Difference between revisions of "User:Khannatanmai"

From Apertium
Jump to navigation Jump to search
 
(102 intermediate revisions by the same user not shown)
Line 1: Line 1:
'''Google Summer of Code 2019: Proposal [Second Draft]'''

'''Anaphora Resolution'''

== Personal Details ==
== Personal Details ==

'''Name:''' Tanmai Khanna
'''Name:''' Tanmai Khanna


Line 15: Line 10:
'''LinkedIn:''' [http://linkedin.com/in/khannatanmai khannatanmai]
'''LinkedIn:''' [http://linkedin.com/in/khannatanmai khannatanmai]


'''Current Job:''' Teaching Assistant for Linguistics courses and an Undergraduate Researcher in the LTRC Lab, IIIT Hyderabad
'''Current Designation:''' Graduate Researcher in the LTRC Lab, IIIT Hyderabad (5th year student) and a Teaching Assistant for Linguistics courses


'''Time Zone:''' GMT+5:30
'''Time Zone:''' GMT+5:30
Line 21: Line 16:
== About Me ==
== About Me ==


'''Professional Interests:''' I’m currently doing research in Computational Linguistics and I have a particular interest in Linguistics and NLP tools, specifically Machine Translation and its components.
'''Open Source Softwares I use:''' I have used Apertium in the past, Ubuntu, Firefox, VLC.

'''Professional Interests:''' I’m currently studying NLP and I have a particular interest in Linguistics and NLP tools, specifically Machine Translation and its components.

'''Hobbies:''' I love Parliamentary Debating, Singing, and Reading.

'''What I want to get out of GSoC'''

I’ve enjoyed using Apertium in various personal and academic projects and it’s amazing to me that I get an opportunity to work with them.

NLP is my passion, and I would love to work with similarly passionate people at Apertium, to develop tools that people actually benefit from. This would be an invaluable experience that classes just can't match.

I am applying for GSoC, as the stipend would allow me to dedicate my full attention to the project during the 3 months.

=== Why is it that I am interested in Apertium and Machine Translation? ===

Apertium is an Open Source Rule-based MT system. I'm a researcher in the IIIT-H LTRC lab, currently working on Machine Translation and it interests me because it’s a complex problem and is often called NLP-Complete by my professors, i.e. it uses most of the tools NLP has to offer and hence if one learns to do good MT they learn most of Natural Language Processing.

Each part of Apertium's mission statement, especially the fact that they focus on Low Resource Languages, excites me to be working with them. While recent trends lean towards Neural Networks and Deep Learning, they fall short when it comes to resource-poor languages. Anaphora Resolution without complex linguistic information is a challenge that I'll be tackling during this Summer of Code with Apertium.

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!

== Project Proposal ==
=== Which of the published tasks am I interested in? What do I plan to do? ===

Anaphora Resolution - Apertium currently uses default male. I wish to build a perfect Anaphora Resolution system, but that is obviously not possible in 3 months so I propose we limit the scope of this project to use complex features to pick an antecedent as a way to increase the fluency and intelligibility of output significantly.

The Anaphora Resolution tool will be language agnostic and will improve the output for most language pairs in Apertium. Pronouns are present in a lot of languages and they change based on their antecedent's gender, number, person, etc. The aim is to figure out the antecedent and choose the correct pronoun accordingly.

Sentences like “The group agreed to release his mission statement”, "The girl ate his apple" are grammatically incoherent and an incorrect pronoun will more often than not confuse people in complex sentences.

What puts people off Machine Translation is the lack of fluency, and this tool will definitely generate more fluent sentences leading to more trust in this tool.

'''NOTE: Anaphora Resolution is one part of resolving long-distance dependencies. The method of resolving this will not be limited to anaphora and can be used for general coreference, agreement and other long-distance dependencies which need to identify the antecedent.'''

This project has promising future prospects:

* Adding Language Specific metrics for better identification
* Extending system to general coreference needed for MT
* Shallow parsing for low resource languages

I want to be a contributor to Apertium even after GSoC and would love to continue this project forward after this summer.

=== Proposed Improvements ===

I will be working with Spanish-English and Catalan-English pairs while developing the tool. However, the features will be made largely language agnostic and I will also evaluate how well it works for other language pairs which need Anaphora Resolution.

Ultimately, the system should be able to do Anaphora Resolution for the following:

* '''Possessive Pronouns'''

Spanish Sentence: La chica comió su manzana

Apertium Translation: The girl ate his apple

After Anaphora [Proposed Translation]: The girl ate '''her''' apple

* '''Zero Pronouns'''

'''Eg. 1.'''
Spanish Sentence: Canta bien

Apertium Translation: It sings well

After Anaphora [Proposed Translation]: He/She/It sings well (Based on context)

'''Eg. 2.'''
Spanish Sentence: La chica está aquí, lleva un vestido rojo

Apertium Translation: The girl is here, spends a red dress

After Anaphora [Proposed Translation]: The girl is here, '''she''' spends/wears a red dress

* '''Reflexive Pronouns'''

Spanish Sentence: Se mató

Apertium Translation: It killed

After Anaphora [Proposed Translation]: He/She killed himself/herself

* '''Long Distance Agreement'''

English Sentence: The table is here. It is red.

Apertium Translation: La mesa es aquí. Es rojo.

Proposed Translation: La mesa es aquí. Es '''roja.'''

----

=== Idea Description ===

Since Apertium deals with low resource language pairs, this module will use very basic linguistic information - POS tags, gender, number information as opposed to parse trees, which require more data to be accurate.

The module will assign '''salience scores''' to all antecedents in the context window.

==== Description of Methods used for the Improvements ====

'''This tool will not deal with First/Second Person pronouns. Here's why:'''

* For most languages only Third Person pronouns are differentiated by gender. Since our task is MT-based Anaphora Resolution, we can leave alone First and Second person as we can translate them correctly without resolving them.
* First and Second Person pronouns are largely resolved in the real world and including them in this tool will produce worse results than leaving them alone.

===== Possessive Pronouns =====

For possessive pronouns, we will use saliency scores from the factors mentioned below (Lappin and Leass Algorithm, Mitkov's Algorithm, Baldwin's Salience Features) and the highest scored antecedent will be picked as the antecedent.

If the antecedent is another pronoun, it's identified antecedent will be attached to this pronoun (dealing with repetition in examples such as "John lost his family, his house, his job and his dignity" ).

===== Reflexive Pronouns =====

Chomsky’s Binding theory (1981) says reflexive pronouns must have local antecedents, which make them easier for us to detect. For Eg.,

* John(i) washed himself(i) [CORRECT]

* John(i) asked Mary to wash himself(i) [INCORRECT]

These will be resolved by identifying the noun closest to the verb. This identification based on simple linguistic information can be done if we can ignore embedded clauses effectively.

===== Zero Pronouns =====

Zero Pronouns will be resolved by '''literally assuming a pronoun before the verb and then resolving it''' using the same way we resolve possessive pronouns.

Due to word order differences in languages, this might give different results for different types of languages. Hence, we can evaluate this method and restrict it to certain languages if needed.

'''NOTE:''' Currently verbs in Spanish without a pronoun are translated to "it + verb". This design choice makes sense due to the existence of '''Pleonastic Pronouns''', i.e. pronouns which don't refer to any entity. For eg., "It is raining".

However, once we start trying to resolve Zero Anaphora, we will have to also distinguish Pleonastic Pronouns from Non-Pleonastic Pronouns which is largely done by semantics.

===== Long Distance Agreement for Adjectives =====

Long distance agreement for adjectives will become easier when the nearest pronoun has been resolved by the previous steps.

Here we can also try to ignore embedded clauses and the problem becomes similar to Reflexive pronouns where we need to identify the subject of the verb, in this case, the verb that connects the noun to the adjective.

==== Existing Algorithms ====

There are several existing algorithms for Anaphora Resolution:
[Add links here]

However, all of these algorithms use Parse Trees or Machine Learning, both of which require a large amount of data to give accurate outputs.

The following algorithms use, or can be modified to use very simple linguistic information already available in the Apertium Pipeline to perform reasonable Anaphora Resolution:

===== Lappin and Leass Algorithm =====

This algorithm considers these following factors in resolving the antecedent of a pronoun:

* '''Recency:''' Entities introduced in recent utterances are more likely to be referred to by a pronoun than entities introduced in utterances further back.
* '''Grammatical Role:''' Entities introduced in subject position tend to get topicalised, and are more likely to be referred to by a pronoun than entities in object positions.
* '''Repetition:''' Entities that have already been referred to frequently are more likely to be pronominalised than those that have not.

George needed a new car. His previous car got totalled, and he had recently come into some money. Jerry went with him to the car dealers. He bought a Nexus.

* '''Parallelism:''' Pronouns are more likely to refer to those entities that do not violate syntactically parallel constructions.

John took Bill to the zoo; Mary took him to the park.

====== Algorithm ======

* Resolving each Pronoun from left to right
* Collect potential referents from up to 4 sentences back
* Filter out antecedents that don’t satisfy agreement/syntax constraints
* Select remaining antecedents with the highest salience value; add pronoun to class.

====== Saliency Factors ======

[[File:SaliencyFactors.png]]
[[File:SaliencyFactors.png|1000px|thumb|left|Saliency Factors]]

* The salience of a referent is the sum of all applicable weights
* The salience of a referent is halved each time a sentence boundary is crossed
* This, along with the weight for being in the current sentence, makes more recent referents more salient

* Lappin and Leass report 86% accuracy for their algorithm on a corpus of Computer manuals

References:

[https://www.aclweb.org/anthology/J94-4002 Original Paper: An Algorithm for Pronominal Resolution]

[https://www.cl.cam.ac.uk/teaching/1011/L104/lec12-2x2.pdf Anaphora Resolution, University of Cambridge]

===== Antecedent Indicators marked by Ruslan Mitkov =====

'''Boosting Indicators''' (Will be updated as I study the data and find more indicators)

* First NPs
* Lexical Reiteration: Lexically reiterated items are likely candidates for antecedents.
* Section Heading Preference
* Collocation Pattern Preference: This preference is given to candidates which have an identical collocation pattern with a pronoun.
* Immediate reference (if it is pronoun then its reference)
* Sequential Instructions

'''Impeding Indicators''' (Will be updated as I study the data and find more indicators)

* Indefiniteness
* Prepositional NPs: NPs which are part of a PP are penalized.

Reference : [https://link.springer.com/content/pdf/10.1023%2FA%3A1011184828072.pdf Multilingual Anaphora Resolution, Ruslan Mitkov]

'''We can add more factors, such as detecting heads of NPs, ignoring NPs that are part of Relative clauses to make the system more accurate.'''

===== Salience Features to find Coreference =====

* Unique in Discourse
If there is a single possible antecedent i in the read-in portion of the entire discourse, then pick i as the antecedent: 8 correct, and 0 incorrect.

* Reflexive
Pick nearest possible antecedent in the read-in portion of the current sentence if the anaphor is a reflexive pronoun: 16 correct, and 1 incorrect.

* Unique in Current + Prior
If there is a single possible antecedent i in the prior sentence and the read-in portion of the current sentence, then pick i as the antecedent: 114 correct, and 2 incorrect.

* Possessive Pronoun
If the anaphor is a possessive pronoun and there is a single exact string match i of the possessive in the prior sentence, then pick i as the antecedent: 4 correct, and 1 incorrect.

*Unique Subject/ Subject Pronoun
If the subject of the prior sentence contains a single possible antecedent i, and the anaphor is the subject of the current sentence, then pick i as the antecedent: 11 correct, and 0 incorrect.

Reference : [https://aclweb.org/anthology/W97-1306 High Precision Coreference with Limited Knowledge and Linguistic Reference, Breck Baldwin]

== Work Plan ==
:: Would be good to line these tasks up with weeks in the program (including the community bonding period). —[[User:Firespeaker|Firespeaker]] ([[User talk:Firespeaker|talk]]) 18:48, 16 March 2019 (CET)


'''Community Bonding Period''' (May 6 - May 27)
* Understand the Apertium pipeline fully
* Modify and Understand individual files
* Get familiar with the files that I need to modify
* Formalise the problem, limit the scope of anaphora resolution (To Anaphora needed for MT)
* Flowchart of the proposed system and Pseudocode
* Study the EuroParl corpus and see which anaphors the method will be able to resolve on paper

'''Week 1''' (May 27)
* Automatic Annotation of anaphora for evaluation (EuroParl Corpus)
* Implement a preliminary scoring system for antecedent indicators [work for Spanish-English and Catalan-English for now]
* Decide on a definite context window

'''Week 2''' (June 3)
* Implement Basic Anaphora for Possessive Pronouns (Python)
* Implement transfer rules for Possessive Pronouns
* A basic prototype ready
* Write the code in C++ (All further coding to be done in C++)
* TEST the prototype with the pipeline

'''Week 3''' (June 10)

* Implement Basic Anaphora for Reflexive Pronouns (On Verbs) [For Spanish and Catalan]
* Implement Basic Anaphora for Zero Pronouns (On Verbs) [For Spanish and Catalan]
* Implement transfer rules for the above
* TEST the system extensively
* Document the outline

'''Week 4''' (June 17)
* Implement the system to work out all possible antecedents
* Add ability to give antecedents a score
* TEST basic sentences with single antecedents, Test the pipeline
* Test and Evaluate for Possessive, Reflexive, Zero Pronouns in Spa-Eng pair
* Test and Evaluate for Possessive, Reflexive, Zero Pronouns in Cat-Eng pair

=== Deliverable #1: Anaphora Resolution for single antecedents, with transfer rules [The full pipeline] ===

'''Evaluation 1: June 24-28'''

'''Week 5''' (June 28)
* Implement Antecedent Indicators:
* Code to Identify Boosting Indicators

'''Week 6''' (July 4)
* Code to Identify Impeding Indicators
* Code to Identify Adjective Agreement Antecedent
* Implement transfer rules for agreement in adjectives for Cat-Eng & Spa-Eng
* Code to remember antecedents for a certain window

'''Week 7''' (July 10)
* Implement remaining transfer rules for anaphora in pronouns Cat-Eng & Spa-Eng
* Give scores to the antecedent indicators
* Code Salience Indicators & Implement tie breaking systems
* Modify the scoring system based on performance in the pairs

'''Week 8''' (July 16)
* Implement fallback for anaphora (in case of too many antecedents or not past certainty threshold)
* TEST Scoring System
* TEST and Evaluate current system and produce precision and recall
* TEST and Evaluate Agreement for Adjectives
* Document Antecedent Indicators, Scoring System, Fallback for Cat-Eng & Spa-Eng

=== Deliverable #2: Anaphora Resolution with antecedent scores, fallback mechanism, for Spa-Eng and Cat-Eng ===

'''Evaluation 2: July 22-26'''

'''Week 9''' [OPTIONAL: If current system not producing good enough results]
* Implement Expectation-Maximization Algorithm using monolingual corpus
* Implement choosing anaphora with max probability: If Scoring System has a tie vs. As an independent system
* Compare and Evaluate the effectiveness of the above two possibilities
* Test EM Algorithm and the implemented system

'''Week 9''' [NOT OPTIONAL] (July 26)

* Insert into Apertium pipeline
* Implement code to accept input in chunks and process it
* Output with anaphora attached

'''Week 10''' (August 1)
* EXTENSIVELY TEST final system
* Try out other language pairs
* Evaluate and find out which features are language agnostic
* Decide on list of features for agnostic anaphora and for language specific anaphora

'''Week 11''' (August 7)
* TEST on multiple pairs and give Evaluation Scores
* TEST for backwards compatibility and ensure it

'''Week 12''' (August 13)
* Wrap up on the final module
* Complete the overall documentation with observations and future prospects

'''Final Evaluations: August 19-26'''

=== Project Completed ===
'''NOTE''': Week 11 and Week 12 have extra time to deal with unforeseen issues and ideas
----

== A description of how and who it will benefit in society ==

It will definitely benefit most users of Apertium and hopefully will attract more people to the tool. I’m from India and for a lot of our languages, we don’t have the data to create reliable Neural MT systems. Similarly, for all resource poor languages, Apertium provides an easy and reliable MT system for their needs. That’s how Apertium benefits society already.

However, what discourages people from using Machine Translation is unintelligible outputs and too much post editing which makes it very time consuming and costly for them. While Apertium aims to make minimal errors, as of now it selects a default male pronoun and that leads to several unintelligible outputs. Fixing that and making the system more fluent and intelligible overall will encourage people to use Machine Translation and will reduce costs of time and money.

== Reasons why Google and Apertium should sponsor it ==

I feel this project has a wide scope as it can affect almost all language pairs and helps almost everyone using Apertium. A decent Anaphora Resolution will give the output an important boost in its fluency and intelligibility, not just for one language, but all of them.

It’s a project which has promising future prospects as well - apart from the fact that language specific features can be added to improve it, we’ll be doing anaphora resolution even for languages which don’t need it to pick the correct pronoun. Doing this will enable Apertium to do Gisting Translation in the future, for which Anaphora Resolution is essential.

This method also paves the way for general coreference resolution for low-resource languages.


'''Hobbies:''' I enjoy playing the bass, singing, reading, and used to be super into Parliamentary Debating in uni.
With this project I aim to help the users of Apertium, I wish to become a regular contributor to Apertium and become equipped to do a lot more Open Source Development in the future for other organisations as well.


== Ideas ==
By funding this project, Google will help improve an important Open Source tool and promote Open Source Development. In a world of Proprietary softwares, this is an invaluable resource for society and supports innovation that everyone can benefit from.


'''Constructions in low resource MT:''' [[User:Khannatanmai/Constructions]]
== Skills and Qualifications ==
I'm currently a third year student at IIIT Hyderabad where I'm studying Computational Linguistics. It is a dual degree where we study Computer Science, Linguistics, NLP and more. I'm working on Machine Translation in the LTRC lab in IIIT Hyderabad and I'm part of the MT group in our university.


== Google Summer of Code 2019 -- Anaphora Resolution ==
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.


'''Proposal ''': [[User:Khannatanmai/GSoC2019Proposal]]
Due to the focused nature of our course, I have worked in several projects, such as building Translation Memory, Detecting Homographic Puns, POS Taggers, Grammar and Spell Checkers, Named Entity Recognisers, Building Chatbots, etc. all of which required a working understanding of Natural Language Processing.


'''Final Report''': [[User:Khannatanmai/GSoC2019Report]]
I am fluent in English, Hindi and have basic knowledge of Spanish.


== Google Summer of Code 2020 -- Markup handling with wordbound blanks ==
The details of my skills and work experience can be found here: [https://drive.google.com/file/d/1ZGAWJQzmDlxJo-_TNr-0wqzIdNTVraPt/view CV]


'''Proposal: Modifying the apertium stream format and eliminating dictionary trimming: [[User:Khannatanmai/GSoC2020Proposal_Trimming]]'''
=== Coding Challenge ===
I successfully completed the coding challenge and proposed an alternate method to process the input as a chunk which resulted in a speedup of more than 2x.


Development of the stream extension: [[User:Khannatanmai/New_Apertium_stream_format]]
The repo can be found at: https://github.com/khannatanmai/apertium


Eliminating Dictionary Trimming: [[User:Khannatanmai/Eliminating_Dictionary_Trimming]]
Files in Repo:
* Code to do basic anaphora resolution (last seen noun), input taken as a stream (byte by byte)
* Code to do basic anaphora resolution (last seen noun), input taken as a stream (as a chunk)
* Speed-Up Report


'''Progress: [[User:Khannatanmai/GSoC2020Progress]]'''
== Non-Summer-Of-Code Plans ==


Documentation of features related to secondary tags: [[User:Khannatanmai/Secondary_tags_features]]
I will have my college vacations during GSoC so will have no other commitments in that period and will be dedicated to GSoC full time, i.e. 40 hours a week.


Alternate stream modification proposal: [[User:Khannatanmai/Alternate_stream_modification]]
I am planning to go on a short trip to London from 30 May to 3rd June so I will start working for Week 1 earlier and finish it before the trip. I will still be able to devote 20 hours in Week 1.


Development of the updated stream extension: [[User:Khannatanmai/Secondary_info_apertium_stream_format]]
I have also kept the workload lighter in Week 1 for the same reason.


'''Development of wordbound blanks: [[User:Khannatanmai/Wordbound_blanks]]'''
:: In aligning your tasks and goals with the GSoC timeline, be sure to take this into account. —[[User:Firespeaker|Firespeaker]] ([[User talk:Firespeaker|talk]]) 18:49, 16 March 2019 (CET)


'''Documentation of wordbound blanks: [[Wordbound_blanks]]'''


'''Final Report: [[User:Khannatanmai/GSoC2020_Final_Report]]'''
[[Category:GSoC 2019 student proposals|Khannatanmai]]

Latest revision as of 04:27, 9 September 2020

Personal Details[edit]

Name: Tanmai Khanna

E-mail address: khanna.tanmai@gmail.com , tanmai.khanna@research.iiit.ac.in

IRC: khannatanmai

GitHub: khannatanmai

LinkedIn: khannatanmai

Current Designation: Graduate Researcher in the LTRC Lab, IIIT Hyderabad (5th year student) and a Teaching Assistant for Linguistics courses

Time Zone: GMT+5:30

About Me[edit]

Professional Interests: I’m currently doing research in Computational Linguistics and I have a particular interest in Linguistics and NLP tools, specifically Machine Translation and its components.

Hobbies: I enjoy playing the bass, singing, reading, and used to be super into Parliamentary Debating in uni.

Ideas[edit]

Constructions in low resource MT: User:Khannatanmai/Constructions

Google Summer of Code 2019 -- Anaphora Resolution[edit]

Proposal : User:Khannatanmai/GSoC2019Proposal

Final Report: User:Khannatanmai/GSoC2019Report

Google Summer of Code 2020 -- Markup handling with wordbound blanks[edit]

Proposal: Modifying the apertium stream format and eliminating dictionary trimming: User:Khannatanmai/GSoC2020Proposal_Trimming

Development of the stream extension: User:Khannatanmai/New_Apertium_stream_format

Eliminating Dictionary Trimming: User:Khannatanmai/Eliminating_Dictionary_Trimming

Progress: User:Khannatanmai/GSoC2020Progress

Documentation of features related to secondary tags: User:Khannatanmai/Secondary_tags_features

Alternate stream modification proposal: User:Khannatanmai/Alternate_stream_modification

Development of the updated stream extension: User:Khannatanmai/Secondary_info_apertium_stream_format

Development of wordbound blanks: User:Khannatanmai/Wordbound_blanks

Documentation of wordbound blanks: Wordbound_blanks

Final Report: User:Khannatanmai/GSoC2020_Final_Report

Retrieved from "https://wiki.apertium.org/w/index.php?title=User:Khannatanmai&oldid=72940"