition (ASR) system, but the performance (Minimization of Word Error Recognition-WER and Maximization of Word Accuracy Rate- WAR) of the methodology is not dependent on the only technique applied in that method. The research work indicates that, performance mainly depends on the category of the noise, the level of the noise and the variable size of the window, frame, frame overlap etc is considered in the existing methods. The main aim of the work presented in this paper is to use variable size of parameters like window size, frame size and frame overlap percentage to observe the performance of algorithms for various categories of noise with different levels and also train the system for all size of parameters and category of real world noisy environment to improve the performance of the speech recognition system. This paper presents the results of Signal-to-Noise Ratio (SNR) and Accuracy test by applying variable size of parameters. It is observed that, it is really very hard to evaluate test results and decide parameter size for ASR performance improvement for its resultant optimization. Hence, this study further suggests the feasible and optimum parameter size using Fuzzy Inference System (FIS) for enhancing resultant accuracy in adverse real world noisy environmental conditions. This work will be helpful to give discriminative training of ubiquitous ASR system for better Human Computer Interaction (HCI). Keywords: ASR Performance, ASR Parameters Optimization, Multi-Environmental Training, Fuzzy Inference System for ASR, ubiquitous ASR system, Human Computer Interaction (HCI)
A Novel Approach for Rule Based Translation of English to Marathiaciijournal
This paper presents a design for rule-based machine translation system for English to Marathi language pair. The machine translation system will take input script as English sentence and parse with the help of Stanford parser. The Stanford parser will be used for main purposes on the source side processing, in the machine translation system. English to Marathi Bilingual dictionary is going to be created. The system will take the parsed output and separate the source text word by word and searches for their corresponding target words in the bilingual dictionary. The hand coded rules are written for Marathi inflections and also reordering rules are there. After applying the reordering rules, English sentence will be syntactically reordered to suit Marathi language
ADVANCEMENTS ON NLP APPLICATIONS FOR MANIPURI LANGUAGEkevig
Manipuri is both a minority and morphologically rich language with genetic features similar to Tibeto Burman languages. It has Subject-Object-Verb (SOV) order, agglutinative verb morphology and ismonosyllabic. Morphology and syntax are not clearly distinguished in this language. Natural Language Processing (NLP) is a useful research field of computer science that deals with processing of a large amount of natural language corpus. The NLP applications encompass E-Dictionary, Morphological
Analyzer, Reduplicated Multi-Word Expression (RMWE), Named Entity Recognition (NER), Part of Speech (POS) Tagging, Machine Translation (MT), Word Net, Word Sense Disambiguation (WSD) etc. In this paper, we present a study on the advancements in NLP applications for Manipuri language, at the same time presenting a comparison table of the approaches and techniques adopted and the results obtained of each of the applications followed by a detail discussion of each work.
Abstract
Part of speech tagging plays an important role in developing natural language processing software. Part of speech tagging means assigning part of speech tag to each word of the sentence. The part of speech tagger takes a sentence as input and it assigns respective/appropriate part of speech tag to each word of that sentence. In this article I surveys the different work have done about odia POS tagging.
________________________________________________
Design of A Spell Corrector For Hausa LanguageWaqas Tariq
In this article, a spell corrector has been designed for the Hausa language which is the second most spoken language in Africa and do not yet have processing tools. This study is a contribution to the automatic processing of the Hausa language. We used existing techniques for other languages and adapted them to the special case of the Hausa language. The corrector designed operates essentially on Mijinguini’s dictionary and characteristics of the Hausa alphabet. After a brief review on spell checking and spell correcting techniques and the state of art in the Hausa language processing, we opted for the data structures trie and hash table to represent the dictionary. The edit distance and the specificities of the Hausa alphabet have been used to detect and correct spelling errors. The implementation of the spell corrector has been made on a special editor developed for that purpose (LyTexEditor) but also as an extension (add-on) for OpenOffice.org. A comparison was made on the performance of the two data structures used.
A New Approach to Parts of Speech Tagging in Malayalamijcsit
Parts-of-speech tagging is the process of labeling each word in a sentence. A tag mentions the word’s
usage in the sentence. Usually, these tags indicate syntactic classification like noun or verb, and sometimes
include additional information, with case markers (number, gender etc) and tense markers. A large number
of current language processing systems use a parts-of-speech tagger for pre-processing.
There are mainly two approaches usually followed in Parts of Speech Tagging. Those are Rule based
Approach and Stochastic Approach. Rule based Approach use predefined handwritten rules. This is the
oldest approach and it use lexicon or dictionary for reference. Stochastic Approach use probabilistic and
statistical information to assign tag to words. It use large corpus, so that Time complexity and Space
complexity is high whereas Rule base approach has less complexity for both Time and Space. Stochastic
Approach is the widely used one nowadays because of its accuracy.
Malayalam is a Dravidian family of languages, inflectional with suffixes with the root word forms. The
currently used Algorithms are efficient Machine Learning Algorithms but these are not built for
Malayalam. So it affects the accuracy of the result of Malayalam POS Tagging.
My proposed Approach use Dictionary entries along with adjacent tag information. This algorithm use
Multithreaded Technology. Here tagging done with the probability of the occurrence of the sentence
structure along with the dictionary entry.
Survey on Indian CLIR and MT systems in Marathi LanguageEditor IJCATR
Cross Language Information Retrieval (CLIR) deals with retrieving relevant information stored in a language different from
the language of user’s query. This helps users to express the information need in their native languages. Machine translation based (MTbased)
approach of CLIR uses existing machine translation techniques to provide automatic translation of queries. This paper covers the
research work done in CLIR and MT systems for Marathi language in India.
Punjabi to Hindi Transliteration System for Proper Nouns Using Hybrid ApproachIJERA Editor
The language is an effective medium for the communication that conveys the ideas and expression of the human
mind. There are more than 5000 languages in the world for the communication. To know all these languages is
not a solution for problems due to the language barrier in communication. In this multilingual world with the
huge amount of information exchanged between various regions and in different languages in digitized format,
it has become necessary to find an automated process to convert from one language to another. Natural
Language Processing (NLP) is one of the hot areas of research that explores how computers can be utilizing to
understand and manipulate natural language text or speech. In the Proposed system a Hybrid approach to
transliterate the proper nouns from Punjabi to Hindi is developed. Hybrid approach in the proposed system is a
combination of Direct Mapping, Rule based approach and Statistical Machine Translation approach (SMT).
Proposed system is tested on various proper nouns from different domains and accuracy of the proposed system
is very good.
ON THE UTILITY OF A SYLLABLE-LIKE SEGMENTATION FOR LEARNING A TRANSLITERATION...cscpconf
Source and target word segmentation and alignment is a primary step in the statistical learning of a Transliteration. Here, we analyze the benefit of a syllable-like segmentation approach for learning a transliteration from English to an Indic language, which aligns the training set word pairs in terms of sub-syllable-like units instead of individual character units. While this has been found useful in the case of dealing with Out-of-vocabulary words in English-Chinese in the presence of multiple target dialects, we asked if this would be true for Indic languages which are simpler in their phonetic representation and pronunciation. We expected this syllable-like method to perform marginally better, but we found instead that even though our proposed approach improved the Top-1 accuracy, the individual-character-unit alignment model
somewhat outperformed our approach when the Top-10 results of the system were re-ranked using language modeling approaches. Our experiments were conducted for English to Telugu transliteration (our method will apply equally well to most written Indic languages); our training consisted of a syllable-like segmentation and alignment of a large training set, on which we built a statistical model by modifying a previous character-level maximum entropy based Transliteration learning system due to Kumaran and Kellner; our testing consisted of using the same segmentation of a test English word, followed by applying the model, and reranking the resulting top 10 Telugu words. We also report the dataset creation and selection since standard datasets are not available.
Speech synthesis we can, in theory, mean any kind of synthetization of speech. For example, it can be the process in which a speech decoder generates the speech signal based on the parameters it has received through the transmission line, or it can be a procedure performed by a computer to estimate some kind of a presentation of the speech signal given a text input. Since there is a special course about the codecs (Puheen koodaus, Speech Coding), this chapter will concentrate on text-to-speech synthesis, or shortly TTS, which will be often referred to as speech synthesis to simplify the notation. Anyway, it is good to keep in mind that irrespective of what kind of synthesis we are dealing with, there are similar criteria in regard to the speech quality. We will return to this topic after a brief TTS motivation, and the rest of this chapter will be dedicated to the implementation point of view in TTS systems. Text-to-speech synthesis is a research field that has received a lot of attention and resources during the last couple of decades – for excellent reasons. One of the most interesting ideas (rather futuristic, though) is the fact that a workable TTS system, combined with a workable speech recognition device, would actually be an extremely efficient method for speech coding). It would provide incomparable compression ratio and flexible possibilities to choose the type of speech (e.g., breathless or hoarse), the fundamental frequency along with its range, the rhythm of speech, and several other effects. Furthermore, if the content of a message needs to be changed, it is much easier to retype the text than to record the signal again. Unfortunately this kind of a system does not yet exist for large vocabularies. Of course there are also numerous speech synthesis applications that are closer to being available than the one discussed above. For instance, a telephone inquiry system where the information is frequently updated, can use TTS to deliver answers to the customers. Speech synthesizers are also important to the visually impaired and to those who have lost their ability to speak. Several other examples can be found in everyday life, such as listening to the messages and news instead of reading them, and using hands-free functions through a voice interface in a car, and so on.
A decision tree based word sense disambiguation system in manipuri languageacijjournal
This paper manifests a primary attempt on building a word sense disambiguation system in Manipuri
language. The paper discusses related attempts made in the Manipuri language followed by the proposed
plan. A database, consisting of 650 sentences, is collected in Manipuri language in the course of the study.
Conventional positional and context based features are suggested to capture the sense of the words, which
have ambiguous and multiple senses. The proposed work is expected to predict the senses of the
polysemous words with high accuracy with the help of the suitable knowledge acquisition techniques. The
system produces an accuracy of 71.75 %.
Isolation and Purification of Secoisolariciresinoldiglucoside oligomers (Lign...IOSR Journals
The present study aimed to extract and purify the compound of Secoisolariciresinoldiglucoside
oligomers (lignan) from flax seed (Linumusitatissimum) and its antioxidant activity. The Lignan was extracted
by solvents which gave the best results were ethanol : 1,4 dioxane (1:1, v:v).SDG release after alkaline
hydrolysisby using a methanolicNaOH , 20 mM, pH=8 at 50 ºC.followed by using following chromatographic
techniques: Liquid-liquid, Sephadex LH-20 column chromatography, thin layer chromatographic (TLC), high
performance liquid chromatographic (HPLC) and Fourier Transform Infra-Red(FTIR) . The EC50 values of
Pure lignan extract (9 μg/ml) was shown possess DPPH radical scavenging activity compared to reference
substances BHT and vitamin C (EC50= 3 and 4.2 μg/ml) respectively, and this was higher than partial pure
lignan component (EC50= 25.5 μg/ml).The total phenolic content of the pure lignanwas higher than partial
pure lignan which gave 22.312 and 14.85 g/ml respectively.
Zooplanktonic Diversity and Trophic Status of Pashupatinath Pond in Relation ...IOSR Journals
Present investigation were carried out to physico chemical characterstics and trophic status of
Pashupatinath pond Mandsour (M.P.).water sample were collected seasonal basis for a period of December
2008 to September 2010 using plastic container .standard procedure were followed during collection
,preservation and analysis of water sample for various physicochemical and biological parameter .
The water quality is remained moderately alkaline PH (8.11) while electrical conductivity (0.2176 ms/cm), TDS
(187ppm) chloride (22.123ppm), Hardness(139.166ppm) and alkalinity (75.33ppm) show low mean value
.Average dissolve oxygen levels were at (7.771 ppm) ,while average nitrate and phosphate The water remained
modertly alkaline PH (7.95) while electrical conductance (0.2165ms/cm),TDS (153.66ppm)chloride (22.83ppm)
hardness (138.66ppm )and alkalinity (62.166 ppm) showed low mean values .Average dissolved oxygen level
were at (7.58 ppm ) while average nitrate and phosphate level were (0.2126ppm) and (0.5868 ppm) respectely
.On the basis of water quality parameter .Mirzapur dam was found to be oligotrophic. A low density of
Zooplankton were also observed during the study period
“Trade-Off between Detection and Resolution of Two Point Objects Under Variou...IOSR Journals
It is a well-experienced fact that whenever one tries to detect a weak object point in the vicinity of an intense point object, viz., a binary star-SIRUS and its companion weak satellite star, there is always loss of resolution of the optical system. In other words, one wants to improve the defectively of the system, there is always a loss of resolution capabilities of the system. Thus, there is a trade-off between Detection and Resolution of optical systems under various imaging situations. In this first paper of discussion of this trade-off, we have derived the Fourier analytical formulation of this problem. This formulation will be used to find out a compatible trade-off between Detection and Resolution in our further publications
Simple Thinking Makes Chemistry Metabolic and Interesting - A Review ArticleIOSR Journals
First of all I give my love to all the students who reside in ‘Chemistry Phobia zone’. In this short article I shall try to share my simple thinking that I achieved during my last 16 years of journey in the route of Chemical Education. Before starting my journey it was a big encounter for me to change the conventional methods and makes chemistry easier and interesting to such student who belongs to Chemistry Paranoia Zone. I believe that students are just like flowers and it is our duty to nourish them properly as a gardener. In my infinitely small area of knowledge, I just try to innovate fourteen (14) teaching methodologies by including thirty four (34) completely new formulae in the chemistry world which were followed by different eminent writers in their books since last 90-95 years namely Sir G.Wilkinson, Prof. F. Albert Cotton, Prof. I.L.Finar, Prof. R.T. Morrison, Prof. R.N. Boyd, Prof. Solomons and Prof. Fryhle, Prof. J.G. Smith etc..
Influence of chimney effect on the radon effective dose of the lung simulated...IOSR Journals
One of the well-known radon prone areas of the world is Ramsar in Iran, which is surrounded by the
Alborz Mountain in its southern part and Caspian Sea on the north. The annual effective dose in the district of
Talesh-Mahalleh is higher than the annual dose limits for radiation workers. In this study, the indoor radon
level and effective dose of the lung were estimated using a Prassi portable radon gas survey meter in a model
house containing top soil samples from different parts of Ramsar. For the extremely hot samples, the effective
dose of the lung in winter season was 27.75±2.55mSv, when the windows and exhaust part of chimney were
closed. However, when the chimney was turned on and the exhaust part of chimney was open, the effective dose
of the lung was reduced to 1.27±0.23mSv. Also the seasonal radon effective doses of the lung with other samples
were reduced to low values. The results suggest by using chimney effect and chimney heaters a significant
lessening of the radon seasonal effective dose in dwellings of Ramsar can be achieved.
Implementation of Marathi Language Speech Databases for Large Dictionaryiosrjce
IOSR journal of VLSI and Signal Processing (IOSRJVSP) is a double blind peer reviewed International Journal that publishes articles which contribute new results in all areas of VLSI Design & Signal Processing. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced VLSI Design & Signal Processing concepts and establishing new collaborations in these areas.
Design and realization of microelectronic systems using VLSI/ULSI technologies require close collaboration among scientists and engineers in the fields of systems architecture, logic and circuit design, chips and wafer fabrication, packaging, testing and systems applications. Generation of specifications, design and verification must be performed at all abstraction levels, including the system, register-transfer, logic, circuit, transistor and process levels
Artificially Generatedof Concatenative Syllable based Text to Speech Synthesi...iosrjce
IOSR journal of VLSI and Signal Processing (IOSRJVSP) is a double blind peer reviewed International Journal that publishes articles which contribute new results in all areas of VLSI Design & Signal Processing. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced VLSI Design & Signal Processing concepts and establishing new collaborations in these areas.Design and realization of microelectronic systems using VLSI/ULSI technologies require close collaboration among scientists and engineers in the fields of systems architecture, logic and circuit design, chips and wafer fabrication, packaging, testing and systems applications. Generation of specifications, design and verification must be performed at all abstraction levels, including the system, register-transfer, logic, circuit, transistor and process levels.
Speech is the important mode of communication and is the current research
topic. The concentration is mostly focused on synthesis and analyzing part.Apart of
synthesizing, text to speech system is developed.Speech synthesis is an artificial production
of human speech.A text to speech system (TTS) is to convert an arbitrary text into speech.In
India different languages have been spoken each being the mother tongue of tens of millions
of people.In this paper,the text to speech system is primarily developed for Telugu, a
Dravidian language predominantly spoken in Indian state of Andhra Pradesh.The
important qualities expected from this system are naturalness and intelligibility.Telugu TTS
can be developed using other synthesis methods like articulatory synthesis,formant synthesis
and concatenative synthesis.This paper describes a development of a Telugu text to speech
system using concatenative synthesis method on mobile based system OMAP 3530 (ARM
Cortex A-8 core) in Linux.
Approach To Build A Marathi Text-To-Speech System Using Concatenative Synthes...IJERA Editor
Marathi is one of the oldest languages in India. This research paper describes the development of Marathi Textto-
Speech System (TTS). In Marathi TTS the input is Marathi text in Unicode. The voices are sampled from real
recorded speech. The objective of a text to speech system is to convert an arbitrary text into its corresponding
spoken waveform. Speech synthesis is a process of building machinery that can generate human-like speech
from any text input to imitate human speakers. Text processing and speech generation are two main components
of a text to speech system. To build a natural sounding speech synthesis system, it is essential that text
processing component produce an appropriate sequence of phonemic units. Generation of sequence of phonetic
units for a given standard word is referred to as letter to phoneme rule or text to phoneme rule. The
complexity of these rules and their derivation depends upon the nature of the language. The quality of a speech
synthesizer is judged by its closeness to the natural human voice and understandability. In this research paper we
described an approach to build a Marathi TTS system using concatenative synthesis method with syllable as a
basic unit of concatenation.
Dynamic Construction of Telugu Speech Corpus for Voice Enabled Text EditorWaqas Tariq
In recent decades speech interactive systems have gained increasing importance. Performance of an ASR system mainly depends on the availability of large corpus of speech. The conventional method of building a large vocabulary speech recognizer for any language uses a top-down approach to speech. This approach requires large speech corpus with sentence or phoneme level transcription of the speech utterances. The transcriptions must also include different speech order so that the recognizer can build models for all the sounds present. But, for Telugu language, because of its complex nature, a very large, well annotated speech database is very difficult to build. It is very difficult, if not impossible, to cover all the words of any Indian language, where each word may have thousands and millions of word forms. A significant part of grammar that is handled by syntax in English (and other similar languages) is handled within morphology in Telugu. Phrases including several words (that is, tokens) in English would be mapped on to a single word in Telugu.Telugu language is phonetic in nature in addition to rich in morphology. That is why the speech technology developed for English cannot be applied to Telugu language. This paper highlights the work carried out in an attempt to build a voice enabled text editor with capability of automatic term suggestion. Main claim of the paper is the recognition enhancement process developed by us for suitability of highly inflecting, rich morphological languages. This method results in increased speech recognition accuracy with very much reduction in corpus size. It also adapts Telugu words to the database dynamically, resulting in growth of the corpus.
DEVELOPMENT OF PHONEME DOMINATED DATABASE FOR LIMITED DOMAIN T-T-S IN HINDIijaia
Maximum digital information is available to fewer people who can read or understand a particular
language. The corpus is the basis for developing speech synthesis and recognition systems. In India, almost
all speech research and development affiliations are developing their own speech corpora for Hindi
language, which is the first language for more than 200 million people. The primary goal of this paper is to
review the speech corpus created by various institutes and organizations so that the scientists and language
technologists can recognize the crucial role of corpus development in the field of building ASR and TTS
systems. This aim is to bring together all the information related to the recording, volume and quality of
speech data in speech corpus to facilitate the work of researchers in the field of speech recognition and
synthesis. This paper describes development of medium size database for Metro rail passenger information
systems using HMM based technique in our organization for above application. Phoneme is chosen as
basic speech unit of the database. The result shows that a medium size database consisting of 630
utterances with 12,614 words, 11572 tokens of phonemes covering 38 phonemes are generated in our
database and it cover maximum possible phonetic context.
A Review on Speech Corpus Development for Automatic Speech Recognition in Ind...Eswar Publications
Corpus development gained much attention due to recent statistics based natural language processing. It has new
applications in Language Technology, linguistic research, language education and information exchange. Corpus based Language research has an innovative outlook which will discard the aged linguistic theories. Speech corpus is the essential resources for building a speech recognizer. One of the main challenges faced by speech scientist is the unavailability of these resources. Very fewer efforts have been made in Indian languages to make these resources available to public compared to English. In this paper we review the efforts made in Indian languages for developing speech corpus for automatic speech recognition.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
A Corpus-Based Concatenative Speech Synthesis System for Marathiiosrjce
IOSR journal of VLSI and Signal Processing (IOSRJVSP) is a double blind peer reviewed International Journal that publishes articles which contribute new results in all areas of VLSI Design & Signal Processing. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced VLSI Design & Signal Processing concepts and establishing new collaborations in these areas.
Design and realization of microelectronic systems using VLSI/ULSI technologies require close collaboration among scientists and engineers in the fields of systems architecture, logic and circuit design, chips and wafer fabrication, packaging, testing and systems applications. Generation of specifications, design and verification must be performed at all abstraction levels, including the system, register-transfer, logic, circuit, transistor and process levels
A Context-based Numeral Reading Technique for Text to Speech Systems IJECEIAES
This paper presents a novel technique for context based numeral reading in Indian language text to speech systems. The model uses a set of rules to determine the context of the numeral pronunciation and is being integrated with the waveform concatenation technique to produce speech out of the input text in Indian languages. For this purpose, the three Indian languages Odia, Hindi and Bengali are considered. To analyze the performance of the proposed technique, a set of experiments are performed considering different context of numeral pronunciations and the results are compared with existing syllable-based technique. The results obtained from different experiments shows the effectiveness of the proposed technique in producing intelligible speech out of the entered text utterances compared to the existing technique even with very less storage and execution time.
Grapheme-To-Phoneme Tools for the Marathi Speech SynthesisIJERA Editor
We describe in detail a Grapheme-to-Phoneme (G2P) converter required for the development of a good quality
Marathi Text-to-Speech (TTS) system. The Festival and Festvox framework is chosen for developing the
Marathi TTS system. Since Festival does not provide complete language processing support specie to various
languages, it needs to be augmented to facilitate the development of TTS systems in certain new languages.
Because of this, a generic G2P converter has been developed. In the customized Marathi G2P converter, we
have handled schwa deletion and compound word extraction. In the experiments carried out to test the Marathi
G2P on a text segment of 2485 words, 91.47% word phonetisation accuracy is obtained. This Marathi G2P has
been used for phonetising large text corpora which in turn is used in designing an inventory of phonetically rich
sentences. The sentences ensured a good coverage of the phonetically valid di-phones using only 1.3% of the
complete text corpora.
Improvement in Quality of Speech associated with Braille codes - A Reviewinscit2006
J. Anurag, P. Nupur and Agrawal, S.S.
School of Information Technology, Guru Gobind Singh Indraprastha University, Delhi, India
Centre for Development of Advanced Computing, Noida, India
Marathi Text-To-Speech Synthesis using Natural Language Processingiosrjce
IOSR journal of VLSI and Signal Processing (IOSRJVSP) is a double blind peer reviewed International Journal that publishes articles which contribute new results in all areas of VLSI Design & Signal Processing. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced VLSI Design & Signal Processing concepts and establishing new collaborations in these areas.
Design and realization of microelectronic systems using VLSI/ULSI technologies require close collaboration among scientists and engineers in the fields of systems architecture, logic and circuit design, chips and wafer fabrication, packaging, testing and systems applications. Generation of specifications, design and verification must be performed at all abstraction levels, including the system, register-transfer, logic, circuit, transistor and process levels
Approach of Syllable Based Unit Selection Text- To-Speech Synthesis System fo...iosrjce
IOSR journal of VLSI and Signal Processing (IOSRJVSP) is a double blind peer reviewed International Journal that publishes articles which contribute new results in all areas of VLSI Design & Signal Processing. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced VLSI Design & Signal Processing concepts and establishing new collaborations in these areas.Design and realization of microelectronic systems using VLSI/ULSI technologies require close collaboration among scientists and engineers in the fields of systems architecture, logic and circuit design, chips and wafer fabrication, packaging, testing and systems applications. Generation of specifications, design and verification must be performed at all abstraction levels, including the system, register-transfer, logic, circuit, transistor and process levels.
Epistemic Interaction - tuning interfaces to provide information for AI supportAlan Dix
Paper presented at SYNERGY workshop at AVI 2024, Genoa, Italy. 3rd June 2024
https://alandix.com/academic/papers/synergy2024-epistemic/
As machine learning integrates deeper into human-computer interactions, the concept of epistemic interaction emerges, aiming to refine these interactions to enhance system adaptability. This approach encourages minor, intentional adjustments in user behaviour to enrich the data available for system learning. This paper introduces epistemic interaction within the context of human-system communication, illustrating how deliberate interaction design can improve system understanding and adaptation. Through concrete examples, we demonstrate the potential of epistemic interaction to significantly advance human-computer interaction by leveraging intuitive human communication strategies to inform system design and functionality, offering a novel pathway for enriching user-system engagements.
State of ICS and IoT Cyber Threat Landscape Report 2024 previewPrayukth K V
The IoT and OT threat landscape report has been prepared by the Threat Research Team at Sectrio using data from Sectrio, cyber threat intelligence farming facilities spread across over 85 cities around the world. In addition, Sectrio also runs AI-based advanced threat and payload engagement facilities that serve as sinks to attract and engage sophisticated threat actors, and newer malware including new variants and latent threats that are at an earlier stage of development.
The latest edition of the OT/ICS and IoT security Threat Landscape Report 2024 also covers:
State of global ICS asset and network exposure
Sectoral targets and attacks as well as the cost of ransom
Global APT activity, AI usage, actor and tactic profiles, and implications
Rise in volumes of AI-powered cyberattacks
Major cyber events in 2024
Malware and malicious payload trends
Cyberattack types and targets
Vulnerability exploit attempts on CVEs
Attacks on counties – USA
Expansion of bot farms – how, where, and why
In-depth analysis of the cyber threat landscape across North America, South America, Europe, APAC, and the Middle East
Why are attacks on smart factories rising?
Cyber risk predictions
Axis of attacks – Europe
Systemic attacks in the Middle East
Download the full report from here:
https://sectrio.com/resources/ot-threat-landscape-reports/sectrio-releases-ot-ics-and-iot-security-threat-landscape-report-2024/
PHP Frameworks: I want to break free (IPC Berlin 2024)Ralf Eggert
In this presentation, we examine the challenges and limitations of relying too heavily on PHP frameworks in web development. We discuss the history of PHP and its frameworks to understand how this dependence has evolved. The focus will be on providing concrete tips and strategies to reduce reliance on these frameworks, based on real-world examples and practical considerations. The goal is to equip developers with the skills and knowledge to create more flexible and future-proof web applications. We'll explore the importance of maintaining autonomy in a rapidly changing tech landscape and how to make informed decisions in PHP development.
This talk is aimed at encouraging a more independent approach to using PHP frameworks, moving towards a more flexible and future-proof approach to PHP development.
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdf91mobiles
91mobiles recently conducted a Smart TV Buyer Insights Survey in which we asked over 3,000 respondents about the TV they own, aspects they look at on a new TV, and their TV buying preferences.
UiPath Test Automation using UiPath Test Suite series, part 3DianaGray10
Welcome to UiPath Test Automation using UiPath Test Suite series part 3. In this session, we will cover desktop automation along with UI automation.
Topics covered:
UI automation Introduction,
UI automation Sample
Desktop automation flow
Pradeep Chinnala, Senior Consultant Automation Developer @WonderBotz and UiPath MVP
Deepak Rai, Automation Practice Lead, Boundaryless Group and UiPath MVP
Essentials of Automations: Optimizing FME Workflows with ParametersSafe Software
Are you looking to streamline your workflows and boost your projects’ efficiency? Do you find yourself searching for ways to add flexibility and control over your FME workflows? If so, you’re in the right place.
Join us for an insightful dive into the world of FME parameters, a critical element in optimizing workflow efficiency. This webinar marks the beginning of our three-part “Essentials of Automation” series. This first webinar is designed to equip you with the knowledge and skills to utilize parameters effectively: enhancing the flexibility, maintainability, and user control of your FME projects.
Here’s what you’ll gain:
- Essentials of FME Parameters: Understand the pivotal role of parameters, including Reader/Writer, Transformer, User, and FME Flow categories. Discover how they are the key to unlocking automation and optimization within your workflows.
- Practical Applications in FME Form: Delve into key user parameter types including choice, connections, and file URLs. Allow users to control how a workflow runs, making your workflows more reusable. Learn to import values and deliver the best user experience for your workflows while enhancing accuracy.
- Optimization Strategies in FME Flow: Explore the creation and strategic deployment of parameters in FME Flow, including the use of deployment and geometry parameters, to maximize workflow efficiency.
- Pro Tips for Success: Gain insights on parameterizing connections and leveraging new features like Conditional Visibility for clarity and simplicity.
We’ll wrap up with a glimpse into future webinars, followed by a Q&A session to address your specific questions surrounding this topic.
Don’t miss this opportunity to elevate your FME expertise and drive your projects to new heights of efficiency.
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Jeffrey Haguewood
Sidekick Solutions uses Bonterra Impact Management (fka Social Solutions Apricot) and automation solutions to integrate data for business workflows.
We believe integration and automation are essential to user experience and the promise of efficient work through technology. Automation is the critical ingredient to realizing that full vision. We develop integration products and services for Bonterra Case Management software to support the deployment of automations for a variety of use cases.
This video focuses on the notifications, alerts, and approval requests using Slack for Bonterra Impact Management. The solutions covered in this webinar can also be deployed for Microsoft Teams.
Interested in deploying notification automations for Bonterra Impact Management? Contact us at sales@sidekicksolutionsllc.com to discuss next steps.
The Art of the Pitch: WordPress Relationships and SalesLaura Byrne
Clients don’t know what they don’t know. What web solutions are right for them? How does WordPress come into the picture? How do you make sure you understand scope and timeline? What do you do if sometime changes?
All these questions and more will be explored as we talk about matching clients’ needs with what your agency offers without pulling teeth or pulling your hair out. Practical tips, and strategies for successful relationship building that leads to closing the deal.
Leading Change strategies and insights for effective change management pdf 1.pdf
G1803013542
1. IOSR Journal of Computer Engineering (IOSR-JCE)
e-ISSN: 2278-0661,p-ISSN: 2278-8727, Volume 18, Issue 3, Ver. I (May-Jun. 2016), PP 35-42
www.iosrjournals.org
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 35 | Page
Development of Text-to-Speech Synthesizer for Pali Language
Suhas Mache1
, C. Namrata Mahender2
1
(Research Scholar, Department Of CS & IT, Dr.Babasaheb Ambedkar Marathwada University Aurangabad,
India)
2
(Asst. Professor, Department Of CS & IT, Dr.Babasaheb Ambedkar Marathwada University Aurangabad,
India)
Abstract : We introduced a new method for Text-To-Speech (TTS) synthesis for Pali language. We discuss the
efforts in collecting speech database of Pali language and relevant design issues in development of TTS system.
This system is based on unit selection concatenative speech synthesis using phonemes, syllables and words as an
elementary unit for Pali speech synthesis. The speech units picked by the selection algorithm optimally. An
important advantage of this approach leads to reduced prosody mismatch and spectral discontinuity that occurs
syllable concatenation. The results obtained from the proposed system are far superior as compared to the
traditional unit based Text-To-Speech system. The most important output of this system is the improved
naturalness and intelligible synthesized speech.
Keywords : Text-To-Speech synthesis, Pali Speech Database, Unit selection, Speech Analysis
I. Introduction
Generating a human like sound by machine is the dream of many scientists from last centuries.
Synthesizing human speech is difficult due to the complexity of human speech. The production of human speech
involves the lungs, vocal fold and vocal tract (oral cavity and nasal cavity) functioning collectively [1] [2]. Text-
to-Speech (TTS) is the process of converting unknown text into sound. It is an artificial production of human
speech. A computer system used for this purpose is called a speech synthesizer and that is implemented in both
software and hardware [3]. The Text-to-Speech synthesis procedure consists of two main phases. The first one is
text analysis, where the input text is transcribed into a phonetic or some other linguistic representation and the
second one is the generation of speech waveforms [4].
Fig.1 Block diagram of Text-to-Speech System
Techniques of speech synthesis [5].
a) Articulatory Synthesis
b) Formant Synthesis
c) Concatenative Synthesis
Unit selection based concatenative speech synthesis system has become popular now a day because of their
highly natural sounding synthetic speech [6].
To develop a Text-to-Speech system for Pali language, we have chosen a unit selection based
concatenative speech synthesis. The development process of Pali language TTS system consist of simplified
phone set of Pali language. That includes vowels, consonant and syllables. The inputted text i.e. character string
is then preprocessed and analyzed [4]. The speech signal is generated by concatenating prerecorded sound units.
In this approach several fundamental periods of pre-recorded phonemes are simply concatenated. The phonemes
are then connected to form words and sentences [7]. The detailed process is explained in section 3.
Input Text
Text Analysis
Text Normalization
Linguistic Analysis
Phonetic Analysis
Grapheme -to-
Phoneme Conversion
Concatenation of Speech
Units &
Waveform-Generation
Speech
Database
Synthesized
Speech
2. Development Of Text-To-Speech Synthesizer For Pali Language
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 36 | Page
II. Previous work
In India 15 official languages are spoken in different forms across different places. From the point of
view of TTS development, the most helpful aspect of Indian scripts is that they are basically phonetic in nature,
and there is one-to-one correspondence between the written and spoken forms of most of the Indian languages.
This makes the task of automatic phonetization simpler [8]. In our study we explore many speech synthesis
methods, techniques, applications and products as possible in our investigation. The world’s first mechanical
speech synthesizer machine was developed by Gerbert [9] after that a continuous development was done. In
present scenario sophisticated speech synthesizer is available in English language as compared in Indian
languages where still a lot is to be done. The available work for Indian languages is discussed bellow.
Table No.1: Text to Speech systems in Indian languages.
Institute /
Organization
Covered
Languages
Synthesis
techniques
Database Performance
JNU Delhi Sanskrit Rule based Database is designed in SQL server,
phones and 2,00000 words
Unlimited TTS -
more than average
C-DAC
Mumbai
Marathi, Odia Unit selection
Concatenative
Festival based speech
synthesis
Units of phones, syllables and words Unlimited TTS -
more than average
IIIT
Hyderabad
Bengali, Hindi,
Kannada, Tamil,
Malayalam,
Marathi, Sanskrit
Letter to Sound
rule
(grapheme-to
phoneme or Akshara-
to-sound)
(1000 sentences of each language)
ph sy wd
47 866 2285
58 890 2145
51 851 2125
48 1191 2077
57 660 2097
35 930 2182
51 997 2310
ph – Phonemes,
sy – Syllables,
wd - Words
Unlimited TTS -
more than average
TIFR
Mumbai
Marathi, Hindi Formant synthesis Phones and words Average
HCU
Hyderabad
Telugu Concatenative
MBROLA based
Diphone Average
CEERI,
Delhi
Hindi Bengali
(partly)
Formant Synthesis
(Klatt – type)
Syllables & Phonemes (Parameter
Data Base)
Excellent
Unlimited TTS-
Average
IIT, Chennai
(Madras)
Hindi, Tamil Concatenative
diphone synthesis
(1400 diphones)
Diphone Syllable (Mainly) Unlimited TTS-
Average
C-DAC Pune Hindi, Indian
English
Concatenative Phonemes and recorded word Limited TTS -
more than average
C-DAC
Noida
Hindi Concatenative Multi form units Diphones, Syllable,
frequent words and phrases.
Domain specific,
Excellent unlimited
TTS Average
C-DAC
Kolkata
Bangla Concatenative Phonemes and new set of signal
units
Unlimited TTS
Average
Utkal
University,
Bhubaneshwar
Oriya Concatenative Phones, Diphones & Triphones Unlimited TTS
Average
IISc
Banglore
(Dhvani)
Bengali,
Gujrati, Hinhi,
Kannada,
Malayalam,
Marathi, Oriya,
Punjabi, Tamil,
Telugu.
Concatenative Phomes (consonant, vowels & half
consonant)
Unlimited TTS -
more than average
TTS system in Indian Languages [9]-[23].
III. Methodology
3.1 Nature of the Pali Scripts
The name pali means lines or canonical text [24], basic units of writing system in Pali are characters
which are an orthographic representation of speech sounds. Pali has been written in variety of scripts including
Bramhi, Devanagari and other Indic scripts. Today Pali is studied mainly by those who wish to read original
Buddhist scriptures. There are non-religious texts in Pali including historical and medical texts. A typical
structure of Pali language script is in syllabic in nature the form of an Akshara are V, CV, CVC and CCCV
3. Development Of Text-To-Speech Synthesizer For Pali Language
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 37 | Page
where C is consonant and V is vowel [25]. There is good correspondence between what is written and what is
spoken.
3.2 Phone Set
An Akshara is an orthographic representation of a speech sound in Pali language. Pali language in
Devnagari script is represented by 8 vowels and 32 consonants making a total of 40 alphabets as shown in fig. 1.
Vowels
अ आ इ ई उ ऊ ए ओ
Consonants
क ख ग घ ड. च छ
ज झ त्र ट ठ ड ढ
ण त थ द ध न ऩ
प फ ब भ म य र
ल स ह ऱ
Fig. No.2 Phone set for Pali language in Devnagari script [27].
3.3 Development of Speech Database
The purpose of building Pali speech synthesizer is to generate high quality synthetic speech. The
database has been collected in order to investigate how well the Pali TTS system is used by humans in
intelligible verbal communication [28]. In recent years speech synthesis technology has progressed remarkably
because of large scale speech corpus. .
3.3.1 Text Corpus
It is important to have an optimal text corpus balanced in terms of phonetic coverage. The text corpus
consist of phone set vowels and consonants, commonly used in daily words, body parts, days, months, names of
worker role, animals, birds, relationships, colors, sessions, budha historical places, names of historical
educational institutes, names of bhikhu-bhikhuni, budha grantha, short stories Tipitak and paragraphs and
sentences from budha’s ascent literature. The text corpus consists of total 676 sentences, 12359 total words and
3226 unique words. In these sentences we tried to cover the possible conversions related to our daily life.
3.3.2 Speech Recording
Data is collected from professional speakers, with very good voice quality. The quality of digital sound
is determined by discrete parameters. The discrete parameters are the sample rate, bit capacity and number of
channels [29]. One important conclusion that we can make at this point is that our digitization standards should
be able to faithfully represent acoustic signals. The Linguistic Data Consortium for Indian Languages (LDC-IL)
under Central Institute of Indian Languages has designed the standards for capturing the speech data according
to application for which the speech data is collected and the devices that were used for recording the speech
samples [30]. To record the speech samples, a process of speaker selection was carried out. A professional male
speaker (Pali language expert) in the age group of 20-25 with very good voice quality was selected. The
recording was done in noise free environment. The speech signals was recorded by using a standard headset
microphone connected to the laptop. A 16 KHz sampling rate is often used for microphone speech. We have
used PRAAT software tool to record the speech at 16 KHz sampling frequency and represented using 16
bits/sample [30].
3.3.3 Speech Processing and Labeling
The most important tasks in building speech database are the speech data segmentation and labeling.
The quality of synthesized speech depends on the accuracy of the labeling process [31]. Manual labeling of
speech data is still the most common form of labeling means to temporally define discrete names to them using
symbols from an appropriately defined set corresponding to acoustic, physiological, phonetic or higher level
linguistics terms [32]. In this work manual speech segmentation and labeling are very tedious and time
consuming task and require much efforts. We record all the necessary speech data and then some processing are
done to remove the noise of the recorded speech. Then normalize the speech by using audacity software tool.
The processed speech was segmented using PRAAT tool and labeled the speech files as “k.wav”.
4. Development Of Text-To-Speech Synthesizer For Pali Language
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 38 | Page
3.3.4 Unit Selection and Speech Generation
One approach to the generation of natural sounding synthesized speech is to select and concatenate
recorded speech units from a large speech database [33][34]. The task of unit selection procedure is to find the
most appropriate recorded speech units in the corpus. Input data received from language processing modules in
the speech synthesizer are sequence of phonemes to be pronounced, whereby prosodic parameters for
pronunciation of each phoneme are provided. These parameters contain data on the fundamental frequency and
duration of the phoneme pronunciation [35]. We propose that the units in a synthesis database can be considered
as a state transition network. For a given textual input which is mapped into present database, the unit selection
algorithm concatenates the corresponding wave files sequentially from left to right. The developed algorithms
detects corresponding audio file present in the speech database and concatenates them.
The algorithm first checks weather the user has entered a word or a sentence by detecting the number
of spaces present in the entered text. The inputted text is split into words, syllables and phonemes. A search is
then made to find a best i.e. first priority of search for word into database. If exact word is found in database,
then it fetches the corresponding wave file in the database. If not then split the word into syllables. Search
syllables into database, if syllables are not found in database split into phonemes and select corresponding wave
files stringing together (concatenating) all wave files and generate the speech.
Following figure shows searching of the optimal sequence of recorded speech units for target Pali word
ऩालरबासा
पा लि भा सा
Fig.3 Searching of Speech Units
IV. Results and Discussion
The result of this work is evaluated to check the naturalness of the synthesized speech. The GUI shown
in Fig. No. 4 where speech is synthesized by exact or complete words and words formed by concatenating
syllables and phones
Fig. 4 Snapshot of Pali Text To Speech Synthesizer GUI
Unit
- 1
Unit
- 2
Unit
- 3
Unit
- 4
1
3
4
6
1
7
2
5
1
1
9
1
0
8
Concate
nated
Unit
Speech
Database
5. Development Of Text-To-Speech Synthesizer For Pali Language
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 39 | Page
The speech waveform and spectrogram plot for the Pali phrase सब्फ ऩाऩस्स अकायण is shown in following
figure.
Fig. 5 Spectrogram and wave form of synthesized speech
Fig.6 Waveform, Spectrum and Cepstrum
The spectrogram shows that the formant changes are not abrupt at the concatenation points. The
average energy of synthetic speech was calculated by short time speech measurement. This measurement can
distinguish between voiced and unvoiced speech segments, since unvoiced speech has significantly smaller
short time energy. The energy of the frames is calculated using the relation
Fig. No.7 Short time energy of synthetic speech
Moreover, spectral changes are uniform across the syllable boundaries and hence reinforce the idea that
the syllable like unit is indeed a good candidate for concatenative speech synthesis.
6. Development Of Text-To-Speech Synthesizer For Pali Language
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 40 | Page
The results of Pali language vowels, consonants and words
Table No.2 Statistical results of vowels
Sr.
No.
Pali
Vowels
Pitch of
Uttr_1
Pitch of
Uttr_2
Pitch of
Uttr_3
Pitch of
Uttr_4
Pitch of
Uttr_5
Mean SD
1 अ 163.187 158.791 156.19 151.48 153.15 156.56 4.65
2 आ 148.307 158.872 151.669 150.75 149.5 151.82 4.14
3 इ 155.838 174.455 153.526 152.16 150.69 157.33 9.76
4 ई 148.4 162.934 150.967 151 155.28 153.72 5.71
5 उ 158.299 158.193 151.534 151.29 149.2 153.70 4.24
6 ऊ 153.286 163.536 161.836 153.9 152.81 157.07 5.17
7 ए 151.477 162.311 148.062 155.38 149.37 153.32 5.74
8 ओ 138.717 146.605 140.412 150.83 130.45 141.40 7.81
Table No.2 Statistical results of consonants
Sr.
No.
Pali
Consonants
Pitch of
Uttr_1
Pitch of
Uttr_2
Pitch of
Uttr_3
Pitch of
Uttr_4
Pitch of
Uttr_5
Mean SD
1 क 155.73 158.36 154.467 160.54 153.46 156.51 2.90
2 ख 152.2 153.48 146.706 164.05 151.75 153.64 6.37
3 ग 149.22 156.33 150.324 150.32 153.74 151.99 2.96
4 घ 143.61 146.45 143.093 157.79 148.07 147.80 5.94
5 च 157.67 159.19 147.87 149.66 155.49 153.97 4.97
6 छ 150.94 157.8 156.442 151.87 148.63 153.14 3.86
7 ज 151.52 154.22 153.343 153.5 155.91 153.70 1.59
8 झ 143.77 153.59 152.89 148.35 146.79 149.08 4.15
9 ट 146.22 162.16 154.159 154.35 157.81 154.94 5.86
10 ठ 135.69 143.47 137.912 139.2 138.02 138.86 2.87
Table No.3 Statistical results of words
Sr.
No.
Word Pitch of
Uttr_1
Pitch of
Uttr_2
Pitch
of
Uttr_3
Pitch of
Uttr_4
Pitch of
Uttr_5
Mean SD Variance
1 लानयीदो 150.368 149.133 147.829 147.126 148.142 148.52 1.26 2.52
2 भक्कक्कट 159.577 157.515 156.735 153.542 150.918 155.66 3.42 6.85
3 भमुयो 150.153 147.311 149.171 143.63 150.067 148.07 2.73 5.46
4 गरुड 134.67 136.701 144.365 139.919 134.041 137.94 4.26 8.52
5 सुलो 157.743 157.122 155.488 158.264 148.427 155.41 4.04 8.08
6 सारयका 153.045 151.671 152.847 150.664 151.761 152 0.97 1.94
7 कोककरो 162.627 150.858 157.704 156.924 149.426 155.51 5.39 10.78
8 ककपऩरो 145.234 165.986 167.629 161.007 155.162 148.50 9.1 18.21
9 काक 150.466 149.372 151.96 149.203 147.434 154.89 1.67 11.29
10 कु कु टो 165.191 159.453 161.212 161.175 158.799 161.17 2.49 4.98
The overall performance of the system is computed by calculating the percentage of correct phonemes (i.e.
consonants and vowels), 1 to 100 digits, short words and connected words calculated on 100 random unknown
words (i.e. words that are not in the database except connected words) of Pali language.
Table No.4 Test Results
Sr. No. Type of Data Accuracy (%)
1 Vowels 100 %
2 Consonants 100 %
3 Syllables 100 %
4 Digits (1 – 100) 100 %
5 Short words 71 %
6 Connected words 42 %
7. Development Of Text-To-Speech Synthesizer For Pali Language
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 41 | Page
V. Conclusion
The developed Text-To-Speech system we observed that with initial experiments showing that word
unit performs better than the phoneme and syllable units. In this experiment speech units picked by the selection
algorithm optimally, to produce a natural sounding synthetic speech. This speech synthesizer is capable of
generating natural sounding synthesized speech with no prosodic modeling.
It was observed that when the database of small speech units, the speech synthesizer is likely to
produce a low quality speech. As the database of units increases, it increases the quality of the synthesizer.
Creation of maximum coverage of units for concatenation synthesis gives greatest naturalness. This system can
generate natural and intelligible synthesized speech for Pali language. Overall these test in table number 4 shows
that the accuracy of the TTS system is 85 %. Here we conclude that the Text to Speech conversion accuracy is
good.
Future scope- The current system is not meant to read words those are formed by combination of Half
Symbol + Full Symbol (i.e. Jodakshar) e.g. फुध्द, याठ्ह, फरीलद्द. Thus in future development of this system such
words will allow be included to make full-fledged system for Pali language.
References
[1]. Archana Balyan, S.S. Agrwal and Amita Dev, Speech Synthesis: Review, International Journal of Engineering Research and
Technology, ISSN 2278-0181 Vol. (2), 2013, 57 – 75.
[2]. Jonathan Williams, Prosody in Text-to-Speech Synthesis Using Fuzzy Logic, M.S. diss., West Virginia University, USA, 2007.
[3]. D.D. Pande, M. Praveen Kumar, A Smart Device for People with Disabilities using ARM7, International Journal of Engineering
Research and Technology, ISSN 2278-0181 Vol.(3), 2014, 614 – 618.
[4]. Sami Lemmetty, Review of Speech Synthesis Technology, M.S. diss., Helsinki University of Technology, Finland, 1999.
[5]. Mark Tatham and Katherine Morton, Developments in Speech Synthesis (John Wiley & Sons, Ltd. ISBN: 0-470-85538-X, 2005)
[6]. Jithendra Vepa and Simon King, Subjective Evaluation of Join Cost and Smoothing Methods for Unit Selection Speech Synthesis,
IEEE Transactions on Audio, Speech and Language Processing, Vol. (14), 2006, 1763 – 1771.
[7]. Baiju Mahananda, C.M.S. Raju, Prahallad Kishore et. al., Building a Prototype Text-to-Speech for Sanskrit, Proc. 4th
International
Sanskrit Computational Linguistic Symposium, New Delhi, 2010, 39-47.
[8]. Krishna N Sridhar, Murthy Hema A, Gonsalves Timothy A., Text-to-speech in Indian languages, Proc. International Conference on
Natural Language Processing, ICON-2002, Mumbai, 2002, 317-26.
[9]. Mattingly, I.G., Speech Synthesis for Phonetic and Phonological Models, T.A. Sebeok (Ed.) Current Trends in Linguistics,
Linguistics and Adjacent Arts and Sciences, Vol.(12), 1974, 2451-2487.
[10]. Computational Linguistics R&D Special Centre for Sanskrit Studies JNU, Delhi, Online available
http://sanskrit.jnu.ac.in/samvacaka/index.jsp (2016).
[11]. Girish N. Jha, Sudhir K. Mishra, R. Chandrashekhar et. al., Developing a Sanskrit Analysis System for Machine Translation, Proc.
National Seminar on Translation Today: state and issues, Dept. of Linguistics, University of Kerala, Trivandrum, 2005, 23-25.
[12]. Development of TEXT to SPEECH SYSTEM in Indian Languages Phase-II, Online available
http://cdac.in/index.aspx?id=mcst_speech_tchnology, 2016.
[13]. Kishore Prahallad, E. Naresh Kumar, Venkatesh Keri, S. Rajendran and Alan W Black, The IIIT-H Indic Speech Databases, Proc.
of Interspeech Portland, Oregon, USA, 2012.
[14]. A. Sen and K. Samudravijaya, Indian Accent Text to Speech System for Web Browsing, International Journal Sadhana, Vol. (27),
February 2002, 113-126.
[15]. Spoken Language Processing, Online available http:// http://speech.tifr.res.in 2016.
[16]. S.S. Agrwal, Rajesh Verma, Shailendra Nigam, Anuradha Sengar, Development of Rules for Unlimited Text To Specch Synthesis
of Hindi, Proc. Conf. on ICA, Rome, 2001, 2-4.
[17]. Speech and Vision Laboratory IIT Madras, Online available http://www.cse.iitm.ac.in/speech, 2016.
[18]. Ramani B, S.L. Christina, G.A. Rachel, S. Solomi V, M.K. Nandwana, A. Prakash, A Common Attribute based Unified HTS
framework for Speech Synthesis in Indian Languages, Proc. 8th ISCA Conf. on Speech Synthesis Workshop, Barcelona, Spain,
2013, 291-296.
[19]. Hindi Text-to-Speech (TTS) System C-DAC, Online available http://cdac.in/index.aspx?id=mlc_gist_speechtech, 2016.
[20]. ESNOLA based Bangla TTS C-DAC, Online available http://cdac.in/index.aspx?id=mc_st_TTS_Bangla, 2016.
[21]. Sanghamitra Mohanti, Syllable Based Indian Language Text To Speech System, International Journal of Advances in Engineering
& Technology ISSN: 2231-1963, May 2011, 138-143.
[22]. Preranasri Mali, A Survey on Text To Speech Translation of Multi Language, International Journal of Research In Advanced
Engineering Technologies ISSN: 2347-2812, 2014, 29-31.
[23]. Dhvani a Text to Speech System for Indian Languages, Online available http://dhvani.sourceforge.net, 2016.
[24]. Omniglot the Online Encyclopedia of writing Systems and Languages, Online available http://www.omniglot.com/writing/pali.htm,
2016.
[25]. Baiju Mahananda, C.M.S. Raju, Ramalinga Reddy Patil, Narayana Jha, et. al., Building a Prototype Text to Speech for Sanskrit,
Proc. 4th
International Conf. on Sanskrit Computational Linguistics Symposium, New Delhi, India, 2010, 39 – 47.
[26]. Branko Markovic, Slobodant Jovicic, Jovan Galic, Dorde Grozdic, Whispered Speech Database: Design, Processing and
Application, Proc. 16th
International Conf. on Text, Speech and Dialogue, Springer ISBN 978-3-642-40585-3-74, 2013, 591 – 598.
[27]. Susila Muljadhav, Pali Bhasha Parichya ऩारी बाषा ऩरयचम, (Ramai Publication, 2015.)
[28]. Hiromichi Kawanami, Tsuyoshi Masuda, Tomoki Toda and Kiyohiro Shikano, Designing Speech Database with Prosodic Variety
for Expressive TTS System, Proc. IREC, Japan, 2002, 2039-2042.
[29]. Primer on Audio PC various issues associated with PC based audio technology, Online available on
http://www.totalrecorder.com/primerpc.htm, 2016.
[30]. Standards for Speech Data Capturing and Annotation, available online http://www.ldcil.org/download/samplingstandards.pdf, 2016.
[31]. Marian Boldea, Cosmin Munteanu, Labeling A Romanian Speech Database, Proc. Second International workshop on Speech and
Computer, Napoca Romania 1997.
8. Development Of Text-To-Speech Synthesizer For Pali Language
DOI: 10.9790/0661-1803013542 www.iosrjournals.org 42 | Page
[32]. Barry, W.J. Fourcin, A.J., Levels of Labeling Computer Speech and Language, Proc. Abstract Computer Science, Australian
National University, Australia 1992.
[33]. Hunt A.J.,Unit selection in a concatenative speech synthesis system using a large speech database, Proc. International Conference
on ICASSP-96, Acoustic, Speech and Signal Processing IEEE Vol.(1), 1996, 373-376.
[34]. Andrew J. Hunt and Alan W. Black, Unit Selection in a Concatenative Speech Synthesis System Using a Large Speech Database,
IEEE Transactions on Audio, Speech and Language Processing, Vol. (3), 1996, 373-376.
[35]. Jerneja Gros and Mario Zganec, An Efficient Unit-selection Method for Concatenative Text-to-speech Synthesis Systems,
International Journal of Computing and Information Technology, 2008, 69-78.