Proceedings of the Sixth Social Media Mining for Health (#SMM4H) Workshop and Shared Task
Bibliographic record
Abstract
Welcome to the 6th Social Media Mining for Health (#SMM4H) Workshop & Shared Task 2021, co-located at the 2021 Annual Conference of the North American Chapter of the Association for Computational Linguistics.Held online in its sixth iteration, #SMM4H 2021 continues to serve as a venue for bringing together data mining researchers interested in building solutions for challenges involved in utilizing social media data for health informatics.For #SMM4H 2021, we accepted 3 workshop papers and 29 shared task system description papers.Each submission was peer-reviewed by two to three reviewers.The accepted workshop papers used social media data, mainly from Twitter, for topics, studies and applications surrounding COVID-19 and pharmacovigilance.Niu et.al. present a study summarizing the evaluation of Twitter sentiments towards non-pharmaceutical interventions for COVID-19 in Canada.Karisani et.al. propose a novel technique that uses both unlabeled and labeled tweets with drug mentions along multiple views to achieve a new state-of-the-art performance in extracting adverse drug effects.Finally, Miranda et.al. present a new annotated corpora in Spanish to identify occupational subgroups on Twitter to estimate risks associated with COVID-19.They also present a summary of the ProfNER shared task organized with the annotated data along the text classification and named entity recognition subtasks.The #SMM4H 2021 shared tasks sought to advance the use of Twitter data (tweets) for pharmacovigilance, medication non-adherence, patient-centered outcomes, tracking cases and symptoms associated with COVID-19 and assessing risks for occupational groups.In addition to re-reruns of adverse drug effects extraction tasks in English and Russian #SMM4H 2021 included new tasks for detecting medication non-adherence, adverse pregnancy outcomes, probable cases of COVID-19, symptoms associated with COVID-19, extracting occupations and professions from Spanish tweets for COVID-19 risk assessment and detecting self reports of breast cancer posts.The eight tasks required methods for binary classification, multi-class classification, and named entity recognition (NER).With 40 teams making prediction submissions, participation in the #SMM4H shared tasks continue to grow.Among the 29 shared task system description papers that were accepted, 9 teams were invited to present their system orally.The organizing committee of #SMM4H 2021 would like to thank the program committee for reviewing the workshop papers and the additional reviewers of system description papers for providing constructive feedback and participating in peer-review.We are also grateful to the organizers of NAACL 2021 for facilitating the organization of the workshop and the Codalab team for providing the platform to organize shared tasks.We would also like to thank the annotators of the shared task datasets, and of course, everyone who submitted a paper or participated in the shared tasks.#SMM4H 2021 would not have been possible without the contributions and participation from all of them.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.018 | 0.022 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.003 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.010 | 0.008 |
| Open science | 0.004 | 0.014 |
| Research integrity | 0.005 | 0.006 |
| Insufficient payload (model declined to judge) | 0.134 | 0.074 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".