---
license: mit
language:
- en
- da
- zh
tags:
- Responsible AI
- Recommender Systems
pretty_name: HealthRec
size_categories:
- 100M<n<1B
---
#### Investigating Recommender Systems from the Healthiness Perspective: Benchmarks, Warnings and Enhancement

This is the resource of our work "Investigating Recommender Systems from the Healthiness Perspective: Benchmarks, Warnings and Enhancement" which aims to investigate recommender systems from the healthiness perspective. 

This resource includes:

**Three constructed datasets** with healthiness-related information under "Benchmarks" path;

**Two healthiness-related metrics** proposed to quantitatively evaluate the healthiness of recommendation results under "HealthinessMetrics" path.

**A novel framwork HealthRec**, which is a generic and model-agnostic framework that can be seamlessly integrated into existing models to promote healthy content dissemination, under "HealthRec_code" path.

More details about these resources can be found in the "readme.md" files in each path.

By providing this resource, we seek to raise awareness of potential risks associated with recommender systems and provide a foundation for building more responsible, healthiness-aware approaches.