Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enhancing Health Mention Classification Performance: A Study on Advancements in Parameter Efficient Tuning

Reem Abdel-Salam, Mary Adewunmi

TL;DR

The paper tackles Health Mention Classification on social media by integrating Part-of-Speech information with parameter-efficient fine-tuning methods, including prompt-tuning, soft-prompting, P-tuning V2, and LoRa, across three public datasets (PHM2017, RHMD, Illness). It demonstrates that POS augmentation and various PEFT strategies can yield consistent F1 improvements over baselines while enabling smaller, more training-efficient models, with notable gains on PHM2017 (91.5 F1) and Illness (95.5 F1). The study also investigates domain adaptation via masked language modeling and cross-dataset pre-training, finding that POS-aware representations and careful PEFT configurations enhance generalization. Overall, the approach offers a practical pathway to improve HMC performance in real-world social media monitoring while reducing computational resources.

Abstract

Health Mention Classification (HMC) plays a critical role in leveraging social media posts for real-time tracking and public health monitoring. Nevertheless, the process of HMC presents significant challenges due to its intricate nature, primarily stemming from the contextual aspects of health mentions, such as figurative language and descriptive terminology, rather than explicitly reflecting a personal ailment. To address this problem, we argue that clearer mentions can be achieved through conventional fine-tuning with enhanced parameters of biomedical natural language methods (NLP). In this study, we explore different techniques such as the utilisation of part-of-speech (POS) tagger information, improving on PEFT techniques, and different combinations thereof. Extensive experiments are conducted on three widely used datasets: RHDM, PHM, and Illness. The results incorporated POS tagger information, and leveraging PEFT techniques significantly improves performance in terms of F1-score compared to state-of-the-art methods across all three datasets by utilising smaller models and efficient training. Furthermore, the findings highlight the effectiveness of incorporating POS tagger information and leveraging PEFT techniques for HMC. In conclusion, the proposed methodology presents a potentially effective approach to accurately classifying health mentions in social media posts while optimising the model size and training efficiency.

Enhancing Health Mention Classification Performance: A Study on Advancements in Parameter Efficient Tuning

TL;DR

The paper tackles Health Mention Classification on social media by integrating Part-of-Speech information with parameter-efficient fine-tuning methods, including prompt-tuning, soft-prompting, P-tuning V2, and LoRa, across three public datasets (PHM2017, RHMD, Illness). It demonstrates that POS augmentation and various PEFT strategies can yield consistent F1 improvements over baselines while enabling smaller, more training-efficient models, with notable gains on PHM2017 (91.5 F1) and Illness (95.5 F1). The study also investigates domain adaptation via masked language modeling and cross-dataset pre-training, finding that POS-aware representations and careful PEFT configurations enhance generalization. Overall, the approach offers a practical pathway to improve HMC performance in real-world social media monitoring while reducing computational resources.

Abstract

Health Mention Classification (HMC) plays a critical role in leveraging social media posts for real-time tracking and public health monitoring. Nevertheless, the process of HMC presents significant challenges due to its intricate nature, primarily stemming from the contextual aspects of health mentions, such as figurative language and descriptive terminology, rather than explicitly reflecting a personal ailment. To address this problem, we argue that clearer mentions can be achieved through conventional fine-tuning with enhanced parameters of biomedical natural language methods (NLP). In this study, we explore different techniques such as the utilisation of part-of-speech (POS) tagger information, improving on PEFT techniques, and different combinations thereof. Extensive experiments are conducted on three widely used datasets: RHDM, PHM, and Illness. The results incorporated POS tagger information, and leveraging PEFT techniques significantly improves performance in terms of F1-score compared to state-of-the-art methods across all three datasets by utilising smaller models and efficient training. Furthermore, the findings highlight the effectiveness of incorporating POS tagger information and leveraging PEFT techniques for HMC. In conclusion, the proposed methodology presents a potentially effective approach to accurately classifying health mentions in social media posts while optimising the model size and training efficiency.
Paper Structure (17 sections, 3 figures, 5 tables)

This paper contains 17 sections, 3 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Materials and Methods
  4. Datasets
  5. Methodology
  6. Incorporating POS information
  7. Masked Language Modeling
  8. Parameter Efficient fine-tuning
  9. Prompt-tuning
  10. Soft-Prompting
  11. P-tuning V2/Prefix-tuning
  12. P-tuning V2 / Prefix-tuning with LoRa
  13. Results and Discussion
  14. Experimental Setup
  15. Results
  16. ...and 2 more sections

Figures (3)

  • Figure 1: Prompt architecture for the RHMD dataset.
  • Figure 2: Modified Soft-prompting architecture to include virtual tokens before and after the text.
  • Figure 3: Modified Soft-prompting architecture to include a virtual token before the text and a hard prompt after the text.