Papers by Soumen Chakrabarti
Topic Sensitive Attention on Generic Corpora Corrects Sense Bias in Pretrained Embeddings (P19-1)
Copied to clipboard
| Challenge: | Existing methods to adapt pretrained embeddings to a large corpus are limited and do not provide sufficient quality. |
| Approach: | They propose to use a small corpus D_T to pretrain embeddings that accurately capture the sense of words in a limited set of focused topics. |
| Outcome: | The proposed embeddings capture the sense of words in a topic in spite of the limited size of the corpus D_T. |
TwiRGCN: Temporally Weighted Graph Convolution for Question Answering over Temporal Knowledge Graphs (2023.eacl-main)
Copied to clipboard
| Challenge: | Recent years have witnessed interest in Temporal Question Answering over Knowledge Graphs (TKGQA) but these methods are highly engineered and do not automatically discover relevant parts of the KG during multi-hop reasoning. |
| Approach: | They propose a scheme to modulate the messages passed through a KG edge during convolution based on the relevance of its associated period to the question. |
| Outcome: | The proposed system outperforms state-of-the-art models on a recent challenging dataset for multi-hop complex temporal QA called TimeQuestions. |
Type-Sensitive Knowledge Base Inference Without Explicit Type Supervision (P18-2)
Copied to clipboard
| Challenge: | State-of-the-art knowledge base completion models make frequent errors when ranking entities that are not compatible with the type required by the relation. |
| Approach: | They propose to enhance each base factorization with two type-compatibility terms between entity-relation pairs and combine the signals in a novel manner. |
| Outcome: | The proposed model achieves 7% MRR gains over baseline models and predicts supervised types better than baseline models. |
mOKB6: A Multilingual Open Knowledge Base Completion Benchmark (2023.acl-short)
Copied to clipboard
| Challenge: | Open knowledge bases (Open KBCs) are constructed from triples of the form, which can be denoted as (s, r, o) by using open information extraction (Open IE) systems. |
| Approach: | They construct a dataset with facts from Wikipedia in six languages . they use open information extraction systems to extract triples from text . |
| Outcome: | The proposed dataset contains facts from Wikipedia in six languages including English . it improves existing Open KB construction pipeline by doing multilingual coreference resolution and keeping only entity-linked triples . |
Cost-Performance Optimization for Processing Low-Resource Language Tasks Using Commercial LLMs (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) exhibit impressive zero/few-shot inference and generation quality for high-resource languages (HRLs). |
| Approach: | They propose to reduce the cost of processing LRLs by code-mixing, translation, and transliteration of LRL to HRLs to ensure that predictive and generative qualities are not compromised. |
| Outcome: | The proposed model reduces the cost of processing LRLs while ensuring that predictive and generative qualities are not compromised. |
AIT-QA: Question Answering Dataset over Complex Tables in the Airline Industry (2022.naacl-industry)
Copied to clipboard
Yannis Katsis, Saneem Chemmengath, Vishwajeet Kumar, Samarth Bharadwaj, Mustafa Canim, Michael Glass, Alfio Gliozzo, Feifei Pan, Jaydeep Sen, Karthik Sankaranarayanan, Soumen Chakrabarti
| Challenge: | Table Question Answering (Table QA) systems have been shown to be highly accurate when trained and tested on open-domain datasets built on top of Wikipedia tables. |
| Approach: | They propose a domain-specific Table QA test dataset to test Table Question Answering systems on open-domain datasets built on top of Wikipedia tables. |
| Outcome: | The proposed methods are highly accurate when tested on open-domain datasets built on top of Wikipedia tables. |
Multi-Row, Multi-Span Distant Supervision For Table+Text Question Answering (2023.acl-long)
Copied to clipboard
Vishwajeet Kumar, Yash Gupta, Saneem Chemmengath, Jaydeep Sen, Soumen Chakrabarti, Samarth Bharadwaj, Feifei Pan
| Challenge: | Existing question answering systems for tables and linked text are relatively unexplored. |
| Approach: | They propose a transformer-based question answering system that copes with distant supervision along both axes of the question and answer. |
| Outcome: | The proposed system beats baselines for HybridQA and OTT-QA with best EM and F1 scores on a held out test set. |
Diverse In-Context Example Selection After Decomposing Programs and Aligned Utterances Improves Semantic Parsing (2025.naacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are well suited for seq2seq translation . a lack of pretraining corpora can hinder the use of LLMs for structured interpretation . |
| Approach: | They propose to decompose available ICE trees into fragments and use additional invocations to map them to corresponding utterances. |
| Outcome: | The proposed method shows visible gains on diverse parsing benchmarks on popular languages. |
Dense Retrieval with Quantity Comparison Intent (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing sparse and dense retrieval systems fragment numerals and units that express quantities in arbitrary ways. |
| Approach: | They propose a dense retrieval system built around a density multi-vector index . they propose eliciting and exploiting quantities and associated comparison intents . |
| Outcome: | The proposed system is faster and more accurate than popular PLMs on two public and one proprietary e-commerce benchmarks. |
Temporal Knowledge Base Completion: New Algorithms and Evaluation Protocols (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing TKBC models heavily overestimate link prediction performance due to imperfect evaluation mechanisms. |
| Approach: | They propose a method that integrates entities, relations and time into a uniform space . they propose improved evaluation protocols for link and time prediction . |
| Outcome: | The proposed method exploits the recurrent nature of some facts/events and temporal interactions between pairs of relations yielding state-of-the-art results. |
CRUSH4SQL: Collective Retrieval Using Schema Hallucination For Text2SQL (2023.emnlp-main)
Copied to clipboard
| Challenge: | Existing Text-to-SQL generators require the entire schema to be encoded with the user text. |
| Approach: | They propose a method that uses an LLM to hallucinate a minimal DB schema . they use the hallucinated schema to retrieve a subset of the actual schema based on multiple dense retrievals . |
| Outcome: | The proposed method leads to significantly higher recall than existing methods. |
Topic Transferable Table Question Answering (2021.emnlp-main)
Copied to clipboard
Saneem Chemmengath, Vishwajeet Kumar, Samarth Bharadwaj, Jaydeep Sen, Mustafa Canim, Soumen Chakrabarti, Alfio Gliozzo, Karthik Sankaranarayanan
| Challenge: | Weakly-supervised table question-answering (TableQA) models have achieved state-of-art performance by using pre-trained BERT transformer to jointly encoding a question and a table to produce structured query for the question. |
| Approach: | They propose a framework for TableQA that incorporates topic-specific vocabulary injection into BERT, a novel text-to-text transformer generator and a logical form re-ranker. |
| Outcome: | The proposed framework provides a reasonably good baseline for topic shift benchmarks. |
Entropy-guided Vocabulary Augmentation of Multilingual Language Models for Low-resource Tasks (2023.findings-acl)
Copied to clipboard
| Challenge: | Multilingual language models (MLLMs) support low-resource languages (LRLs) but LRL words are under-represented in wordpiece/subword vocabularies, leading to low task accuracy . |
| Approach: | They propose an entropy-based vocabulary augmented language model to detect LRL words with undesirable wordpiece segmentations. |
| Outcome: | The proposed model improves performance and limits on wordpiece augmentation strategies for multiple diverse LRLs. |
Joint Completion and Alignment of Multilingual Knowledge Graphs (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for knowledge graph completion are incomplete, as curators struggle to keep up with the real world. |
| Approach: | They propose a multitask approach to solve missing facts in incomplete Knowledge Graphs . they add a relation representation to the existing KG embedding scheme . |
| Outcome: | The proposed system outperforms existing models in seven languages compared to existing models . it also outperformed existing models, underscoring the value of joint alignment and completion. |
Question Answering Over Temporal Knowledge Graphs (2021.acl-long)
Copied to clipboard
| Challenge: | Temporal Knowledge Graphs (Temporal KGs) provide temporal scopes (start and end times) on each edge in the Knowledge . Lack of broad coverage datasets has been limiting progress in this area . |
| Approach: | They propose a transformer-based solution that exploits recent advances in Temporal Knowledge Graph embeddings and achieves an increase of 120% in accuracy over the next best performing method. |
| Outcome: | The proposed solution improves on the only known dataset by 340x . it increases accuracy by 120% over the baseline solution . |
Improved Sentiment Detection via Label Transfer from Monolingual to Synthetic Code-Switched Text (P19-1)
Copied to clipboard
| Challenge: | Existing sentiment detection methods are trained on sentiment-labeled monolingual text. |
| Approach: | They propose a method for synthesizing labeled code-switched text from monolingual text. |
| Outcome: | The proposed method improves sentiment labeling accuracy for three languages. |
MutantPrompt: Prompt Optimization via Mutation Under a Budget on Modest-sized LMs (2025.findings-acl)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have revolutionized the way we learn and process information, but identifying optimal prompts remains a challenge for low-resource languages. |
| Approach: | They propose a framework that leverages multi-armed bandit algorithms to efficiently identify optimal prompts tailored to low-resource languages. |
| Outcome: | The proposed framework is able to find optimal prompts for low-resource languages and significantly improves performance across multiple low-level tasks. |
Small Language Models Fine-tuned to Coordinate Larger Language Models improve Complex Reasoning (2023.emnlp-main)
Copied to clipboard
| Challenge: | Recent attempts at prompt decomposition toward solving complex, multi-step reasoning problems depend on the ability of the LLM to simultaneously decompose and solve the problem. |
| Approach: | They propose a decomposition generator that decomposes complex problems into subproblems that require fewer reasoning steps. |
| Outcome: | The proposed method can produce competitive or even better performance compared to its larger successor, GPT-4. |
Alignment-Augmented Consistent Translation for Multilingual Open Information Extraction (2022.acl-long)
Copied to clipboard
| Challenge: | Existing training data in English is limited to supervised Open Information Extraction (OpenIE) . |
| Approach: | They propose a model that automatically converts English sentences into other languages by using Alignment-Augmented Constrained Translation (AACTrans) they train a generative OpenIE model that outputs for each sentence relations in the first stage and all extractions containing the relation in the second stage. |
| Outcome: | The proposed model outperforms existing models on Spanish, Portuguese, Chinese, Hindi and Telugu on 5 languages. |
Efficient Continual Pre-training of LLMs for Low-resource Languages (2025.naacl-industry)
Copied to clipboard
| Challenge: | Open-source large language models (LLMs) are a promising tool for low-resource languages . however, there is still a substantial performance gap between high-resourced languages and LRLs . |
| Approach: | They develop an algorithm to select a subset of texts from a larger corpus and use it to select tokens for LLMs. |
| Outcome: | The proposed algorithm reduces the cost of continual pre-training (CPT) with large amounts of language-specific data. |
OpenIE6: Iterative Grid Labeling and Coordination Analysis for Open Information Extraction (2020.emnlp-main)
Copied to clipboard
| Challenge: | OpenIE generates extractions iteratively, requiring repeated encoding of partial outputs. |
| Approach: | They propose an iterative open information extraction system that generates extractions iterativly, requiring repeated encoding of partial outputs. |
| Outcome: | The proposed system beats the previous systems by as much as 4 pts in F1 while being much faster. |
NLP Service APIs and Models for Efficient Registration of New Clients (2020.findings-emnlp)
Copied to clipboard
| Challenge: | State-of-the-art NLP inference uses enormous neural architectures and models trained for GPU-months, well beyond the reach of most consumers of NLP. |
| Approach: | They propose a centralized NLP service that can be customized to suit clients . they propose NER, sentiment labeling, and predictive language modeling to improve client experience. |
| Outcome: | The proposed model can be used to improve word usage and salience across clients without re-training or fine-tuning. |
IMoJIE: Iterative Memory-Based Joint Open Information Extraction (2020.acl-main)
Copied to clipboard
| Challenge: | Recent neural OpenIE systems are statistical or rule-based for Open Information Extraction. |
| Approach: | They propose an extension to CopyAttention that produces the next extraction conditioned on all previously extracted tuples. |
| Outcome: | The proposed model outperforms CopyAttention by 18 pts and a BERT-based strong baseline by 2 ptes. |