Search Results for author: Jugal Kalita

Found 50 papers, 7 papers with code

Generalization to Mitigate Synonym Substitution Attacks

no code implementations EMNLP (DeeLIO) 2020 Basemah Alshemali, Jugal Kalita

Our findings show that replacing the embeddings of the important words in the input samples with the average of their synonyms’ embeddings can significantly improve the generalization of DNN-based classifiers.

Adversarial Attack

Solving Arithmetic Word Problems Using Transformer and Pre-processing of Problem Texts

no code implementations ICON 2020 Kaden Griffith, Jugal Kalita

This paper outlines the use of Transformer networks trained to translate math word problems to equivalent arithmetic expressions in infix, prefix, and postfix notations.

Explaining Math Word Problem Solvers

no code implementations24 Jul 2023 Abby Newcomb, Jugal Kalita

Automated math word problem solvers based on neural networks have successfully managed to obtain 70-80\% accuracy in solving arithmetic word problems.

Training-free Neural Architecture Search for RNNs and Transformers

1 code implementation1 Jun 2023 Aaron Serianni, Jugal Kalita

Ultimately, our analysis shows that the architecture search space and the training-free metric must be developed together in order to achieve effective results.

Image Classification Language Modelling +1

Enhancing Translation for Indigenous Languages: Experiments with Multilingual Models

no code implementations27 May 2023 Atnafu Lambebo Tonja, Hellina Hailu Nigatu, Olga Kolesnikova, Grigori Sidorov, Alexander Gelbukh, Jugal Kalita

This paper describes CIC NLP's submission to the AmericasNLP 2023 Shared Task on machine translation systems for indigenous languages of the Americas.

Machine Translation Transfer Learning +1

Abstractive Text Summarization Using the BRIO Training Paradigm

no code implementations23 May 2023 Khang Nhut Lam, Thieu Gia Doan, Khang Thua Pham, Jugal Kalita

This paper presents a straightforward but effective technique to improve abstractive summaries by fine-tuning pre-trained language models, and training them with the BRIO paradigm.

Abstractive Text Summarization

Spatiotemporal Transformer for Stock Movement Prediction

no code implementations5 May 2023 Daniel Boyle, Jugal Kalita

Financial markets are an intriguing place that offer investors the potential to gain large profits if timed correctly.

Utilizing Priming to Identify Optimal Class Ordering to Alleviate Catastrophic Forgetting

no code implementations24 Dec 2022 Gabriel Mantione-Holmes, Justin Leo, Jugal Kalita

While efforts are being made to quell catastrophic forgetting, there is a lack of research that looks into the importance of class ordering when training on new classes for incremental learning.

Incremental Learning

CAMeMBERT: Cascading Assistant-Mediated Multilingual BERT

no code implementations22 Dec 2022 Dan DeGenaro, Jugal Kalita

Large language models having hundreds of millions, and even billions, of parameters have performed extremely well on a variety of natural language processing (NLP) tasks.

Knowledge Distillation

Building a Chatbot on a Closed Domain using RASA

no code implementations12 Aug 2022 Khang Nhut Lam, Nam Nhat Le, Jugal Kalita

In this study, we build a chatbot system in a closed domain with the RASA framework, using several models such as SVM for classifying intents, CRF for extracting entities and LSTM for predicting action.


Automatically Creating a Large Number of New Bilingual Dictionaries

no code implementations12 Aug 2022 Khang Nhut Lam, Feras Al Tarouti, Jugal Kalita

This paper proposes approaches to automatically create a large number of new bilingual dictionaries for low-resource languages, especially resource-poor and endangered languages, from a single input bilingual dictionary.

Creating Reverse Bilingual Dictionaries

no code implementations NAACL 2013 Khang Nhut Lam, Jugal Kalita

Bilingual dictionaries are expensive resources and not many are available when one of the languages is resource-poor.

Reverse Dictionary

Creating Lexical Resources for Endangered Languages

no code implementations WS 2014 Khang Nhut Lam, Feras Al Tarouti, Jugal Kalita

This paper examines approaches to generate lexical resources for endangered languages.

Automatically constructing Wordnet synsets

no code implementations ACL 2014 Khang Nhut Lam, Feras Al Tarouti, Jugal Kalita

Manually constructing a Wordnet is a difficult task, needing years of experts' time.


Phrase translation using a bilingual dictionary and n-gram data: A case study from Vietnamese to English

no code implementations WS 2015 Khang Nhut Lam, Feras Al Tarouti, Jugal Kalita

Past approaches to translate a phrase in a language L1 to a language L2 using a dictionary-based approach require grammar rules to restructure initial translations.

Towards Multimodal Vision-Language Models Generating Non-Generic Text

no code implementations ICON 2021 Wes Robbins, Zanyar Zohourianshahzadi, Jugal Kalita

To address this, recent work has used optical character recognition to supplement visual information with text extracted from an image.

Descriptive Optical Character Recognition +1

ZoDIAC: Zoneout Dropout Injection Attention Calculation

1 code implementation28 Jun 2022 Zanyar Zohourianshahzadi, Jugal Kalita

Recently the use of self-attention has yielded to state-of-the-art results in vision-language tasks such as image captioning as well as natural language understanding and generation (NLU and NLG) tasks and computer vision tasks such as image classification.

Image Captioning Image Classification +1

Incremental Deep Neural Network Learning using Classification Confidence Thresholding

no code implementations21 Jun 2021 Justin Leo, Jugal Kalita

This in turn leads to the concept of incremental learning where a model with its own architecture and initial trained set of data can identify unknown classes during the testing phase and autonomously update itself if evidence of a new class is detected.

Classification Incremental Learning

Improving Computer Generated Dialog with Auxiliary Loss Functions and Custom Evaluation Metrics

no code implementations4 Jun 2021 Thomas Conley, Jack St. Clair, Jugal Kalita

This research joins the quest by creating a dialog generating Recurrent Neural Network (RNN) and by enhancing the ability of this network with auxiliary loss functions and a beam search.

Solving Arithmetic Word Problems with Transformers and Preprocessing of Problem Text

no code implementations2 Jun 2021 Kaden Griffith, Jugal Kalita

This paper outlines the use of Transformer networks trained to translate math word problems to equivalent arithmetic expressions in infix, prefix, and postfix notations.

Classifying Malware Images with Convolutional Neural Network Models

no code implementations30 Oct 2020 Ahmed Bensaoud, Nawaf Abudawaood, Jugal Kalita

In this paper, we use several convolutional neural network (CNN) models for static malware classification.

Classification General Classification +2

Multi-task learning for natural language processing in the 2020s: where are we going?

no code implementations22 Jul 2020 Joseph Worsham, Jugal Kalita

Multi-task learning (MTL) significantly pre-dates the deep learning era, and it has seen a resurgence in the past few years as researchers have been applying MTL to deep learning solutions for natural language tasks.

Multi-Task Learning

Abstractive and mixed summarization for long-single documents

no code implementations3 Jul 2020 Roger Barrull, Jugal Kalita

The lack of diversity in the datasets available for automatic summarization of documents has meant that the vast majority of neural models for automatic summarization have been trained with news articles.

Introducing Aspects of Creativity in Automatic Poetry Generation

1 code implementation ICON 2019 Brendan Bena, Jugal Kalita

Specifically, we generate poems that express emotion and elicit the same in readers, and poems that use the language of dreams---called dream poetry.

Language Modelling

Adversarial Analysis of Natural Language Inference Systems

no code implementations7 Dec 2019 Tiffany Chien, Jugal Kalita

So it is no surprise that many adversarial (challenge) datasets have been created that cause models trained on standard datasets to fail dramatically.

Multi-Task Learning Natural Language Inference

Solving Arithmetic Word Problems Automatically Using Transformer and Unambiguous Representations

1 code implementation2 Dec 2019 Kaden Griffith, Jugal Kalita

This paper outlines the use of Transformer networks trained to translate math word problems to equivalent arithmetic expressions in infix, prefix, and postfix notations.

Moving Towards Open Set Incremental Learning: Readily Discovering New Authors

no code implementations28 Oct 2019 Justin Leo, Jugal Kalita

This paper also develops a new metric that measures multiple attributes of clustering open set data.

Author Attribution Clustering +2

Toward Robust Image Classification

no code implementations19 Sep 2019 Basemah Alshemali, Alta Graham, Jugal Kalita

Neural networks are frequently used for image classification, but can be vulnerable to misclassification caused by adversarial images.

Classification General Classification +1

Noise Flooding for Detecting Audio Adversarial Examples Against Automatic Speech Recognition

no code implementations25 Dec 2018 Krishan Rajaratnam, Jugal Kalita

Limited-vocabulary speech classifiers, such as the Speech Commands model, are used quite frequently in a variety of applications, particularly in managing automated attendants in telephony contexts.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

Parallel Attention Mechanisms in Neural Machine Translation

no code implementations29 Oct 2018 Julian Richard Medina, Jugal Kalita

Recent papers in neural machine translation have proposed the strict use of attention mechanisms over previous standards such as recurrent and convolutional neural networks (RNNs and CNNs).

Machine Translation Translation

Abstractive Summarization Using Attentive Neural Techniques

2 code implementations20 Oct 2018 Jacob Krantz, Jugal Kalita

However, we show that these metrics are limited in their ability to effectively score abstractive summaries, and propose a new approach based on the intuition that an abstractive model requires an abstractive evaluation.

Abstractive Text Summarization Machine Translation +1

Exploring Sentence Vector Spaces through Automatic Summarization

no code implementations16 Oct 2018 Adly Templeton, Jugal Kalita

To our knowledge, this is the first time specific dimensions of sentence embeddings have been connected to sentence properties.

Sentence Embeddings

Exploring Sentence Vectors Through Automatic Summarization

no code implementations ICLR 2018 Adly Templeton, Jugal Kalita

Vector semantics, especially sentence vectors, have recently been used successfully in many areas of natural language processing.

Sentence Embeddings

Deep Learning applied to NLP

2 code implementations9 Mar 2017 Marc Moreno Lopez, Jugal Kalita

Convolutional Neural Network (CNNs) are typically associated with Computer Vision.

General Classification Image Classification

Detecting and Extracting Events from Text Documents

no code implementations15 Jan 2016 Jugal Kalita

Events of various kinds are mentioned and discussed in text documents, whether they are books, news articles, blogs or microblog feeds.

Event Detection Philosophy

Cannot find the paper you are looking for? You can Submit a new open access paper.