Papers by Dipti Sharma

8 papers
Towards Speech to Speech Machine Translation focusing on Indian Languages (2023.eacl-demo)

Copied to clipboard

Challenge: SSMT is a web application for translating videos from one language to another by cascading multiple language modules.
Approach: They introduce an SSMT pipeline for translating videos from one language to another by cascading multiple language modules.
Outcome: The proposed system can get 3.5+ MOS score for English to Hindi using human intervention.
Towards Automated Semantic Role Labelling of Hindi-English Code-Mixed Tweets (D19-55)

Copied to clipboard

Challenge: a new system for semantic role labelling of Hindi-English code-mixed tweets is proposed . code-mixing is a largely observed phenomenon in colloquial usage and on social media .
Approach: They propose a system for automating Semantic Role Labelling of Hindi-English code-mixed tweets.
Outcome: The proposed system gives an overall accuracy of 84% for Argument Classification, a 10% increase over the existing rule-based model.
Fine-tuning Pre-trained Named Entity Recognition Models For Indian Languages (2024.naacl-srw)

Copied to clipboard

Challenge: Named Entity Recognition (NER) is a useful component in NLP applications.
Approach: They propose to use annotated named entity corpora to classify a given entity into a category within a textual document.
Outcome: The proposed model achieves an F1 score of 0.80 on an unseen dataset for Indian languages.
The LTRC Hindi-Telugu Parallel Corpus (2022.lrec-1)

Copied to clipboard

Challenge: a qualitative corpus of 700K parallel sentences was created using multiple methods such as extract, align and review of Hindi-Telugu corpora.
Approach: They propose to create a Hindi-Telugu parallel corpus of different technical domains using different methods including extract, align and review.
Outcome: The proposed corpus is the largest, publicly available domain parallel corpus for Hindi-Telugu.
Universal Dependency Parsing for Hindi-English Code-Switching (N18-1)

Copied to clipboard

Challenge: Code-switching data often need additional processes such as language identification, normalization and/or back-transliteration to be processed.
Approach: They propose a neural stacking model that leverages part-of-speech tags and syntactic tree annotations in tweets to parse code-switching data.
Outcome: The proposed model is 1.5% better than the augmented model and 3.8% better than one which uses first-best normalization and/or back-transliteration.
HAWP: a Dataset for Hindi Arithmetic Word Problem Solving (2022.lrec-1)

Copied to clipboard

Challenge: Word problem solving is a challenging and interesting task in NLP.
Approach: They propose to use equations to solve Hindi arithmetic word problems . they propose to also use equation equivalence to evaluate word problem solvers .
Outcome: The proposed dataset is based on 2336 arithmetic word problems in Hindi . it also includes baseline systems and evaluation techniques .
No more beating about the bush : A Step towards Idiom Handling for Indian Language NLP (L18-1)

Copied to clipboard

Challenge: idioms are a part of natural language and are difficult to learn with a parallel corpora database.
Approach: They propose to use a parallel idiom dataset to train two NLP subtasks . they show significant improvement in the two subtask training without the idiomatic dataset .
Outcome: The proposed model improves on baseline models with the idiom dataset for two NLP applications.
A Simple and Effective Dependency Parser for Telugu (2020.acl-srw)

Copied to clipboard

Challenge: Existing dependency parsers for Telugu use hand-crafted features based on linguistic information like part-of-speech and morphology which are expensive to annotate.
Approach: They propose to replace linguistic feature templates with a minimal feature function for Telugu . they train a BERT model on the Telugus Wikipedia data and use contextual vector representations to train the parser.
Outcome: The proposed parser achieves state-of-the-art for Telugu using contextual vector representations . the proposed model trains on the Telugus Wikipedia data and trains with a greedy transition based approach .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations