Papers by Jennifer Foster

20 papers
Reference-Free Evaluation of Taxonomies (2026.findings-acl)

Copied to clipboard

Challenge: Taxonomies are used to classify items, ideas or organisms based on shared characteristics.
Approach: They introduce two reference-free metrics for quality evaluation of taxonomies in the absence of labels.
Outcome: The proposed metrics correlate well with F1 against ground truth taxonomies on five taxonomies and improve hierarchical classification when used with label hierarchies.
How to Make Neural Natural Language Generation as Reliable as Templates in Task-Oriented Dialogue (2020.emnlp-main)

Copied to clipboard

Challenge: Neural Natural Language Generation (NLG) systems are well known for their unreliability.
Approach: They propose a data augmentation approach which restricts the output of a neural network and guarantees reliability.
Outcome: The proposed approach scored 100% in semantic accuracy on the E2E NLG Challenge dataset, the same as a template system.
Cross-lingual Parsing with Polyglot Training and Multi-treebank Learning: A Faroese Case Study (D19-61)

Copied to clipboard

Challenge: Cross-lingual dependency parsing involves transferring syntactic knowledge from one language to another.
Approach: They compare two approaches to cross-lingual dependency parsing using monolingual source models and a polyglot model which is trained on the combination of all source languages.
Outcome: The proposed methods improve low-resource dependency parsers by transferring syntactic knowledge from one language to another.
Improving Document-Level Sentiment Analysis with User and Product Context (2020.coling-main)

Copied to clipboard

Challenge: Existing work that improves document-level sentiment analysis by encoding user and product information has been limited to considering only the text of the current review.
Approach: They propose to incorporate all available historical review text belonging to the author of the review in question and investigate the inclusion of his- torical reviews associated with the current product.
Outcome: The proposed model improves on IMDB, Yelp 2013 and Yelpan 2014 datasets by more than 2 percentage points in the best case.
Shape of Synth to Come: Why We Should Use Synthetic Data for English Surface Realization (2020.acl-main)

Copied to clipboard

Challenge: In the Surface Realization Shared Tasks of 2018 and 2019, there was little difference in absolute performance between systems trained with and without synthetic data.
Approach: They propose to use synthetic data to explore approaches to surface realization from Universal-Dependency-like trees to surface strings for several languages.
Outcome: The proposed method improves performance of a previously state-of-the-art system by 8 BLEU points over the previous system on the English dataset.
Investigating the Saliency of Sentiment Expressions in Aspect-Based Sentiment Analysis (2023.findings-acl)

Copied to clipboard

Challenge: In a set of masking experiments, we examine the extent to which the tokens identified as salient by LIME and a gradient-based method are being used by the classifier.
Approach: They use a BERT BASE model to mask the sentiments of an English dataset and find that both methods produce faithful rationales.
Outcome: The proposed classifier outperforms both the gradient-based and black-box saliency methods on the SemEval 2016 english dataset.
FarExStance: Explainable Stance Detection for Farsi (2025.coling-main)

Copied to clipboard

Challenge: FarExStance is a new dataset for explainable stance detection in Farsi . it contains extractive explanations as evidence for stance labels and claims .
Approach: They propose a dataset for explainable stance detection in Farsi with extractive explanations as evidence.
Outcome: The proposed model is the most accurate on stance detection, while the best explanation is from few-shot Claude-3.5-Sonnet.
Out-of-Distribution Generalization in Natural Language Processing: Past, Present, and Future (2023.emnlp-main)

Copied to clipboard

Challenge: Existing literature on the generalization of machine learning models to out-of-distribution data is lacking.
Approach: They propose to present the first comprehensive review of recent progress, methods, and evaluations on the generalization challenge from an OOD perspective in natural language understanding.
Outcome: The proposed survey provides the first comprehensive review of recent progress, methods, and evaluations on the generalization challenge from an OOD perspective in natural language understanding.
Semantic-Aware Dynamic Retrospective-Prospective Reasoning for Event-Level Video Question Answering (2023.acl-srw)

Copied to clipboard

Challenge: Event-Level Video Question Answering (EVQA) requires complex reasoning across video events to obtain the visual information needed to provide optimal answers.
Approach: They propose a semantic-aware dynamic retrospective-prospective reasoning approach for video-based question answering that explicitly uses the Semantic Role Labeling (SRL) structure of the question in the dynamic reasoning process.
Outcome: The proposed model outperforms existing models on a trafficQA benchmark dataset.
Exploiting Rich Textual User-Product Context for Improving Personalized Sentiment Analysis (2023.findings-acl)

Copied to clipboard

Challenge: Typical approaches do not exploit the potential of historical reviews or do not make full use of user/product associations.
Approach: They propose to use historical reviews to initialize user and product representations and incorporate textual associations via a user-product cross-context module.
Outcome: The proposed method outperforms existing state-of-the-art models on IMDb, Yelp and Longformer benchmarks.
gaBERT — an Irish Language Model (2022.lrec-1)

Copied to clipboard

Challenge: We compare gaBERT to multilingual BERT and the monolingual Irish WikiBERT and show that gaBERt provides better representations for downstream parsing tasks.
Approach: They propose a monolingual BERT model for the Irish language that provides better representations for a downstream parsing task.
Outcome: The proposed model performs better than the multilingual BERT and the monolingual Irish WikiBERT on a parsing task.
FoodTaxo: Generating Food Taxonomies with Large Language Models (2025.acl-industry)

Copied to clipboard

Challenge: a recent study shows that LLMs are useful for automating taxonomies from a seed taxonomy to a set of known concepts.
Approach: They propose to use Large Language Models for automated taxonomy generation and completion.
Outcome: The proposed approach is based on an open-source LLM (Llama-3).
Revisiting Tri-training of Dependency Parsers (2021.emnlp-main)

Copied to clipboard

Challenge: Pre-trained word embeddings and self-training have been used in dependency parsing tasks for years.
Approach: They compare tri-training and pretrained word embeddings in dependency parsing . they use language-specific FastText and ELMo embedds and multilingual BERT embedders .
Outcome: The proposed methods are tri-training and pretrained word embeddings.
Treebank Embedding Vectors for Out-of-Domain Dependency Parsing (2020.acl-main)

Copied to clipboard

Challenge: a recent advance in monolingual dependency parsing is the idea of a treebank embedding vector . this allows the model to prefer training data from one treebank over another at test time .
Approach: They propose a method to predict a treebank vector for sentences that do not come from a particular treebank . they also explore what happens when they move away from predefined treebank embedding vectors .
Outcome: The proposed method can predict treebank vectors for sentences that do not come from a treebank used in training with sufficient accuracy for nine out of ten languages.
Do Stochastic Parrots have Feelings Too? Improving Neural Detection of Synthetic Text via Emotion Recognition (2023.findings-emnlp)

Copied to clipboard

Challenge: Recent advances in generative AI have shone a spotlight on high-performance synthetic text generation technologies.
Approach: They propose to use emotion-driven pretrained language models to generate synthetic text that lacks emotional coherence.
Outcome: The proposed detector achieves significant improvements across a range of synthetic text generators, various sized models, datasets, and domains.
Is It Dish Washer Safe? Automatically Answering “Yes/No” Questions Using Customer Reviews (N19-3)

Copied to clipboard

Challenge: Using Amazon reviews, we find that the answer to a question is only in 45% of cases.
Approach: They combine Amazon reviews with consumer reviews and manually analyse 400 questions from four domains to find that reviews directly contain the answer to the question . they then compare QA systems that use reviews in addition to the questions to see if they can be useful for other question types.
Outcome: The proposed system outperforms the chance baseline but not by a large margin.
English Machine Reading Comprehension Datasets: A Survey (2021.emnlp-main)

Copied to clipboard

Challenge: a survey of English Machine Reading Comprehension datasets is carried out . the aim is to provide a concise yet informative overview of the landscape .
Approach: They survey 60 English Machine Reading Comprehension datasets to provide a resource for other researchers interested in this problem.
Outcome: The proposed survey covers 60 English MRC datasets with a view to providing a resource for other researchers interested in the problem.
Fact or Factitious? Contextualized Opinion Spam Detection (P19-2)

Copied to clipboard

Challenge: In analytic analysis of fake reviews, we compare a number of techniques to detect them.
Approach: They propose a machine learning approach that fine-tunes contextualised embeddings to detect fake reviews.
Outcome: The proposed approach fine-tunes state of the art contextualised embeddings to show that it is effective in detecting fake reviews and lay the groundwork for future research in this area.
Improving Unsupervised Question Answering via Summarization-Informed Question Generation (2021.emnlp-main)

Copied to clipboard

Challenge: Question Generation (QG) is the production of meaningful questions given a set of input passages and corresponding answers.
Approach: They propose a method which uses questions generated heuristically from news summaries as a source of training data for a QG system.
Outcome: The proposed method outperforms previous unsupervised models on three in-domain datasets and three out-of-domain ones.
TwittIrish: A Universal Dependencies Treebank of Tweets in Modern Irish (2022.acl-long)

Copied to clipboard

Challenge: Modern Irish is a minority language lacking computational resources for accurate automatic syntactic parsing of user-generated content.
Approach: They propose to use a treebank to facilitate natural language parsing of user-generated content in Irish.
Outcome: The proposed treebank enables natural language processing of user-generated content in Irish.

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