Papers by Mona Diab

DeSePtion: Dual Sequence Prediction and Adversarial Examples for Improved Fact-Checking (2020.acl-main)

Challenge:	Modern NLP systems rely on word embeddings as input units to encode statistical semantic and syntactic properties of words.
Approach:	They propose to use discrete cosine transform to compress word sequences in order-preserving manner.
Outcome:	The proposed model preserves syntactic information in semantic probing tasks . it is comparable to vector averaging but mediocre in performance.

Copied to clipboard

Christopher Hidey, Tuhin Chakrabarty, Tariq Alhindi, Siddharth Varia, Kriste Krstovski, Mona Diab, Smaranda Muresan

Efficient Large Scale Language Modeling with Mixtures of Experts (2022.emnlp-main)

Challenge:	Fact Extraction and Verification datasets provide a resource for end-to-end fact-checking, requiring retrieval of evidence from Wikipedia to validate a veracity prediction.
Approach:	They propose a system that is resilient to attacks by multiple propositions, temporal reasoning, ambiguity and lexical variation and a sequence of evidence sentences and veracity relation predictions.
Outcome:	The proposed system is resilient to three realistic “attacks” and obtains state-of-the-art results due to improved evidence retrieval.

Copied to clipboard

Mikel Artetxe, Shruti Bhosale, Naman Goyal, Todor Mihaylov, Myle Ott, Sam Shleifer, Xi Victoria Lin, Jingfei Du, Srinivasan Iyer, Ramakanth Pasunuru, Giridharan Anantharaman, Xian Li, Shuohui Chen, Halil Akin, Mandeep Baines, Louis Martin, Xing Zhou, Punit Singh Koura, Brian O’Horo, Jeffrey Wang, Luke Zettlemoyer, Mona Diab, Zornitsa Kozareva, Veselin Stoyanov

Few-shot Learning with Multilingual Generative Language Models (2022.emnlp-main)

Challenge:	Mixture of Experts layers (MoEs) enable efficient scaling of language models . large autoregressive language models such as GPT-3 can be adapted to a wide range of tasks .
Approach:	They propose to use Mixture of Experts layers to enable efficient scaling of language models . they find that MoEs are substantially more compute efficient than dense models compared to MoE models - but only when they are more modestly trained .
Outcome:	The proposed model outperforms dense models in a wide range of tasks and domains.

Copied to clipboard

Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O’Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li

Automatic Generation of Model and Data Cards: A Step Towards Responsible AI (2024.naacl-long)

Challenge:	Large-scale generative language models such as GPT-3 are competitive few-shot learners.
Approach:	They train multilingual generative language models on a corpus covering a diverse set of languages and study their few- and zero-shot learning capabilities.
Outcome:	The proposed model outperforms GPT-3 on 171 out of 182 directions with 32 training examples and surpasses the official supervised baseline in 45 directions.

Copied to clipboard

Jiarui Liu, Wenkai Li, Zhijing Jin, Mona Diab

Multi-Domain Goal-Oriented Dialogues (MultiDoGO): Strategies toward Curating and Annotating Large Scale Dialogue Data (D19-1)

Challenge:	Existing models and datasets are incomplete and lack consistent documentation.
Approach:	They propose an automated generation approach using Large Language Models (LLMs) their paper establishes a comprehensive dataset and develops 'CardGen' pipeline .
Outcome:	The proposed approach exhibits enhanced completeness, objectivity, and faithfulness in generated model and data cards, a significant step in responsible AI documentation practices ensuring better accountability and traceability.

Copied to clipboard

Denis Peskov, Nancy Clarke, Jason Krone, Brigi Fodor, Yi Zhang, Adel Youssef, Mona Diab

Dialogue Systems & Conversational Agents

Challenge:	a large number of goal-oriented dialogue datasets are limited in their size, linguistic diversity, domain coverage, or annotation granularity.
Approach:	They propose a multi-domain goal-oriented dialogue dataset that uses a crowd-sourced worker and a trained annotator to curate and annotate large scale data.
Outcome:	The proposed dataset is 8 times the size of the largest comparable dialogue dataset available to the public.

Evaluating Large Language Model Biases in Persona-Steered Generation (2024.findings-acl)

Copied to clipboard

Andy Liu, Mona Diab, Daniel Fried

GRASS: Compute Efficient Low-Memory LLM Training with Structured Sparse Gradients (2024.emnlp-main)

Challenge:	a recent wave of powerful new large language models has raised concerns that their expressed opinions may be biased towards certain political, national or moral viewpoints.
Approach:	They define an incongruous persona as a persona with multiple traits where one trait makes its other traits less likely in human survey data.
Outcome:	The results show that LLMs are less steerable towards incongruous personas than congruous ones . the models that are fine-tuned with RLHF are more steerable, especially towards stances associated with political liberals and women .

Copied to clipboard

Aashiq Muhamed, Oscar Li, David Woodruff, Mona Diab, Virginia Smith

A Multitask Learning Approach for Diacritic Restoration (2020.acl-main)

Challenge:	Existing projection-based methods that project gradients into a lower-dimensional subspace can introduce computational and memory overheads.
Approach:	They propose a novel approach that leverages sparse projections to transform gradients into structured sparser updates.
Outcome:	The proposed approach significantly reduces memory usage for optimizer states and minimizes memory footprint, computation, and communication costs, leading to substantial throughput improvements.

Copied to clipboard

Sawsan Alqahtani, Ajay Mishra, Mona Diab

Challenge:	Diacritics are used to specify pronunciations and meanings in many languages like Arabic.
Approach:	They propose to use multi-task learning to optimize diacritic restoration with related NLP problems . they use Arabic as a case study since it has sufficient data resources for tasks .
Outcome:	The proposed model outperforms baseline models and is comparable to the state-of-the-art models.

ALERT: Adapt Language Models to Reasoning Tasks (2023.acl-long)

Copied to clipboard

Ping Yu, Tianlu Wang, Olga Golovneva, Badr AlKhamissi, Siddharth Verma, Zhijing Jin, Gargi Ghosh, Mona Diab, Asli Celikyilmaz

Emotion Detection and Classification in a Multigenre Corpus with Joint Multi-Task Deep Learning (C18-1)

Challenge:	Large language models have shown increasing in-context learning capabilities with scaling up the model and data sizes.
Approach:	They propose a benchmark and suite of analyses to evaluate reasoning skills of large language models.
Outcome:	The proposed model compares pre-trained and fine-tuned models on tasks that require reasoning skills to solve.

Copied to clipboard

Shabnam Tafreshi, Mona Diab

Adapting High-resource NMT Models to Translate Low-resource Related Languages without Parallel Data (2021.acl-long)

Challenge:	Sentence-level emotion detection is a challenging task due to subjectivity of emotion.
Approach:	They propose a model to address genre robustness in a multi-task learning problem . they use a genre-based corpus to train a neural net model with different genres .
Outcome:	The proposed model improves the results across different genres compared to a single model trained on a genre.

Copied to clipboard

Wei-Jen Ko, Ahmed El-Kishky, Adithya Renduchintala, Vishrav Chaudhary, Naman Goyal, Francisco Guzmán, Pascale Fung, Philipp Koehn, Mona Diab

Multilinguality

Challenge:	linguistic overlap between low-resource languages and high-resourced languages is a major obstacle for training high-quality machine translation systems.
Approach:	They exploit linguistic overlap to facilitate translation to and from low-resource languages . they use monolingual data and parallel data in related high-resourced languages based on their method .
Outcome:	The proposed method significantly improves translation into low-resource language compared to baselines on 7 languages from three different language families.

Gender bias amplification during Speed-Quality optimization in Neural Machine Translation (2021.acl-short)

Copied to clipboard

Adithya Renduchintala, Denise Diaz, Kenneth Heafield, Xian Li, Mona Diab

Multilinguality

Challenge:	et al., 2002) show that gendered noun translation performance degrades faster than BLEU.
Approach:	They propose to use greedy search, quantization, AANs and shallow decoders to speed up decoding . they find minimal degradation of BLEU, but gendered noun translation degrades faster .
Outcome:	The proposed model degrades gendered noun translation performance faster than other models.

BeSt: The Belief and Sentiment Corpus (2022.lrec-1)

Copied to clipboard

Jennifer Tracey, Owen Rambow, Claire Cardie, Adam Dalton, Hoa Trang Dang, Mona Diab, Bonnie Dorr, Louise Guthrie, Magdalena Markowska, Smaranda Muresan, Vinodkumar Prabhakaran, Samira Shaikh, Tomek Strzalkowski

Sentiment Analysis

Challenge:	a corpus of propositional content is a set of cognitive attitudes of different agents towards a text . propositional attitudes are a cognitive attitude, including belief and sentiment, towards .
Approach:	They propose a corpus which records cognitive state: who believes what, who has what sentiment . they use newswire and discussion forums in Chinese, English, and Spanish .
Outcome:	The proposed corpus records who believes what (i.e., factuality) and who has what sentiment towards what.

WASA: A Web Application for Sequence Annotation (L18-1)

Copied to clipboard

Fahad AlGhamdi, Mona Diab

Semantic Compression for Word and Sentence Embeddings using Discrete Wavelet Transform (2024.findings-acl)

Challenge:	a major barrier to research on CS has been the lack of large multilingual, multi-genre CS-annotated corpora.
Approach:	They propose a web-based annotation system that manages large-scale CS data annotation.
Outcome:	The proposed system can manage large-scale multilingual code switching (CS) data annotation.

Copied to clipboard

Rana Salama, Abdou Youssef, Mona Diab

Diversity, Density, and Homogeneity: Quantitative Characteristic Metrics for Text Collections (2020.lrec-1)

Challenge:	Discrete Wavelet Transforms (DWT) can be applied to NLP capturing a variety of linguistic and semantic properties.
Approach:	They propose to use Discrete Wavelet Transforms to analyze word and sentence embeddings . they show that DWT can reduce dimensionality of embeddables by 50-93% .
Outcome:	The proposed paradigm reduces embeddings' dimensionality by 50-93% while maintaining their quality.

Copied to clipboard

Yi-An Lai, Xuan Zhu, Yi Zhang, Mona Diab

Information Retrieval

Challenge:	Existing descriptive statistics are inadequate to summarize text collections by quantitative measures.
Approach:	They propose a set of characteristic metrics that quantitatively measure the dispersion, sparsity, and uniformity of a text collection.
Outcome:	The proposed metrics are highly correlated with text classification performance of a renowned model, which could inspire future applications.

Evaluation of Unsupervised Compositional Representations (C18-1)

Copied to clipboard

Hanan Aldarmaki, Mona Diab

Context-Aware Cross-Lingual Mapping (N19-1)

Challenge:	averaging is a powerful representation of word meanings, but it has drawbacks for some words that do not carry semantic significance.
Approach:	They evaluated various compositional models on extrinsic supervised and unsupervised evaluation benchmarks.
Outcome:	The proposed models outperform context-sensitive models on several extrinsic supervised and unsupervised evaluation benchmarks.

Copied to clipboard

Hanan Aldarmaki, Mona Diab

Multilinguality

Challenge:	Cross-lingual word vectors are typically obtained by fitting an orthogonal matrix that maps the entries of a bilingual dictionary from a source to a target vector space.
Approach:	They propose an alternative to word-level mapping that better reflects sentence-level cross-lingual similarity by directly mapping the averaged embeddings of aligned sentences in a parallel corpus.
Outcome:	The proposed approach outperforms context-independent word mapping in translation retrieval.

Sentence and Clause Level Emotion Annotation, Detection, and Classification in a Multi-Genre Corpus (L18-1)

Copied to clipboard

Shabnam Tafreshi, Mona Diab

Information Extraction & Text Mining

Challenge:	Existing methods for predicting emotion categories are limited due to their multi-label nature . e.g. anger, joy, sadness are difficult to predict due to inherent multi-genre nature - a problem that is often overlooked in single-genrete text.
Approach:	They propose to expand existing annotated data to include 8 emotions from Plutchik's Wheel of Emotions . they explore the effectiveness of clause annotation in sentence-level emotion detection and classification .
Outcome:	The proposed system is the first to target the clause level and provides emotion classification for movie reviews datasets.

Methods for Measuring, Updating, and Visualizing Factual Beliefs in Language Models (2023.eacl-main)

Copied to clipboard

Peter Hase, Mona Diab, Asli Celikyilmaz, Xian Li, Zornitsa Kozareva, Veselin Stoyanov, Mohit Bansal, Srinivasan Iyer

Discrete Cosine Transform as Universal Sentence Encoder (2021.acl-short)

Challenge:	Pretrained language models store a large amount of factual information that can be elicited by prompting or finetuning.
Approach:	They propose methods to measure model factual beliefs and update incorrect beliefs in models . they propose a new visualization tool that shows relationships between stored model beliefs .
Outcome:	The proposed methods improve models' consistency and accuracy, the authors show . their methods outperform existing methods in more difficult settings, the paper shows .

Copied to clipboard

Nada Almarwani, Mona Diab

Analyzing the Role of Semantic Representations in the Era of Large Language Models (2024.naacl-long)

Challenge:	Modern sentence encoders capture underlying linguistic characteristics of words . Discrete Cosine Transform (DCT) is an efficient alternative to averaging .
Approach:	They propose to use a Discrete Cosine Transform to generate universal sentence representations in different languages.
Outcome:	The proposed model captures the underlying syntactic characteristics of a given text without compromising practical efficiency.

Copied to clipboard

Zhijing Jin, Yuen Chen, Fernando Gonzalez Adauto, Jiarui Liu, Jiayi Zhang, Julian Michael, Bernhard Schölkopf, Mona Diab

Efficient Convolutional Neural Networks for Diacritic Restoration (D19-1)

Challenge:	Existing studies show the benefits of semantic representations in NLP tasks . Existing work using AMR is concerned with trainable models .
Approach:	They propose an AMR-driven chain-of-thought prompting method that uses AMR . they propose to use it to predict which input examples AMR may help or hurt on .
Outcome:	The proposed method hurts performance more than it helps on five different tasks.

Copied to clipboard

Sawsan Alqahtani, Ajay Mishra, Mona Diab

CASA-NLU: Context-Aware Self-Attentive Natural Language Understanding for Task-Oriented Chatbots (D19-1)

Challenge:	Diacritic restoration is a computational task that requires a computer to understand written texts.
Approach:	They propose to use Temporal Convolutional Neural Networks (TCN) to restore missing diacritics for each character in written text.
Outcome:	The proposed model improves on TCN in Arabic, Yoruba, and Vietnamese.

Copied to clipboard

Arshit Gupta, Peng Zhang, Garima Lalwani, Mona Diab

Dialogue Systems & Conversational Agents

Challenge:	Prior work on contextual NLU has been limited in terms of the types of contextual signals used and the understanding of their impact on the model.
Approach:	They propose a context-aware self-attentive NLU model that uses multiple signals over a variable context window, such as previous intents, slots, dialog acts and utterances, in addition to the current user uttered.
Outcome:	The proposed model outperforms a baseline model on two conversational datasets yielding a gain of up to 7% on the IC task.

AnswerSumm: A Manually-Curated Dataset and Pipeline for Answer Summarization (2022.naacl-main)

Copied to clipboard

Alexander Fabbri, Xiaojian Wu, Srini Iyer, Haoran Li, Mona Diab

Summarization

Challenge:	Community Question Answering (CQA) fora lack a dataset to produce answer summarizations . a novel dataset of 4,631 CQA threads is used to generate answer summaries .
Approach:	They propose a dataset of 4,631 CQA threads for answer summarization curated by professional linguists.
Outcome:	The proposed approach boosts summarization performance according to automatic evaluation.

Active Learning for Rumor Identification on Social Media (2021.findings-emnlp)

Copied to clipboard

Parsa Farinneya, Mohammad Mahdi Abdollah Pour, Sardar Hamidian, Mona Diab

Detecting Urgency Status of Crisis Tweets: A Transfer Learning Approach for Low Resource Languages (2020.coling-main)

Challenge:	Existing methods for rumor tracking depend on a significant amount of labeled data.
Approach:	They propose an Active-Transfer Learning strategy to identify rumors with limited amount of annotated data.
Outcome:	The proposed approach achieves faster convergence in terms of the F-score while requiring fewer annotated samples (42% of the whole dataset for the best model).

Copied to clipboard

Efsun Sarioglu Kayi, Linyong Nan, Bohan Qu, Mona Diab, Kathleen McKeown

FEQA: A Question Answering Evaluation Framework for Faithfulness Assessment in Abstractive Summarization (2020.acl-main)

Challenge:	We train monolingual and cross-lingual classifiers on the extracted features of tweets . we use a few state-of-the-art contextual embeddings to extract features of the tweets.
Approach:	They propose to use tweets to train a dataset of English and two low-resource languages to train zero-shot transfer models.
Outcome:	The proposed model performs well in English and in low-resource languages . the proposed model is based on state-of-the-art embeddings and semi-supervised methods .

Copied to clipboard

Esin Durmus, He He, Mona Diab

Summarization

Challenge:	Existing automatic metrics do not capture errors in abstractive summarization models.
Approach:	They propose an automatic question answering metric for faithfulness that leverages recent advances in reading comprehension.
Outcome:	The proposed metric has significantly higher correlation with human faithfulness scores on highly abstracted summaries.

Towards Responsible Natural Language Annotation for the Varieties of Arabic (2022.findings-acl)

Copied to clipboard

A. Bergman, Mona Diab

Investigating Cultural Alignment of Large Language Models (2024.acl-long)

Challenge:	In NLP, there is a tendency to aim for broader coverage, often overlooking cultural and (socio)linguistic nuance.
Approach:	They propose a playbook for responsible dataset creation for polyglossic, multidialectal languages . they focus on Arabic annotation of social media content as an example .
Outcome:	The proposed model is based on Arabic annotation of social media content.

Copied to clipboard

Badr AlKhamissi, Muhammad ElNokrashy, Mai Alkhamissi, Mona Diab

Identifying Nuances in Fake News vs. Satire: Using Semantic and Linguistic Cues (D19-50)

Challenge:	Large Language Models (LLMs) are used to represent the diversity of human experience and culturally sensitive topics.
Approach:	They propose a method leveraging anthropological reasoning to enhance cultural alignment by prompting LLMs with different pretraining data mixtures in Arabic and English.
Outcome:	The proposed method enables users to better represent the diversity of human experience and the plurality of different cultures.

Copied to clipboard

Or Levi, Pedram Hosseini, Mona Diab, David Broniatowski

Detecting Hallucinated Content in Conditional Neural Sequence Generation (2021.findings-acl)

Challenge:	a blurry line between fake news and protected-speech satire has been a struggle for social media platforms . purveyors of fake news have begun to masquerade as satirical sites to avoid being demoted .
Approach:	They propose to automatically classify fake news versus satire based on language differences . they hypothesize that nuances could be identified using semantic and linguistic cues .
Outcome:	The proposed method can identify nuances between fake news and satire based on language differences . the proposed method is compared to the language-based baseline and is highly scalable .

Copied to clipboard

Chunting Zhou, Graham Neubig, Jiatao Gu, Mona Diab, Francisco Guzmán, Luke Zettlemoyer, Marjan Ghazvininejad

Text Generation

Challenge:	Neural sequence models can generate fluent sentences, but they can also hallucinate additional content not supported by the input.
Approach:	They propose a task to predict whether each token in the output sequence is hallucinated and collect manually annotated evaluation sets for this task.
Outcome:	The proposed method outperforms baseline methods on machine translation and abstractive summarization datasets and achieves significant improvements in both supervised and unsupervised settings.

Text Characterization Toolkit (TCT) (2022.aacl-demo)

Copied to clipboard

Daniel Simig, Tianlu Wang, Verna Dankers, Peter Henderson, Khuyagbaatar Batsuren, Dieuwke Hupkes, Mona Diab

ToKen: Task Decomposition and Knowledge Infusion for Few-Shot Hate Speech Detection (2022.emnlp-main)

Challenge:	Text Characterization Toolkit (TCT) is a tool that researchers can use to study characteristics of large datasets.
Approach:	They propose a text characterization toolkit that researchers can use to study characteristics of large datasets.
Outcome:	The proposed tool can be used to study characteristics of large datasets and to understand the influence of attributes on models’ behaviour.

Copied to clipboard

Badr AlKhamissi, Faisal Ladhak, Srinivasan Iyer, Veselin Stoyanov, Zornitsa Kozareva, Xian Li, Pascale Fung, Lambert Mathias, Asli Celikyilmaz, Mona Diab

Depth-Wise Attention (DWAtt): A Layer Fusion Method for Data-Efficient Classification (2024.lrec-main)

Challenge:	Hate speech detection is complex and requires commonsense reasoning and social nuance . prior work has shown that even humans cannot achieve a high agreement on whether a post constitutes HS .
Approach:	They frame a few-shot learning task to decompose a hate speech detection task into its "constituent" parts. they show that infusing commonsense knowledge from reasoning datasets improves the performance even further.
Outcome:	The proposed method outperforms baseline methods in the 16-shot case.

Copied to clipboard

Muhammad ElNokrashy, Badr AlKhamissi, Mona Diab