Automated Detection of Substance Use-Related Social Media Posts Based on Image and Text Analysis

@article{Roy2017AutomatedDO,
  title={Automated Detection of Substance Use-Related Social Media Posts Based on Image and Text Analysis},
  author={Arpita Roy and Anamika Paul and Hamed Pirsiavash and Shimei Pan},
  journal={2017 IEEE 29th International Conference on Tools with Artificial Intelligence (ICTAI)},
  year={2017},
  pages={772-779}
}
Nowadays, teens and young adults spend a significant amount of time on social media. According to the national survey of American attitudes on substance abuse, American teens who spend time on social media sites are at increased risk of smoking, drinking and illicit drug use. Reducing teens’ exposure to substance use-related social media posts may help minimize their risk of future substance use and addiction. In this paper, we present a method for automated detection of substance userelated… CONTINUE READING

Figures, Tables, Results, and Topics from this paper.

Key Quantitative Results

  • Our best model achieved 90% prediction accuracy and 75% FMeasure, which are significantly better than models that use image or text features alone.
  • For example, compared with the best Image models, the combined model achieved a 60% increase of precision and 39% increase of F1 measure. Similarly, compared with the best text model, the combined model achieved a 147% increase of precision and 88% increase of F1 score.

Citations

Publications citing this paper.

Interpreting Social Media-Based Substance Use Prediction Models with Knowledge Distillation

  • 2018 IEEE 30th International Conference on Tools with Artificial Intelligence (ICTAI)
  • 2018
VIEW 5 EXCERPTS

References

Publications referenced by this paper.
SHOWING 1-10 OF 34 REFERENCES