Difference between revisions of "User:Pmodi/GSOC 2020 proposal: Hindi-Punjabi"

From Apertium
Jump to navigation Jump to search
Line 231: Line 231:
 
'''Second Evaluation(July 27th - July 31st)'''
 
'''Second Evaluation(July 27th - July 31st)'''
 
| style="text-align:center" | ~ 15,000
 
| style="text-align:center" | ~ 15,000
  +
| style="text-align:center" | <20% (hin-pan) <br> <25%(pan-hin)
|
 
  +
| style="text-align:center" | >85% (hin-pan) <br> >80&(pan-hin)
|
 
 
|-style="background-color:#008a00;"
 
|-style="background-color:#008a00;"
 
| style="text-align:center" | Week ELEVEN
 
| style="text-align:center" | Week ELEVEN
Line 278: Line 278:
 
'''Final evaluation(August 24th - August 31st)
 
'''Final evaluation(August 24th - August 31st)
 
| style="text-align:center" | ~ 17,000
 
| style="text-align:center" | ~ 17,000
  +
| style="text-align:center" | ~15% (hin-pan) <br> <20% (pan-hin)
|
 
  +
| style="text-align:center" | ~90% (hin-pan) <br> ~87% (pan-hin)
|
 
 
|}
 
|}
   

Revision as of 20:50, 27 March 2020

Contact Information

Name: Priyank Modi
Email: priyankmodi99@gmail.com
Current Designation: Undergraduate Researcher in the LTRC Lab, IIIT Hyderabad (completing 6th semester/3rd year in April '20) and a Teaching Assistant for Linguistics courses
IRC: pmodi
Timezone: GMT +0530 hrs
Linkedin: https://www.linkedin.com/in/priyank-modi-81584b175/
Github: https://github.com/priyankmodiPM

Why I am interested in Apertium

Apertium is an Open Source Rule-based machine translation system. Being an undergrad researcher at the LTRC lab in IIIT-H currently working on understanding the nuances of Indian languages and developing systems which improve our analysis of the same, Machine Translation interests me because it’s a complex problem which tries to achieve something most people believe is only achievable by humans.

Translating data to other languages, and especially low - resource languages gives the speakers of those languages access to valuable data and can help in several domains, such as education, news, judiciary, etc. The dictionaries made in the process are crucial for low resource languages and can even help making spell checkers etc.

The most striking factor for me is the fact that while recent trends to find a solution to MT lean towards Neural Networks and Deep Learning, which fall short when it comes to resource-poor languages, Apertium looks to tackle this using a rule based approach. Not only is this beneficial because of the level of understanding it provides instead of simply blaming data for poor results, it actually shows that it can perform better for low resource languages(even for the pair I present in my proposal).

A tool which is rule-based and open source really helps the community with language pairs that are resource - poor and gives them free translations for their needs and that is why I want to work on improving on it. I want to work with Apertium and GSoC so I can contribute to an important Open Source Tool while also honing my own skills, and I hope to become a part of this amazing community of developers!

Which of the published tasks are you interested in? What do you plan to do?

Adopt an unreleased language pair. I plan on developing the Hindi-Punjabi language pair in both directions i.e. hin-pan and pan-hin. This'll involve improving the monolingual dictionaries for both languages, the hin-pan bilingual dictionary and writing suitable transfer rules to bring this pair to a releasable state.

My Proposal

Why Google and Apertium should sponsor it

  • Both Hindi and Punjabi are widely spoken languages, both by number of speakers and geographic spread. Despite that, Punjabi especially has very limited online resources.
  • Services like Google Translate give unsatisfactory results when it comes to translation of this pair(see Section 2.1) On the contrary, I was able to achieve close to human translation for some sentences using minimal rules and time(see Section 3 : Coding Challenge).
  • I believe the Apertium architecture is suited perfectly for this pair and can replace the current state-of-art translator for this pair.
  • This is an important project(since it adds diversity to Apertium and translation systems in general) which requires at least 2-3 months of dedicated work and can be an important resource.

How and who it will benefit in society

As mentioned above, the Apertium community is strongly committed to under-resourced and minoritised/marginalised languages and Google helps its own way via programs like GSoC and GCI. There exists a good amount of vernacular literature and scriptures which could be circulated to a larger group of people if this project is successful. It'll be an important open source dictionary resource for both languages. My larger aim from this project is to develop a chain of pairs covering Indian languages. Since Urdu and Punjabi share their roots, at least one more pair can be developed with minimum effort. I plan of working towards the Hindi-English pair next year since by then I'll have a good understanding of the architecture a cross-language-family pair.

Google Translate : Analysis and comparison

Google Translate provides an interface to translate the pair in question. I have analysed the results of the translation into Punjabi from Google. The numerical results(computed on a small set of sentences from the coding challenge. The human translation which has been reviewed by 3 annotators is also available in the repo) are given below(source-target):

  • hin-pan: 79.23% WER
  • hin-pan: 56.56% PER
  • pan-hin: 82.23% WER
  • pan-hin: 57.83% WER

The results are simply poor, especially when it comes to longer sentences with less frequently used words. It is rather easy to see that Google Translate doesn't try to capture the case or tense in sentences, rather picks the most commonly used form of that particular root. NER is very limited, in contrast to the Apertium module which captures it well(because of it's 3 stage transfer mechanism I believe). The use of English as intermediate(which seems to be the case here as well because some words translate to English and fail to convert to Punjabi maybe because of some errors in parsing,as pointed by Hector) causes problems, such as the incorrect choice of tense in the verbs, wrong choice/dissappearance of some pronouns and the inability to handle copula constructions as well as verbal clauses(especially when other words occur between two sub-clauses). Here is an example of some of these form the Hindi test text:

Original source text (Hindi):

गिरजा आज फिर उस औरत को साथ लाया था.वही दुबली पतली मोटी-मोटी आंखें तीखी नाक और सांवले रंग वाली औरत.
Girija brought that woman with him again today. The same thin, big-eyed, pointy nosed and dusky woman.

Google translation (Punjabi):

ਚਰਚ ਨੇ todayਰਤ ਨੂੰ ਅੱਜ ਵਾਪਸ ਲਿਆਇਆ, ਉਹੀ ਪਤਲੀ womanਰਤ ਜਿਹੜੀ ਸੰਘਣੀ ਅੱਖਾਂ, ਤਿੱਖੀ ਨੱਕ ਅਤੇ ਹਨੇਰਾ ਰੰਗ.
The Church brought back todayਰਤ today. The same thin woman which big-eyed, pointy nose and dark colored.

Translation achieved using Apertium model(Punjabi):

ਗਿਰਜਾ ਅਜ੍ਜ ਫਿਰ ਉਸ ਔਰਤ ਨੂੰ ਨਾਲ ਲਾਇਆ ਸੀ.ਉਹੀ ਦੁਬਲੀ ਪਤਲੀ ਮੋਟੀ-ਮੋਟੀ ਅੱਖਾਂ ਤਿਖੀ ਨੱਕ ਅਤੇ ਸਾਉਲੇ #ਰਂਗ ਵਾਲੀ ਔਰਤ.
Girija brought that woman with him again today. The same thin, big-eyed, pointy nosed and dusky woman.

It is not difficult to see that most translations provided by Google Translate lead to a change in meaning. One clear reason that can be seen is that Google Translate relies on the n-grams available to it, and in case of rarely used words, it fails to translate those and worse, fails to capture the tense. In complex sentences, the chunking(stage 1 and 2 as per apertium model) fails hence leading to a failure in capturing meaning and very often, even generating any syntactically correct sentence.

Current state of dictionaries

A released module already exists for Hindi(as part of the urd-hin pair). However there still exist a lot of anomalies in the Hindi mono-dictionary. I've compiled a preliminary version of list of some these here[insert link]. Apart from these, the existing state of the hin-pan bi-dictionary also needs massive improvement. The first step of this project will be to revise these lists of issues and come up with a sustainable solution. It'll be crucial that the changes made, especially to the Hindi mono-dictionary do no affect the urd-hin pair(and the hindi-begali, hindi-marathi and hindi-gujarati pairs which also have little but some work done) in a negative way.

Resources

[to be added - under confirmation for public use]
Hindi-Punjabi Dictionary - wiktionary
Punjabi-Hindi dictionary - Glosbe (awaiting confirmation)
Punjabi Articles - Wikipedia
Punjabi Dictionary - Wiktionary

Workplan

PHASE DURATION GOALS OF THE WEEK BIDIX WER Coverage
COMMUNITY BONDING PERIOD
  • START:April 26th
  • END:May 17th
  • List and discuss implementation choices of hin-pan bidix and urd-hin pair
  • Reading up on the details of Transfer rules(whether or not a 3-stage transfer is the best way for this pair) and assigning weights
  • Finding Language Resources
  • Making Frequency lists
Post Application Period : CLOSED CATEGORIES
  • START:May 18th
  • END:May 24th
  • Function words(voc prn, cnj, det, prn, post, gen_endings)
  • Transfer rules for post-positions
Post Application Period : Adjectives
  • START:May 25th
  • END:May 31st
  • Punjabi mono-dictionary : adj and adv
  • Expanding bilingual dictionary
  • Lexical selection rules for adj and adv
Week ONE: Verbal Paradigms
  • START:June 1st
  • END:June 7th
  • Punjabi mono-dictionary : Verbal paradigms(vblex, vbser, vaux)
  • Expanding bilingual dictionary
  • Lexical selection rules for verbs
  • testvoc : adj, adv
~ 3,000
Week TWO: Dictionary Expansion
  • START:June 8th
  • END:June 14th
  • Expanding bilingual dictionary
  • Lexical selection rules
~ 5,000
Week THREE: Dictionary Expansion
  • START:June 15th
  • END:June 21st
  • Expanding bilingual dictionary
  • Lexical selection rules
~ 6,500 < 25%(hin-pan) > 75%(hin-pan)
>70%(pan-hin)
Week FOUR: More works on verbs and testing
  • START:June 15th
  • END:June 21st
  • Expanding bilingual dictionary
  • Lexical selection rules
  • Manual Disambiguation of rules hin-pan(src-trg)
~ 7,500
Week FIVE : focus on Nouns
  • START:June 22nd
  • END:June 28th
  • Expanding bilingual dictionary
  • Lexical selection rules
~ 9,000
Week SIX : Expanding Dictionaries
  • START:June 29th
  • END:July 5th
  • Expanding bilingual dictionary
  • Lexical selection rules

First Evaluation(June 29th - July 3rd)

~ 10,500
Week SEVEN : Expanding Dictionaries
  • START:July 6th
  • END:July 12th
  • Expanding bilingual dictionary
  • Lexical selection rules
  • Manual disambiguation of rules(pan-hin)
~ 12,000
Week EIGHT : Transfer rules(hin-pan)
  • START:July 13th
  • END:July 19th
  • Expanding bilingual dictionary
  • Lexical selection rules
  • Transfer rules(hin-pan)
~ 13,000
Week NINE : Transfer rules
  • START:July 20th
  • END:July 26th
  • Expanding bilingual dictionary
  • Lexical selection rules
  • Transfer rules : pan-hin
~ 14,000
Week TEN
  • START:July 27th
  • END:August 2nd
  • Expanding bilingual dictionary
  • Lexical selection rules

Second Evaluation(July 27th - July 31st)

~ 15,000 <20% (hin-pan)
<25%(pan-hin)
>85% (hin-pan)
>80&(pan-hin)
Week ELEVEN
  • START:August 3rd
  • END:August 9th
  • Expanding bilingual dictionary
  • Lexical selection rules
  • Disambiguation rules
  • Transfer rules
~ 16,000
Week TWELVE : Testvoc
  • START:August 10th
  • END:August 16th
  • Testvoc hin-pan
  • Add rules, words
~ 16,500
Week THIRTEEN : Finishing up
  • START:August 17th
  • END:August 23rd
  • Testvoc pan-hin
  • Add rules, words

Personal Code freeze : August 22nd

~ 17,000
Week FOURTEEN : Review
  • START:August 24th
  • END:August 30th
  • Review and documentation

Final evaluation(August 24th - August 31st)

~ 17,000 ~15% (hin-pan)
<20% (pan-hin)
~90% (hin-pan)
~87% (pan-hin)

Skills

I'm currently a third year(commencing start of April '20 ) student at IIIT Hyderabad where I'm studying Computational Linguistics. It is a dual degree course where we study Computer Science, Linguistics, NLP and more.

I've been interested in linguistics from the very beginning and due to the rigorous programming courses, I'm also adept at several programming languages like Python, C++, XML, Bash Scripting, etc. I'm skilled in writing Algorithms. Data Structures, and Machine Learning Algorithms as well.

I also have a lot of experience studying and generating data which I feel is essential in solving any problem, especially the one mentioned in this proposal. My paper on 'Hindi TimeBank: An ISO-TimeML Annotated Reference Corpus' recently got accepted in 16th Joint ACL - ISO Workshop on Interoperable Semantic Annotation at LREC 2020. I am working on extending the same for Punjabi using Transfer learning.

I am also closely involved with the committee conducting Asia-Pacific Linguistics Olympiad(which holds a camp, mentors and prepares students for the International Linguistics Olympiad) and help with the organisation and judging for the same.

Due to the focused nature of our courses, I have worked in several projects, such as building Anaphora Resolution systems, Abstractive Summarizers(using Pointer-generators, hierarchical attention and transformers), POS Taggers, Named Entity Recognisers, simple Q-A systems, a Linux based shell etc. all of which required a working understanding of Natural Language Processing scripting. Some of these projects aren't available on GitHub because of the privacy settings but can be provided if required.

I am fluent in English, Hindi and Punjabi.

Coding challenge

I've completed the coding challenge for translation from Hindi into Punjabi. You can find my work here : Coding challenge repository
Original corpus : source lang-hin
Translated output : target lang-pan
Human Translation : target lang-pan(human)

Results : Source - Hindi, Target - Punjabi (evaluator output included in repo)
(to be checked and revised since WER and PER before and after removing unknown words remains same)
WER achieved : 15.30 %
PER achieved : 93.45 %

Non-Summer-of-Code plans for the Summer

Since I'll be having my college summer vacations for almost the entire duration of the project, I can easily spend 30-40 hours per week on the project. Since, the academic schedule might vary a little bit due to lock downs for prevention of COVID-19, I'll be starting work early and cover the problems in the post-application period. I've also kept workload slightly heavier in the first 2 weeks to cover up any unlikely, uncertain extensions in academics that might show up. Even then, I can spend around 20 hours a week in any case(note that this is a very unlikely situation and even then this period won't last more than a couple of weeks since the coursework is already underway online and is expected to be over before start of the project).