Opioid death projections with AI-based forecasts using social media language
Abstract Targeting of location-specific aid for the U.S. opioid epidemic is difficult due to our inability to accurately predict changes in opioid mortality across heterogeneous communities. AI-based language analyses, having recently shown promise in cross-sectional (between-community) well-being a...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-03-01
|
| Series: | npj Digital Medicine |
| Online Access: | https://doi.org/10.1038/s41746-023-00776-0 |
| _version_ | 1827608130659287040 |
|---|---|
| author | Matthew Matero Salvatore Giorgi Brenda Curtis Lyle H. Ungar H. Andrew Schwartz |
| author_facet | Matthew Matero Salvatore Giorgi Brenda Curtis Lyle H. Ungar H. Andrew Schwartz |
| author_sort | Matthew Matero |
| collection | DOAJ |
| description | Abstract Targeting of location-specific aid for the U.S. opioid epidemic is difficult due to our inability to accurately predict changes in opioid mortality across heterogeneous communities. AI-based language analyses, having recently shown promise in cross-sectional (between-community) well-being assessments, may offer a way to more accurately longitudinally predict community-level overdose mortality. Here, we develop and evaluate, TrOP (Transformer for Opiod Prediction), a model for community-specific trend projection that uses community-specific social media language along with past opioid-related mortality data to predict future changes in opioid-related deaths. TOP builds on recent advances in sequence modeling, namely transformer networks, to use changes in yearly language on Twitter and past mortality to project the following year’s mortality rates by county. Trained over five years and evaluated over the next two years TrOP demonstrated state-of-the-art accuracy in predicting future county-specific opioid trends. A model built using linear auto-regression and traditional socioeconomic data gave 7% error (MAPE) or within 2.93 deaths per 100,000 people on average; our proposed architecture was able to forecast yearly death rates with less than half that error: 3% MAPE and within 1.15 per 100,000 people. |
| first_indexed | 2024-03-09T07:06:37Z |
| format | Article |
| id | doaj.art-c72696cd242e4b48a97c1473aa846145 |
| institution | Directory Open Access Journal |
| issn | 2398-6352 |
| language | English |
| last_indexed | 2024-03-09T07:06:37Z |
| publishDate | 2023-03-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Digital Medicine |
| spelling | doaj.art-c72696cd242e4b48a97c1473aa8461452023-12-03T09:32:20ZengNature Portfolionpj Digital Medicine2398-63522023-03-016111110.1038/s41746-023-00776-0Opioid death projections with AI-based forecasts using social media languageMatthew Matero0Salvatore Giorgi1Brenda Curtis2Lyle H. Ungar3H. Andrew Schwartz4Department of Computer Science, Stony Brook UniversityDepartment of Computer and Information Science, University of PennsylvaniaNational Institute on Drug Abuse, National Institutes of HealthDepartment of Computer and Information Science, University of PennsylvaniaDepartment of Computer Science, Stony Brook UniversityAbstract Targeting of location-specific aid for the U.S. opioid epidemic is difficult due to our inability to accurately predict changes in opioid mortality across heterogeneous communities. AI-based language analyses, having recently shown promise in cross-sectional (between-community) well-being assessments, may offer a way to more accurately longitudinally predict community-level overdose mortality. Here, we develop and evaluate, TrOP (Transformer for Opiod Prediction), a model for community-specific trend projection that uses community-specific social media language along with past opioid-related mortality data to predict future changes in opioid-related deaths. TOP builds on recent advances in sequence modeling, namely transformer networks, to use changes in yearly language on Twitter and past mortality to project the following year’s mortality rates by county. Trained over five years and evaluated over the next two years TrOP demonstrated state-of-the-art accuracy in predicting future county-specific opioid trends. A model built using linear auto-regression and traditional socioeconomic data gave 7% error (MAPE) or within 2.93 deaths per 100,000 people on average; our proposed architecture was able to forecast yearly death rates with less than half that error: 3% MAPE and within 1.15 per 100,000 people.https://doi.org/10.1038/s41746-023-00776-0 |
| spellingShingle | Matthew Matero Salvatore Giorgi Brenda Curtis Lyle H. Ungar H. Andrew Schwartz Opioid death projections with AI-based forecasts using social media language npj Digital Medicine |
| title | Opioid death projections with AI-based forecasts using social media language |
| title_full | Opioid death projections with AI-based forecasts using social media language |
| title_fullStr | Opioid death projections with AI-based forecasts using social media language |
| title_full_unstemmed | Opioid death projections with AI-based forecasts using social media language |
| title_short | Opioid death projections with AI-based forecasts using social media language |
| title_sort | opioid death projections with ai based forecasts using social media language |
| url | https://doi.org/10.1038/s41746-023-00776-0 |
| work_keys_str_mv | AT matthewmatero opioiddeathprojectionswithaibasedforecastsusingsocialmedialanguage AT salvatoregiorgi opioiddeathprojectionswithaibasedforecastsusingsocialmedialanguage AT brendacurtis opioiddeathprojectionswithaibasedforecastsusingsocialmedialanguage AT lylehungar opioiddeathprojectionswithaibasedforecastsusingsocialmedialanguage AT handrewschwartz opioiddeathprojectionswithaibasedforecastsusingsocialmedialanguage |