Stakeholder Insights in Alzheimer’s Disease: Natural Language Processing of Social Media Conversations
Bibliographic record
Abstract
BACKGROUND: Social media data may be especially effective for studying diseases associated with high stigma, such as Alzheimer's disease (AD). OBJECTIVE: We primarily aimed to identify issues/challenges experienced by patients with AD using natural language processing (NLP) of social media posts. METHODS: We searched 130 public social media sources between January 1998 and December 2021 for AD stakeholder social media posts using NLP to identify issues/challenges experienced by patients with AD. Issues/challenges identified by ≥10% of any AD stakeholder type were described. Illustrative posts were selected for qualitative review. Secondarily, issues/challenges were organized into a conceptual AD identification framework (ADIF) and representation of ADIF categories within clinical instruments was assessed. RESULTS: We analyzed 1,859,077 social media posts from 30,341 AD stakeholders (21,011 caregivers; 7,440 clinicians; 1,890 patients). The most common issues/challenges were Worry/anxiety (34.2%), Pain (33%), Malaise (28.7%), Confusional state (27.1%), and Falls (23.9%). Patients reported a markedly higher volume of issues/challenges than other stakeholders. Patient posts reflected the broader scope of patient burden, caregiver posts captured both patient and caregiver burden, and clinician posts tended to be targeted. Less than 5% of the high frequency issues/challenges were in the "function and independence" and "social and relational well-being" categories of the ADIF, suggesting these issues/challenges may be difficult to capture. No single clinical instrument covered all ADIF categories; "social and relational well-being" was least represented. CONCLUSION: NLP of AD stakeholder social media data revealed a broad spectrum of real-world insights regarding patient burden.
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 | 0.014 | 0.050 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.008 | 0.006 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.005 | 0.009 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".