Artificial Intelligence for Social Good: A Survey

Zheyuan Ryan Shi,Claire Wang,Fei Fang

DOI: https://doi.org/10.48550/arXiv.2001.01818

2020-01-07

Abstract:Artificial intelligence for social good (AI4SG) is a research theme that aims to use and advance artificial intelligence to address societal issues and improve the well-being of the world. AI4SG has received lots of attention from the research community in the past decade with several successful applications. Building on the most comprehensive collection of the AI4SG literature to date with over 1000 contributed papers, we provide a detailed account and analysis of the work under the theme in the following ways. (1) We quantitatively analyze the distribution and trend of the AI4SG literature in terms of application domains and AI techniques used. (2) We propose three conceptual methods to systematically group the existing literature and analyze the eight AI4SG application domains in a unified framework. (3) We distill five research topics that represent the common challenges in AI4SG across various application domains. (4) We discuss five issues that, we hope, can shed light on the future development of the AI4SG research.

Computers and Society,Artificial Intelligence

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to systematically review and analyze the research progress in the field of "Artificial Intelligence for Social Good" (AI4SG), and identify the existing research and technical challenges in this field. Specifically, the authors address this problem through the following four aspects:

1. **Literature Review**: It provides the most comprehensive collection of AI4SG literature to date, covering more than 1,000 contributed papers, and quantitatively analyzes the distribution and trends of these literatures from two dimensions: application areas and artificial intelligence technologies used.

2. **Conceptual Approaches**: Three conceptual approaches are proposed to systematically group the existing literature and analyze eight AI4SG application areas within a unified framework. This helps to more clearly understand the research status and future directions in different fields.

3. **Research Topics**: Five representative research topics are distilled, which reflect the common challenges of AI4SG in different application areas and are helpful to guide future research work.

4. **Future Development Directions**: Five issues are discussed, which are aimed at providing inspiration for the future development of AI4SG research, including aspects such as evaluation criteria and deployment strategies.

Through the above methods, the paper not only summarizes the existing achievements in the AI4SG field, but also points out the future research directions, aiming to promote further development and innovation in this field.