தமிழ்கருவி

2007-இல் எனது முதல் தமிழ் மென்பொருளை உருவாக்கினேன். இன்று தொலைந்த மென்பொருள்கலில் ஒன்று. எப்படி தொலயவிட்டேன் ? காலம்.

மிதம் https://egovindia.wordpress.com/2007/01/09/tamil-transliteration-tool-using-gtk-toolkit-for-gnome-environment/

Screen Shot 2018-08-11 at 1.47.00 AM

 

Language Transformations

Question  of Translation

How can you convert a text like “Me Amor!” to “என் உயிரே!” [from Spanish to தமிழ்] ? Lets  assume we have Spanish to English and Tamil to English translators [bidirectional with English] then we can convert Spanish to English then to Tamil. Likewise one can translate between any two languages from a clique of languages [so far as the clique is defined such that each language can be translated to at least one other language in clique].

Development – Theory

Language can exist as text (print/message/document) or speech (audio, conversations) etc. Ideas are represented in any language. Ideas originate from one language and move to another, or sometimes originate iñ many lañguages simultaneously. Ideas cañ cross from oñe language to añother via text or speech.

In mathematical terms if we write L as set of lañguages = { L1, L2, .. Ln} and then if we define each language as a tuple Li = (Ti,Si) then we may further define mathematical function operating on text and converting it to speech as :

TTSi : Ti -> Si

we may define a function speech recognition as,

ASRi : Si -> Ti

we may also define a translation function as,

TXij : Li -> Lj

Essentially what we can do is by representing the language as a node in a graph with two text and speech parts to it, we may connect these nodes to each other via the edges – functions – like ASR and TTS, and to nodes of other languages via translators function edge.

In a graph with only two languages [English, Tamil] with all edges representing functions like TTS, ASR within same language and functions like Translator between two languages (one for each direction) we see a graph like the following:

Screen Shot 2018-08-03 at 11.51.08 PM

Fig. 1: Language transformation graph. Nodes represent languages and their components. Edges represent functions like TTS, ASR [for same language] and Translators [directional between languages]. Clearly we may see this is a directed graph with ability to go from a specific language to another language in text or speech or both forms, provided a path exists from source to target language. Using such a graph with no orphan nodes, we may have universal translation powers from language A to language B [so far as bidirectional connectivity is present with at least one neighbor].

Problems to Ponder

So the curious reader now having a background of representing the translation problem as a graph problem of reaching node B from node A, can use rich set of path finding algorithms and shortest distance algorithms may attempt to answer some of these questions:

  1. What is the graph criteria for a language to have no translations ?
  2. What is the graph criteria for a language to not be able to have virtual assistant ? [Siri, Cortana, Alexa etc.]
  3. Conversely, to 2, what is minimum criteria [necessary but not sufficient] to have a virtual assistant [that can speak and listen] ?
  4. Given two paths to translating from language A -> F, which are of two different lengths which one would you choose and why? Assume all jumps have a uniform information loss. What if information loss at each edge is non-uniform, how can you optimized such a problem ?
  5. How would you introduce a new language into this graph so that it maybe translated to all other languages [unidirectionally] ?
  6. How would you introduce a new language into this graph so that it can be bi-directionally translated ?
  7. How can you represent the transliteration function in this graph ?

Answers will be posted soon! Feel free to leave your comments in section below.

-Muthu

Dispatches from Kovai, India

Kovai – Coimbatore, India – also known as Manchester of India, and various monickers for its beautiful weather, Tamil dialect, and cool weather at foothills of the Nilgiris-Ooty mountain range and clear waters from Sravani river, it is a microcosm of all that is nice and beautiful in முல்லை Tamilnadu.

Our team of fearless open-source enthusiasts añd contributing developers of open-tamil project, Shrinivasan and our newest contributor this year Syed Abuthahir made an impressive presentation at the conference, and gained new clients for the open-tamil library demonstrating the prowess of various tools, particularly from the web-interface http://tamilpesu.us

Screen Shot 2018-07-19 at 12.49.53 AM

Growth and evolution of Open-Tamil – presented at TNAU, Coimbatore. Full slides here

Syed has gladly shared the presentation with me, and I put it up on slideshare for all of us to read/comment. Thanks to all earnest contributions of time, effort, code and thought towards this project over the years!

Best Regards,

-Muthu

India A.I. report – highlights

ஏற்கணவே எழுதிணபடி  இந்திய செயற்கை நுண்ணறிவு அறிக்கை வெளியிட்ட குழுவின் தலைவர்,  IIT-சென்னையைச் சேர்ந்த பேரா. காமகோடி. இந்த அறிக்கையில், முக்கியமான விஷயங்ககள் கீழே படம் வடிவங்களில் பாற்க;

India-AI-report-1

படம் 1: இந்திய செயற்கை நுண்ணறிவு அறிக்கை – மாற்றுத்திறணாளிகள் பற்றி

India-AI-report-2

படம் 2: இந்திய செயற்கை நுண்ணறிவு அறிக்கை – இந்தியமொழிகள் பற்றி

Tournament Model

Muthu@SFO-May-2018.jpg

This year I had chance to speak at my undergraduate institution – a well recognized engineering school in Trichy, India – about various things concerning my professional development and understanding of Science, Engineering and innovation in my short career as software developer and scientist-in-training.

Primarily, my goal was to communicate the tournament model and how we may enjoy our time in educational institutions pursuing a quest for truth regardless of some of the outcomes – just because they are governed by the tournament model.

Consider your task: to pick a winner in 2-player games from a group of N (say 128 or 64 players – like a typical Tennis tournament [or teams of smaller sizes for IPL or World Cup cricket tournaments]) then goal is to organize the games as a championship format with league rounds and knock-out tournaments to eventual final which decides the winner. This is the tournament model.

An alternate version where number of teams/players participating is not a power of 2, we may setup the model as follows algorithm/pseudocode;

  1. Enter all teams/payers in a double-ended-queue [deque]
  2. Select first-2 teams in queue and let them play;
  3. Take the winner of this game and enqueue to the end of queue; discard the loser (obviously!)
  4. Now we have N-1 teams/players in the queue.
  5. Repeat steps 2-4, till number of players is 1.
  6. We have a winner!

Key insight of tournament model is the fact that small differences between entities participating in the model can be amplified by the model making winners, and effects like the Matthew effect can ensure initial advantages snowball over time [esp. in industries like entertainment, social networking etc.]

The tournament model decides frequency of India vs Pakistan cricket matches, why Nadal vs Federer is most likely grand-slam final match up; the system decides success of professional actors and actresses. Why are Kamal Hassan and Rajinikanth more famous than other talented male actors of their generation (e.g. Sathyaraj, Karthik, Prabhu, etc.)[not to mention other female actresses – a whole other question]. Modern day movie star rivalries are also plenty, to wit – Danush vs Simbu etc. in their ascent to fame.

Many principles of randomness of outcomes, and regression toward mean explain the outcomes in retrospect; but none of the techniques have an ability to explain these phenomenon in a predictive manner which one may seek.

Hence as students approaching a potentially lifetime of work in field of engineering or science, I recommend everyone to aspire to understand the fundamental pieces – to learn the instruments, notes, chords, scales of their musical pieces – not just the piece itself- so in the future you can compose your own orchestral music; so that you can build tools for future challenges that you may face – surely different from challenges you were taught to resolve – using an open ended approach to learning.

Tournament model also helps you handle failures – be it product, strategy, problem areas in life. Usually, losing at something is by not making the grade or placing second or being edge out is by being marginally “less” in some way, shape or form, compared to competition.

What is your experience with managing technology projects, and their outcomes ? Leave your comment below.

-M.A.

Namashkaar!

A.I./ML for Hindi Language Processing

Sometimes its good to look around and learn from what’s happening in other realms of Indian language processing. In my limited experience language efforts in computing for Indian language revolve around the Dravidian languages, Bengali, Marathi or Hindi. சில நேரங்களில் குண்டு சட்டியில் குதிரை ஓட்டுரமாதிரி கணினி மொழியியல் ஆயிடக்கூடாது – தனிபட்டபடியும் சரி – மொழிகளுக்கிடையிலும் சரி.

Some good project efforts in Hindi Language processing (open-source) are reviewed in this blog; [there are  projects like open-tamil API for Hindi, e.g. a get_letters like function, provided by tokenizer project here (with caveat that it is a small function only compared to expansive open-tamil), but we talk about the ML/A.I. focused projects here].

  1. Hindi word embedding called Hindi2vec (along lines of word2vec project). The idea is to associate similar words (e.g. ‘பல்’,’நாக்கு’,’வாய்’) with similar vectors within a neighborhood of each other using concepts of linear-algebra – vector spaces and matrices. So when you search or mistype or want to classify there is a neighborhood of known words closer to the potentially unknown word input from the user; such known neighborhood identification can help decision making and drive various learning, classification or dialogue systems.
  2. Hindi Transliteration Model project and the DeepTrans project– this is a really cool where they developed a reference data set of English to Hindi and trained a model for transliteration from English to Hindi of user input.
    1. We can do this in Tamil with the as we have many transliteration schemes as set out in open-tamil, but the even a same user is not strictly going to follow the scheme strictly, nor do different users follow the same scheme – in all these cases a machine learning A.I. model maybe more robust by virtue of learning the underlying rules. Very interesting project, and fairly simple to implement for Tamil from open-tamil transliterate module and SciKit Learn or other frameworks with high 95% correct prediction rate.
  3. Hindi-English parallel dictionary with 8MB size (probably 500,000 words or so I imagine) here – this can be a good jump starting point for translation projects if such existed for Tamil. e.g. Can we have a parallel dictionary English – Tamil for the simple TVU word list/dictionary ?
  4. Hindi Sentiment Analysis project does a ternary [good, bad, neutral] classification of text. They do this by using a CDAC-model which is super curious to me; maybe CDAC-India (Pune) has a Tamil POS-Tagger too ? Probably they do.
    1. Tamil POS-Taggers widely reported; AU-KBC Chennai has a POS-Tagger, probably the best for Tamil; Dr. Vasu Renganathan has a POS-Tagger, but both these works are not available currently for open-source use, however their techniques are openly shared via their papers in INFITT conferences.
    2. Sorkandu project can also be revived for making an open-source POS-Tagger
  5. Emotion Recognition in Hindi Speech project – this work from IIT KGP students builds a reference audio data set with known emotion labels and build some kind of a machine learning model, and then they get 5x better than random coin-toss/guess for the audio emotion recognition from speech.
    1. We probably don’t have any work on this direction in the open, but interestingly NIST in USA sponsored a Tamil Key Word Search (KWS), reports of which were published by a Singapore team in academic journals. More interestingly the KWS challenge released 2 hrs of speech data with tagged information. In USA, government released data usually qualifies for public-domain – e.g. pictures from NASA etc. so maybe there is a way to get this data. கடவுளுக்கு தான் வெளிச்சம்!

While we know, Google ASR, Youtube online translation of English videos into Tamil closed-captioning, foreign languages to Tamil Translation, Transliteration inputs all use perhaps the most advanced models in Tensorflow on cloud hardware, none of this technology is directly usable for free – maybe for a price via their Google cloud API offerings – and we probably don’t know all the details of how they achieved these magical software applications for Tamil language – anyones guess like mine is using the massive data sets they have from our Tamil news groups, emails, websites, and user input + Tensorflow A.I / ML magic. At least, we have to be grateful for Google-aandavar like some friends commented on freetamilcomputing group. 🙂

Surprisingly, to my knowledge, there are no planned efforts, ongoing or completed open-source projects like these in Tamil. Maybe another avenue for growth, and in this case Hindi projects (at least in open-source domain) seem to have forged ahead!

Shukriya.

-Muthu

 

 

iMessage troubles with Telugu language display

I’m somewhat late to hear this news, but recently there was a bug in iMessage application of iOS and Mac platforms causing it to crash/freeze up as reported here.

The crashing character consists of 5 codepoints.

U+0C1C [Lo] TELUGU LETTER JA
U+0C4D [Mn] TELUGU SIGN VIRAMA
U+0C1E [Lo] TELUGU LETTER NYA
U+200C [Cf] ZERO WIDTH NON-JOINER
U+0C3E [Mn] TELUGU VOWEL SIGN AA

and has a complex ligature form.

Sadly the poor usage of Telugu language Unicode block is potential cause to discover the bug so late; however the root cause of bug is some unsightly buffer overflow in Apple’s codebase.

கஷ்டகாலம்.

Who knows if Tamil Unicode has any equivalent horrible bugs living in Android/Apple/Mac/Linux/Windows platforms. Hopefully not.