Papers by Milton King
Exploring BERT-Based Classification Models for Detecting Phobia Subtypes: A Novel Tweet Dataset and Comparative Analysis (2024.lrec-main)
Copied to clipboard
| Challenge: | Phobias are characterized by an intense and irrational fear of specific objects, situations, or activities despite there being no real risk or only a minor threat involved. |
| Approach: | They propose to use a dataset of 811,569 English tweets from user timelines spanning 102 phobia subtypes over six months to classify users into 65 specific phobias. |
| Outcome: | The proposed dataset includes 47,614 self-diagnosed phobia users and a high f1 score for binary classification and multi-class classification. |
Sense of the Day: Short Timeframe Temporal-Aware Word Sense Disambiguation (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing models that consider the predominant sense of a lemma can be tailored to a specific timeframe and based on the timeframe of the text. |
| Approach: | They use Twitter to explore whether different senses are favoured within specific timeframes and how they are used to create short timeframe temporal-aware word sense disambiguation models. |
| Outcome: | The proposed model outperforms temporal agnostic models and author-aware models. |
Leveraging distributed representations and lexico-syntactic fixedness for token-level prediction of the idiomaticity of English verb-noun combinations (P18-2)
Copied to clipboard
| Challenge: | Verb-noun combinations (VNCs) are ambiguous between literal and idiomatic usages in English. |
| Approach: | They propose and evaluate models for classifying verb-noun combinations as idiomatic or literal, based on averaging word embeddings and a variety of approaches to forming distributed representations. |
| Outcome: | The proposed model outperforms a previous model based on skip-thoughts and averaging word embeddings. |
Evaluating Approaches to Personalizing Language Models (2020.lrec-1)
Copied to clipboard
| Challenge: | a large amount of text is not available for training a user-specific language model, which suggests a need to personalize language models with only a small amount of data. |
| Approach: | They propose three approaches to personalize a language model that was trained on a large background corpus using a relatively small amount of text from an individual user. |
| Outcome: | The proposed techniques outperform language model adaptation based on demographic factors. |