Multilingual Computing at C-DAC Mumbai

The Center for Development of Advanced Computing (CDAC) Mumbai is actively involved in Research and Development into the areas of Language computing and has produced many products targeted to bridge the language barrier in the era of digitization.

Knowledge Based Computer Systems (KBCS) at CDAC Mumbai is working in the field Natural Language Processing (NLP), Machine learning (ML), Data Mining (DM), Expert systems (ES), etc. and following are some of the prominent works in the area of multilingual computing

1. MaTra
2. Xlit
3. StatMT
4. SuTra
5. Rupantar
6. ChitranTran

MaTra:

MaTra is a Fully-Automatic English to Hindi Indicative Machine Translation System. The approach taken by MaTra is 'Transfer Based' and is very well appreciated in the research community. MaTra is targeted to work with text in open domains like World Wide Web documents and news stories.

System uri : http://cdacmumbai.in/matra/

Salient Features:

• Hybrid (Rule Based and Statistical) approach to Machine Translation
• Uses target language independent intermediate structured representation and so can be easily adopted for English to other Indian languages machine translation systems

Figure 1:MaTra: Machine Translation System

Xlit:

Xlit is a transliteration tool to convert words from English to Indian languages and back, without losing the phonetic characteristics. It transliterates the words from English to Indian language, eg. converts 'bharat' to 'भारत', 'school' to 'स्कूल', etc. It also suggests more than one option for the given word, like भरत, भारात, बहारत, etc. for 'bharat'. Prototypes are available for Hindi, Marathi, Urdu and Kannada. XlitHindi – an extension for the OpenOffice Writer is available for download at the URL:

http://extensions.services.openoffice.org/project/xlithindi

System uri: http://cdacmumbai.in/xlit/editor/

Salient Features:

• Can be easily integrated into any desktop or web application
• Uses generalized framework for developing any language pair transliteration system

Figure 2 Xlit: A transliteration system

StatMT:

StatMT is a Statistical Machine Translation (SMT) system which translates the source language sentences (e.g., English) to target language sentences (e.g. Hindi, Marathi, Bengali, etc.) using statistical models. Stat MT system is part of English to Indian Languages Machine Translation System (E-ILMT) consortium and aims to design and deploy a Machine Translation System from English to Indian Languages in Tourism and Healthcare Domains. More information about the system can be found at http://www.cdacmumbai.in/e-ilmt.

SuTra:

Sutra is a multi-user translation assistance tool that makes intelligent suggestions to translators on possible reuse of translations from older version systems or systems with similar domains. The aim is to reduce the translator's efforts and make available translated versions of applications in least possible time. System is released under open source and is available at http://sourceforge.net/projects/sutra/.

Figure 3SuTra: An Intelligent Suggestive Translator for Localisation

Rupantar:

Rupantar is an utility to write in Indian languages using Roman Script. It also allows you to convert text from one script to other script, ex. 'रमेश' in Hindi to ‘ரமேஷ்’ in Tamil. It uses a key map based technique for writing and conversion.

System uri: http://www.cdacmumbai.in/rupantar

Salient Features:

• Easy integration with other desktop and web applications
• Fast and lightweight application

Figure 4: Rupantar

ChitranTran:

ChitranTran is an utility to extract and transliterate text from images. Available prototype can extract text in English and Hindi and can tranliterate that to other Indian languages.

System uri: http://202.141.152.1/xlit/chitrantran/

Salient Features:

• Easy integration with other desktop and web applications
• Support for Indian language text extraction and transliteration

Figure 5 ChitranTran

Abdul Kalam Pitches A Multilingual, Mobile Web at WWW conference 2011

APJ Abdul Kalam, Former President of India, speaking at the World Wide Web conference in Hyderabad made a pitch for a multilingual web, saying that in its current form, the World Wide Web has its shortcomings – “The language barrier is the biggest hinderance in making the Web truly democratic. Originally the lingua franca of the web was mainly English, and while the situation has started to change, much more needs to be done. The development of a country is directly determined by the amount of content in the countrys native language available on the web.” More interestingly, Kalam suggested cross-lingual access to the web, saying that knowledge grows by sharing, and language should not be an impediment here. He said that rural folk need to be convinced that the web is useful for them, and at present, the community on the web tends to generate content for its own consumption.

Some of the important points from his speech:
Kalam had the following suggestions for the development of the World Wide Web
- To look for solutions on how a mobile device can provide integrated solutions of 3G and 4G applications in its mother tongue. For a farmer, the price of agricultural products, for a fisherman, the market price of fish.
- For Web 2.0 and 3.0 (the semantic web) to enhance services in native languages and the web to offer access without any barriers of language, cost, creed or geographical barriers.
- For the mobile to become a personal authentication device, and for money transactions through the mobile to be highly secure.
- For sensors incorporated into a mobile device holder, to be able to transmit data related to a patient, and get the doctors advice / consultancy
- More societal applications, given large bandwidth that 4g offers, which involve farmers and villagers who are less empowered. The future of the web is going to shift from connecting the corporates to connecting the individual in the rural society.
Kalam also spoke of a societal grid, combining the National Knowledge Network, a Healthcare Grid, an e-governance grid and a Rural grid.

Courtesy and more detail news here at Medianama

SIGAI Workshop on Emerging Research Trends in Artificial Intelligence (ERTAI - 2011)

SIGAI Workshop on Emerging Research Trends in Artificial Intelligence (ERTAI - 2011) 19th - 21st June, 2011, C-DAC, E-City Bengaluru, India Supported by Computer Society of India (CSI)

About The Workshop

Based on the success of the previous ERTAI workshop, conducted in 2010, CSI- SIGAI has decided to announce the next ERTAI workshop to be held during June 19-21 2011 at C-DAC Electronics City Bengaluru. The backdrop of this workshop remains the same as that of last year. That is, through ERTAI, we plan to continue to provide a forum where those pursuing research in AI can exchange ideas and seek guidance. And those who are seeking to enter into AI research can also get a valuable feel of current research going on in various streams in AI in industry as well as in academia. ERTAI 2011 will enable new and aspiring research scholars to identify relevant and useful research topics and get guidance on their approach and direction.

We invite papers that describe work-in-progress by various research scholars spanning many areas including language processing, multi-agent systems, web mining, information retrieval, semantic web, e-learning, optimization problems, pattern recognition, etc. We also invite suggestions on relevant topics for invited talks -- both in technical areas and and research methology areas. The detailed program for the workshop is being finalized and will be announced shortly.

Proposed Structure of Workshop

It will be a three day programme consisting of,

• Invited talks covering current trends, specific challenges, etc. in Artificial Intelligence
• Invited talks on mentoring research scholars on publication, research methodology, etc.
• Presentations by those currently pursuing research in AI area.

We will have a panel of experienced researchers to evaluate and mentor the research presentations.

Call For Papers

For the research presentations, we are now inviting brief research papers of 5-6 pages, outlining the problem being addressed, approach followed vis-à-vis existing approaches, current status / results, and future plans. A subset will be short-listed for presentation, based on a formal review process. Papers must have significant AI content to be considered for presentation. Relevant topics include (but are not limited to):

	Knowledge Representation		Reasoning
	Model-Based Learning		Expert Systems
	Data Mining		State Space Search
	Cognitive Systems		Vision & Perception
	Intelligent User Interfaces		Reactive AI
	Ambient Intelligence		Artificial Life
	Evolutionary Computing		Fuzzy Systems
	Uncertainty in AI		Machine Learning
	Constraint Satisfaction		Ontologies
	Natural Language Processing		Pattern Recognition
	Intelligent Agents		Soft Computing
	Planning & Scheduling		Neural Networks
	Case-Based Reasoning

A one-page call for papers for the ERTAI - 2011 workshop may be obtained from here

Target Audience

Target audience will be primarily:

• Faculty members pursuing research involving AI as the base or as a tool for an application.
• Faculty members interested in pursuing research and exploring areas / options.
• Research scholars working for a post graduate degree.
• Students seriously interested in research, specifically on AI.

Important Dates

• Full paper submission deadline - April 30, 2011
• Acceptance intimation - May 25, 2011
• Camera ready copy due -  June 05, 2011

ERTAI Secretariat

ERTAI Secretariat
Centre for Development of Advanced Computing
68, Electronics City, Bengaluru 560100.India.
Telephone: +91 80 28523300
Fax: +91 80 28522590
Email: csi.sigai@gmail.com This e-mail address is being protected from spambots. You need JavaScript enabled to view it
Web: http://sigai.cdacmumbai.in/

Advanced Statistics and Data Mining - Summer School

A Summer School on Advanced Statistics and Data Mining is being organized by the Artificial Intelligence Department of the Computer Science Faculty of the Univ. Politécnica de Madrid. from June 28, 2010 to July 09,2010 in Madrid, Spain. More information can be found at:

http://www.dia.fi.upm.es/index.php?page=presentation&hl=es_ES

ACL 2010

The 48th Annual Meeting of the Association for Computational Linguistics will be held in Uppsala, Sweden, July 11–16, 2010. The conference will be organized by the Department of Linguistics and Philology at Uppsala University.
More details can be found at : http://acl2010.org/index.htm

Software tools to assist you use your language

Writing in ones mother tongue is something every one likes; as the best way of expressing your thoughts is in the language you have been brought up in. Here are some tools to enable you do this.

• Xlit - CDAC Mumbai Transliteration Service
• Xlit - Transliteration Editor
• Rupantar - utility to write in Indian Languages using Rupantar scheme
• English Morphological Analysis
• Shabdanjali - English to Hindi & vice versa bilingual dictionary
• MaTra

 Please, feel free to suggest/comment/query.

One Day Workshop on Emerging Trends on Artificial Intelligence

CDAC Mumbai is organizing a Workshop on AI, with support from CSI (Computer Society of India). Research inclined students, teachers and professionals are encouraged to participate in this event. You may expect to learn more details about doing research, get to meet some high profile people involved in AI research in India, also the recent advances, challenges in AI, among many other things.
Please feel free to pass this post to your friends also. Looking forwawrd to meet you there.
For any queries please feel free to write to us. Below are few more details about the participation process:

ERTAI-2010: Call For Participation

The Special Interest Group on Artificial Intelligence (SIGAI) is organizing a Workshop on Emerging Research Trends in Artificial Intelligence (ERTAI - 2010) on 17th April, 2010 at C-DAC Navi Mumbai, supported by the Computer Society of India (CSI), with an objective to bring into focus high quality research being conducted in the field of Artificial Intelligence in India, and also, to provide a common platform where those pursuing research can exchange ideas and seek guidance.

For more details on the workshop, you may refer the call for participation and register for the workshop. The registration details for the workshop are available here.

Rupantar: OpenOffice Writer extension installation and uses

Please reply to this post if you have any issues regarding installation of the Rupantar OpenOffice extention .
Here are some problems you may face while installing the Rupantar in OpenOfice.You may come across these while installing other extensions also. These are not specific to Rupantar.

1] For error:
(com.sun.star.registry.CannotRegisterImplementationException) { { Message = "", Context = (com.sun.star.uno.XInterface) @0 } } } }

>>>> Please install 'openoffice.org-java-common' package from Synaptic manager [specific to Linux machines]

2] For error: Unable to load java while installing the extension

>>>> Please check that your machine has java 6 or above, installed. If you have it and you are still facing the same problem, now you need to mention path of the JRE to be used, in OpenOffice. Please do it this way:

* Go to Openoffice Writer: Tools --->Options --->OpenOffice.org ----->java.
* Now check 'use a Java runtime environment' and if you see one/more JREs there, select it/one of them.
* click 'OK'. You may try installing the extension now.

If you don't see any JRE in the window, click 'Add' and select the JRE you want to use from the file browser window.


3] If your OOo starts not responding while you are trying the extension, please check your installed JAVA. To run the extension properly you need "Sun Java" and not the "Free software foundation Java". Check this in Tools --->Options --->OpenOffice.org ----->java; and make sure that, the Java being used is Sun Java.

Please feel free to suggest/comment/query!

Prakash B. Pimpale
Xlit Team

Open Source Seminar Series: NLTK and GNU build system (20th March 2010)

What:
CDAC Mumbai, Kharghar centre, has been conducting a series of lecture demos to introduce you to the wide range of high quality software available in open source. This months seminars include:

1] NLTK - NLTK [Natural Language Processing Toolkit] is a complete Open Source suite of Python modules, linguistic data and documentation for research and development in Natural Language Processing. It provides utilities for many NLP tasks like chunking, stemming and parsing to name a few. It is a useful tool for students, professionals and researchers interested in Natural Language Processing and is widely used by NLP community all over the world.

and

2] GNU Build System - The GNU build system provides an environment to a computer programmer that helps them make their software/packages portable across platforms (at least for Unix-like operating systems). It also makes the build process easier for the user, allowing him/her to use just a small set of commands to build the software/package from its source code and install it.


When: Saturday,  20th March 2010.  09:30 am to 12:30 pm

Where: CDAC, NEAR BHARTI VIDYAPEETH, KHARGHAR,  NAVI MUMBAI.


Last date to register: March 17th, 2010


Registration fee: Rs. 200/- per participant. To register, please mail to the address below, giving your name, institution, contact address (including email and phone), and a local cheque/ DD for the fee.

Contact:

Course Administration, CDAC (formerly NCST),
Raintree Marg, Near Bharati Vidyapeeth Engg. College,
Opp. Kharghar Railway Station,
Sector 7, CBD Belapur, Navi Mumbai 400614, India.
(5 minutes walk from Kharghar Railway station.)

Telephone: +91-22-27565303 Telefax: +91-22-27560004

Website: http://ossrc.org.in  Email: course_kh@cdacmumbai.in

Contact us

C-DAC, Mumbai (Formerly NCST)
Raintree Marg, Near Bharati Vidyapeeth,
Opp. Kharghar Railway Station,
Sector 7, CBD Belapur,
Navi Mumbai 400614, India.
Telephone: +91-22-27565303 Telefax: +91-22-27560004
Website: www.cdacmumbai.in
email us: kbcs@cdacmumbai.in

Xlit: CDAC Transliteration API

Welcome to the Xlit Ajax API usage-discussion blogpost !!

Note : Please note that usage of this free API for commercial applications is 'STRICTLY PROHIBITED'. Also, in case of interested in using this API,
please write to us so that your webpage can be linked from our repository.

Here, you can discuss any suggestions and/or issues involved with using the API for transliteration or give your feedback.

Xlit : Ajax API Usage Documentation

One of the easier ways to start learning about Xlit Ajax API is using the command line tool 'wget'; execute the following command:

wget --post-data 'srclang=en&trgtlang=hi&inpword=Hello&inputval=5&ui=xml' "http://www.cdacmumbai.in/xlit/editor/getXlatn"

Understanding the command:

This command performs a language transliteration (/xlit/editor/getXlatn) for hello (inpword=hello) from English to an Indian Language (here Hindi : srclang=en&trgtlang=hi). The other language options are listed elsewhere in this document. The response has a Content-Type of 'text/';charset=utf-8. The 'inputval' determines the # of possible alternatives of the given input word in the target language (limited to 5 as default higher limit).

The response returned to this request takes the format as specified with 'ui' in the query. The possible values that ui can take are :

'ui=xml' - This ui format returns the response in the form of a simple xml tree.[Content-Type: text/xml]

'ui=plain' - This ui format returns the response in the form of plain text readable in any text editor supporting unicode font rendering. [Content-Type: text/plain]

Note: Not including the variable 'ui' in the query, invokes the default response with 'Content-Type:text/html'

The following section shows a code snippet that demonstrates API access from Javascript:
*[url=http://www.cdacmumbai.in/xlit/editor/getXlatn]

xmlHttp= GetXmlHttpObject(); xmlHttp.open("POST", url, true); xmlHttp.setRequestHeader("Content-Type","text/html;charset=utf-8;"); xmlHttp.setRequestHeader("Content-Type","application/x-www-form-urlencoded"); xmlHttp.setRequestHeader("accept-charset","UTF-8"); var formData ="srclang=en&trgtlang=hi&inpword="+encodeURIComponent('Hello')+"&inputval=5&ui=plain"; xmlHttp.send(formData);

Similar API access from other languages/non-javascript environments like Flash, Java, PHP and Python, etc. is easily possible. You would be required to only construct the url and establish a connection.

Other Indian languages' support is in the pipeline for Marathi [mr], Tamil [tm], Telugu [tl], Kanada [kn] and Malayalam [ml].

We encourage you to participate and help us mould the API so that it serves you better day after day!

Post your comments/feedback/queries here.You may also wish to write to us at kbcs[AT]cdacmumbai.in

Sarvesh Nikumbh
Xlit Team

XlitHindi installation issues and Error Messages

Here are some problems you may face while installing the XlitHindi in OpenOfice.You may come across these while installing other extensions also. These are not specific to Xlithindi.

1] For error:
(com.sun.star.registry.CannotRegisterImplementationException) { { Message = "", Context = (com.sun.star.uno.XInterface) @0 } } } }

>>>> Please install 'openoffice.org-java-common' package from Synaptic manager [specific to Linux machines]

2] For error: Unable to load java while installing the extension

>>>> Please check that your machine has java 6 or above, installed. If you have it and you are still facing the same problem, now you need to mention path of the JRE to be used, in OpenOffice.Please do it this way:

* Go to Openoofice Writer: Tools --->Options --->OpenOffice.org ----->java.
* Now check 'use a Java runtime environment' and if you see one/more JREs there, select it/one of them.
* click 'OK'. You may try installing the extension now.

If you don't see any JRE in the window, click 'Add' and select the JRE you want to use from the file browser window.


Error Messages:
While using the XlitHindi you may come across the error message saying: "Transliteration Failed. May be a Network problem"
It means either your computer is not connected to INTERNET or OpenOffice is facing problem in connecting to internet or our server is down.
Upon such error please check your internet connection.If you are behind proxy please set the Proxy settings in OpenOffice as:

Go to Tools----->Options --->Internet--->Proxy.
Now select System . Say 'OK' and try using xlitHindi now; if that doesn't work select 'Manual' instead and fill the 'HTTP Proxy' and 'Port' numbers.

If  your OOo starts not responding  while you are trying the extension, please check your installed JAVA. To run the extension properly you need  "Sun Java" and not the "Free software foundation Java". Check this in Tools --->Options --->OpenOffice.org ----->java;  and make sure that, the Java being used is Sun Java. 

Please feel free to suggest/comment/query!

Prakash B. Pimpale
Xlit Team

XlitHindi extension for OOo released

KBCS, CDAC Mumbai has released its transliteration extension for the OpenOffice writer. Using this tool enables you to write in Hindi very easily without knowing the Hindi typing.It will be helpful for the people who use Hindi for the e-communication and documentation.The details of the English to other Indian Language transliteration system can be read at:http://www.cdacmumbai.in/xlit. This tool is really helpful for those who need the data entry in Indian Language, the Display or processing in Indian Language, like the government web sites meant for the citizen, the companies willing to target the non-English knowing consumer [having access to Internet], etc.
The XlitHindi extension can be downloaded from the OpenOffice extension repository [to directly reach XlitHindi page use this link http://extensions.services.openoffice.org/project/xlithindi ].
For any queries or problems you face, please feel free to comment.

Summary of the ML Workshop at CDAC Mumbai

Machine Learning Workshop

C-DAC Mumbai (formerly NCST) conducted a two day workshop on Machine Learning (ML) on 19-20th December, 2008 at the Kharghar, Navi Mumbai campus. The focus of the workshop was an introduction to various Machine Learning techniques and their usefulness in real life applications using an open source tool for ML called Weka.

The response for the workshop was overwhelming; as indicated by the number of participants, their background and the feedback we received from them.

The distribution of the 52 participants is as below:

Teachers -- 32

Students -- 07

Professionals -- 13

The participants were from across the country. The professionals were with different profiles like researchers, analysts, consultants and project leaders from various companies like TCS, Bank of America and Info Edge among others.

Almost all the participants appreciated the efforts of the speakers, lab sessions, materials provided, administration etc. They also suggested to conduct such events annually for advanced fields like this, with possibly longer duration.
The workshop covered Artificial Neural networks, Decision Trees,Clustering, Support Vector Machines and general issues in any Machine Learning technique.

View Last 2008 workshop announcement

Prashant More
KBCS

Here we are with a new URL for MaTra2

We are happy to announce the new URL of the MaTra2, its: http://www.cdacmumbai.in/matra/
Now you can use the English to Hindi MT system by clicking the above link.
For any queries contact us at kbcs@cdacmumbai.in .

Two day Workshop on Machine Learning (19-20 Dece. 2008)

Machine Learning

Machine Learning (ML) is a subfield of artificial intelligence (AI) that is concerned with the design and development of algorithms and techniques that allow computers to "learn". ML is one of the active areas of research in computer science and currently has a large repository of practically useable techniques and algorithms for a wide range of tasks. Typical ML problems include automatic clustering of a set of items, automatic classification (spam mail, documents, etc.), automatic learning/refinement of rules for a diagnostic system, predictive modelling, etc.

ML methods have evolved from various domains such as Statistics, Information theory, Biology and Control theory. ML has a wide spectrum of applications including natural language processing (NLP), pattern recognition, search engines, medical diagnosis, bioinformatics and chemical informatics, fraud detection, stock market analysis, speech and handwriting recognition, robotics, intelligent computer games etc.

About the Workshop

The workshop is meant to provide a comprehensive introduction to Machine Learning, focusing on conceptual understanding of popular ML algorithms and practical applications. Apart from covering popular ML techniques such as Artificial Neural Network (ANN), Support Vector Machine (SVM) and Genetic Algorithms (GA) we will discuss modelling of a problem for using machine learning including input-output transformation. Participant will get hands on experience with these algorithms; using toolkits such as Weka.

Workshop Outline

* Introduction to Machine Learning
* Inductive Learning
* Artificial Neural Network
* Support Vector Machine
* Clustering
* Input-output transformation for ML
* Application case studies
* Lab sessions

Target Audience

The workshop is targeted at academic and industry professionals interested in Machine Learning; professionals working in the area of information retrieval, language processing, document analysis, speech recognition and students interested in the area of Machine Learning. Some familiarity with computer programming will be desirable – language does not matter.

Registration Details

Registration Fee: Rs. 1000/- per participant for academic & non-profit organizations, and Rs. 1500/- per participant for others, payable by a crossed demand draft drawn in favour of 'C-DAC Mumbai' payable at Mumbai. The fee covers lunch, refreshments, and workshop material. The complete registration form fee should be sent to the Course Administration section C-DAC, Kharghar, Mumbai.

Accommodation Availability

Limited shared Non-A/C accommodation is available at the Navi Mumbai Campus Hostel, at Rs.150/- per person per day.

Knowledge Based Computer Systems Division

The Knowledge Based Computer Systems (KBCS) division carries out research and development in selected subfields of Artificial Intelligence. Its core areas of research are Natural Language Processing, Expert Systems, Case Based Reasoning, Information Retrieval, Data Mining, Soft Computing, and Planning and Scheduling.

About C-DAC Mumbai

The Centre for Development of Advanced Computing, Mumbai (formerly National Centre for Software Technology) is a scientific society involved in the Research and Development into various areas of Software Technology and related disciplines with an objective to create focus on advanced information technologies, high-end academics & training relevant to R&D societies.

Contact Us

Knowledge Based Computer Systems (KBCS) Division
Centre for Development of Advanced Computing (Formerly NCST)
Raintree Marg, Near Bharti Vidyapeeth
Opp. Kharghar Railway Station
Sector 7, CBD Belapur
Navi Mumbai 400 614

E-mail: Click here to mail
Telephone: +91-22-27565303-05 Fax: +91-22-27560004


for more information Visit: http://www.cdacmumbai.in

MaTra : An English to Hindi machine translation system

MaTra
Translating Languages ....
Revolutionizing Lives................!

India is a linguistically rich nation with 22 national languages recognized by the constitution of India, of which Hindi is the official language. Besides, there are 844 different dialects that are practised in various parts of the country. Being a developing nation India is in great need of information revolution for the growth of the geographically distributed multilingual society. Internet has helped in bridging this geographical barrier but the language barrier remains unanswered as only less than 10% of the population speaks English. 'MaTra' is an attempt to break this barrier.

The current version of 'MaTra' is a step towards the mission of information revolution for the millions of lives in this nation. MaTra, a fully-automatic indicative English-Hindi machine translation system, translates text in English to Hindi. Though the system is designed to support every domain, it focuses on ‘News’ and ‘Medical’ domains as of now. The system works well for simple sentences.

We are working on translation of other types of sentences such as compound, complex, interrogative, exclamatory, etc.

Using MaTra is simple, to try it please click here or try this :

http://www.cdacmumbai.in/matra

For recent updates on MaTra you can follow this Blog!