Maunendra Sankar Desarkar
YOU?
Author Swipe
View article: DIWALI -- Diversity and Inclusivity aWare cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context
DIWALI -- Diversity and Inclusivity aWare cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context Open
Large language models (LLMs) are widely used in various tasks and applications. However, despite their wide capabilities, they are shown to lack cultural alignment \citep{ryan-etal-2024-unintended, alkhamissi-etal-2024-investigating} and p…
View article: MorphTok: Morphologically Grounded Tokenization for Indian Languages
MorphTok: Morphologically Grounded Tokenization for Indian Languages Open
Tokenization is a crucial step in NLP, especially with the rise of large language models (LLMs), impacting downstream performance, computational cost, and efficiency. Existing LLMs rely on the classical Byte-pair Encoding (BPE) algorithm f…
View article: NLIP at BEA 2025 Shared Task: Evaluation of Pedagogical Ability of AI Tutors
NLIP at BEA 2025 Shared Task: Evaluation of Pedagogical Ability of AI Tutors Open
View article: DIWALI - Diversity and Inclusivity aWare cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context
DIWALI - Diversity and Inclusivity aWare cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context Open
View article: NLIP_Lab-IITH Multilingual MT System for WAT24 MT Shared Task
NLIP_Lab-IITH Multilingual MT System for WAT24 MT Shared Task Open
This paper describes NLIP Lab's multilingual machine translation system for the WAT24 shared task on multilingual Indic MT task for 22 scheduled languages belonging to 4 language families. We explore pre-training for Indic languages using …
View article: NLIP_Lab-IITH Low-Resource MT System for WMT24 Indic MT Shared Task
NLIP_Lab-IITH Low-Resource MT System for WMT24 Indic MT Shared Task Open
In this paper, we describe our system for the WMT 24 shared task of Low-Resource Indic Language Translation. We consider eng $\leftrightarrow$ {as, kha, lus, mni} as participating language pairs. In this shared task, we explore the finetun…
View article: DAC: Quantized Optimal Transport Reward-based Reinforcement Learning Approach to Detoxify Query Auto-Completion
DAC: Quantized Optimal Transport Reward-based Reinforcement Learning Approach to Detoxify Query Auto-Completion Open
View article: Transformer based Multitask Learning for Image Captioning and Object Detection
Transformer based Multitask Learning for Image Captioning and Object Detection Open
In several real-world scenarios like autonomous navigation and mobility, to obtain a better visual understanding of the surroundings, image captioning and object detection play a crucial role. This work introduces a novel multitask learnin…
View article: BoK: Introducing Bag-of-Keywords Loss for Interpretable Dialogue Response Generation
BoK: Introducing Bag-of-Keywords Loss for Interpretable Dialogue Response Generation Open
The standard language modeling (LM) loss by itself has been shown to be inadequate for effective dialogue modeling. As a result, various training approaches, such as auxiliary loss functions and leveraging human feedback, are being adopted…
View article: CharSpan: Utilizing Lexical Similarity to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages
CharSpan: Utilizing Lexical Similarity to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages Open
View article: trie-nlg: trie context augmentation to improve personalized query auto-completion for short and unseen prefixes
trie-nlg: trie context augmentation to improve personalized query auto-completion for short and unseen prefixes Open
View article: Trie-NLG: Trie Context Augmentation to Improve Personalized Query Auto-Completion for Short and Unseen Prefixes
Trie-NLG: Trie Context Augmentation to Improve Personalized Query Auto-Completion for Short and Unseen Prefixes Open
Query auto-completion (QAC) aims to suggest plausible completions for a given query prefix. Traditionally, QAC systems have leveraged tries curated from historical query logs to suggest most popular completions. In this context, there are …
View article: Towards Improvement of Grounded Cross-lingual Natural Language Inference with VisioTextual Attention
Towards Improvement of Grounded Cross-lingual Natural Language Inference with VisioTextual Attention Open
Natural Language Inference (NLI) has been one of the fundamental tasks in Natural Language Processing (NLP). Recognizing Textual Entailment (RTE) between the two pieces of text is a crucial problem. It adds further challenges when it invol…
View article: CharSpan: Utilizing Lexical Similarity to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages
CharSpan: Utilizing Lexical Similarity to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages Open
We address the task of machine translation (MT) from extremely low-resource language (ELRL) to English by leveraging cross-lingual transfer from 'closely-related' high-resource language (HRL). The development of an MT system for ELRL is ch…
View article: SelectNoise: Unsupervised Noise Injection to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages
SelectNoise: Unsupervised Noise Injection to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages Open
In this work, we focus on the task of machine translation (MT) from extremely low-resource language (ELRLs) to English. The unavailability of parallel data, lack of representation from large multilingual pre-trained models, and limited mon…
View article: Towards Low-resource Language Generation with Limited Supervision
Towards Low-resource Language Generation with Limited Supervision Open
We present a research narrative aimed at enabling language technology for multiple natural language generation (NLG) tasks in low-resource languages (LRLs). With approximately 7,000 languages spoken globally, many lack the resources requir…
View article: On Text Style Transfer via Style-Aware Masked Language Models
On Text Style Transfer via Style-Aware Masked Language Models Open
Text Style Transfer (TST) is performable through approaches such as latent space disentanglement, cycle-consistency losses, prototype editing etc. The prototype editing approach, which is known to be quite successful in TST, involves two k…
View article: DivHSK: Diverse Headline Generation using Self-Attention based Keyword Selection
DivHSK: Diverse Headline Generation using Self-Attention based Keyword Selection Open
Diverse headline generation is an NLP task where given a news article, the goal is to generate multiple headlines that are true to the content of the article but are different among themselves. This task aims to exhibit and exploit semanti…
View article: Dial-M: A Masking-based Framework for Dialogue Evaluation
Dial-M: A Masking-based Framework for Dialogue Evaluation Open
In dialogue systems, automatically evaluating machine-generated responses is critical and challenging. Despite the tremendous progress in dialogue generation research, its evaluation heavily depends on human judgments. The standard word-ov…
View article: ComplAI: Theory of A Unified Framework for Multi-factor Assessment of Black-Box Supervised Machine Learning Models
ComplAI: Theory of A Unified Framework for Multi-factor Assessment of Black-Box Supervised Machine Learning Models Open
The advances in Artificial Intelligence are creating new opportunities to improve lives of people around the world, from business to healthcare, from lifestyle to education. For example, some systems profile the users using their demograph…
View article: On Text Style Transfer via Style Masked Language Models
On Text Style Transfer via Style Masked Language Models Open
Text Style Transfer (TST) is performable through approaches such as latent space disentanglement, cycle-consistency losses, prototype editing etc. The prototype editing approach, which is known to be quite successful in TST, involves two k…
View article: DialoGen: Generalized Long-Range Context Representation for Dialogue Systems
DialoGen: Generalized Long-Range Context Representation for Dialogue Systems Open
Long-range context modeling is crucial to both dialogue understanding and generation. The most popular method for dialogue context representation is to concatenate the last-$k$ utterances in chronological order. However, this method may no…
View article: HyperHawkes: Hypernetwork based Neural Temporal Point Process
HyperHawkes: Hypernetwork based Neural Temporal Point Process Open
Temporal point process serves as an essential tool for modeling time-to-event data in continuous time space. Despite having massive amounts of event sequence data from various domains like social media, healthcare etc., real world applicat…
View article: Towards Robust and Semantically Organised Latent Representations for Unsupervised Text Style Transfer
Towards Robust and Semantically Organised Latent Representations for Unsupervised Text Style Transfer Open
Recent studies show that auto-encoder based approaches successfully perform language generation, smooth sentence interpolation, and style transfer over unseen attributes using unlabelled datasets in a zero-shot manner. The latent space geo…
View article: Effective utilization of labeled data from related tasks using graph contrastive pretraining
Effective utilization of labeled data from related tasks using graph contrastive pretraining Open
Contrastive pretraining techniques for text classification has been largely studied in an unsupervised setting. However, oftentimes labeled data from related past datasets which share label semantics with current task is available. We hypo…
View article: Multi-Context Based Neural Approach for COVID-19 Fake-News Detection
Multi-Context Based Neural Approach for COVID-19 Fake-News Detection Open
When the world is facing the COVID-19 pandemic, society is also fighting another battle to tackle misinformation. Due to the widespread effect of COVID 19 and increased usage of social media, fake news and rumors about COVID-19 are being s…
View article: Towards Fair Evaluation of Dialogue State Tracking by Flexible Incorporation of Turn-level Performances
Towards Fair Evaluation of Dialogue State Tracking by Flexible Incorporation of Turn-level Performances Open
Dialogue State Tracking (DST) is primarily evaluated using Joint Goal Accuracy (JGA) defined as the fraction of turns where the ground-truth dialogue state exactly matches the prediction. Generally in DST, the dialogue state or belief stat…
View article: Meta-X$_{NLG}$: A Meta-Learning Approach Based on Language Clustering for Zero-Shot Cross-Lingual Transfer and Generation
Meta-X$_{NLG}$: A Meta-Learning Approach Based on Language Clustering for Zero-Shot Cross-Lingual Transfer and Generation Open
Recently, the NLP community has witnessed a rapid advancement in multilingual and cross-lingual transfer research where the supervision is transferred from high-resource languages (HRLs) to low-resource languages (LRLs). However, the cross…
View article: Graph Neural Network Enhanced Language Models for Efficient Multilingual Text Classification
Graph Neural Network Enhanced Language Models for Efficient Multilingual Text Classification Open
Online social media works as a source of various valuable and actionable information during disasters. These information might be available in multiple languages due to the nature of user generated content. An effective system to automatic…
View article: Towards Robust and Semantically Organised Latent Representations for Unsupervised Text Style Transfer
Towards Robust and Semantically Organised Latent Representations for Unsupervised Text Style Transfer Open
Sharan Narasimhan, Suvodip Dey, Maunendra Desarkar. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2022.