Artificial intelligence and social work / edited by Milind Tambe, Eric Rice.

Format:

Book

Contributor:

Tambe, Milind, 1965- editor.

Rice, Eric (Eric R. W.), editor.

Series:

Artificial Intelligence for Social Good

Artificial intelligence for social good

Language:

English

Subjects (All):

Social service--Technological innovations.

Social service.

Artificial intelligence.

Social justice.

Physical Description:

1 online resource (xi, 257 pages) : digital, PDF file(s).

Edition:

1st ed.

Place of Publication:

Cambridge : Cambridge University Press, 2018.

Summary:

This book marries social work and artificial intelligence to provide an introductory guide for using AI for social good. Following an introductory chapter laying out approaches and ethical principles of using AI for social work interventions, the book describes in detail an intervention to increase the spread of HIV information by using algorithms to determine the key individuals in a social network of homeless youth. Other chapters present interdisciplinary collaborations between AI and social work students, including a chatbot for sexual health information and algorithms to determine who is at higher stress among persons with Type 2 Diabetes. For students, academic researchers, industry leaders, and practitioners, these real-life examples from the USC Center for Artificial Intelligence in Society demonstrate how social work and artificial intelligence can be used in tandem for the greater good.

Contents:

Cover

Half-title

Series information

Title page

Copyright information

Contents

Contributors

Part I

1 Merging Social Work Science and Computer Science for Social Good

Motivations for Our Research

What Is Social Work

What Is AI

The Unlikely Partnership

New Science for Both Sides

The Structure of the Book

2 The Causes and Consequences of Youth Homelessness

The Extent of Youth Homelessness

Causes of Youth Homelessness

Thrown Out of Home/Run Away from Home

Aging Out of Foster Care

Sexual- and Gender-Minority Youth

Travelers

Consequences of Youth Homelessness

Experiences of Violence

Mental Health

Substance Use/Abuse

Contraception and Pregnancy

The Importance of Social Networks

HIV Prevention for Homeless Youth

Why We Need Artificial Intelligence

References

3 Using Social Networks to Raise HIV Awareness among Homeless Youth

Introduction

Related Work

HEALER's Design

Network Construction Application

DIME Solver

HEALER Design

DIME Problem Statement

Uncertain Network

Influence Model

Problem Statement

The Value of Information

Computational Hardness

HEAL: DIME Problem Solver

POMDP Model

HEAL

DOSIM: A New Algorithm for the DIME Problem

Experimental Results

Conclusion

4 Influence Maximization in the Field: The Arduous Journey from Emerging to Deployed Application

Pilot Study Pipeline

Results from the Field

Challenges Uncovered

Conclusion, Limitations, and Lessons Learned

Acknowledgments

5 Influence Maximization with Unknown Network Structure

Exploratory Influence Maximization

The ARISEN Algorithm

Initial weights

Refining the weights

Experiments

Part II.

6 Maximizing the Spread of Sexual Health Information in a Multimodal Communication Network of Young Black Women

Methods

YBW Quantitative Network Survey - Data Collection

Mathematically Modeling Information Diffusion in the YBW Network

Algorithms for Solving the Maximum Influence Problem

Results and Discussion

Conclusions and Future Work

7 Minimizing Violence in Homeless Youth

Data Collection

Dependent Variable

Model

Voter Model

Uncertain Voter Model

Greedy Minimization

Uncertainty in Time

Discussion

Multiple Waves

Utilizing Other Variables

Implications from Social Work Perspective

8 Artificial Intelligence for Improving Access to Sexual Health Necessities for Youth Experiencing Homelessness

Homelessness and HIV Risk

Our Innovation: "Smart Kiosks"

Problem Definition

Weighted K-Center Problem

Datasets

Area Division

Model Parameters

Points V

Existing Centers V''

Distances d(v, r)

Weights W(v)

Proposed Approach

Approximation Guarantees of the Greedy Approach

Population Estimation

Results

Discussion and Conclusions

9 Know-Stress: Predictive Modeling of Stress among Diabetes Patients under Varying Conditions

Data Description

A Brief Background of the Data Set Used

Description of Variables Presented in the Dataset

Aim of Our Analysis

More Detailed Description of the Aims of the Project

Predicted Variable Creation

Determining Optimal Cutoff

Feature Selection and Various Data Sets

Experimental Setup

Models of Classification

Decision Tree

Logistic Regression

Feature Selection

Metrics of Evaluation

Experimental Results and Discussion

Discussion on the Results

Optimal Cutoff Range

Model Stability.

Feature Selection

Future Work

10 A Multidisciplinary Study on the Relationship between Foster Care Attributes and Posttraumatic Stress Disorder Symptoms on Foster Youth

Literature Review

PTSD Symptoms among Youth with Homelessness and Foster Care Histories

Machine Learning within the Social Sciences

Data Augmentation

Data Samples and Procedures

Data: Selected Variables and Outcomes for PTSD Prediction

Data Preprocessing for Models

Modeling Techniques and Evaluations

Evaluation Techniques

Model 0 (baseline): Logistic Regression

Model 1: Neural Networks

Model 2: Decision Tree

Model 3: Bayesian Networks

11 Artificial Intelligence to Predict Intimate Partner Violence Perpetration

Data Set Description

Data Analysis

Theoretical Quantitative Analysis

Statistical Analysis

P-value

LASSO

Support Vector Machines (SVM)

Random Forest

Final Rankings

Methodology

Features

Baseline

Learning Algorithms

Neural Networks

Deep Support Vector Machine

12 SHIHbot: Sexual Health Information on HIV/AIDS, chatbot

Motivating Social Problem

Overview

Data Gathering

NPCEditor

Deployment

Messenger as the User Interface

Intermediate Web Service as the Proxy

NPCEditor as the Backend

Evaluation Metrics

Linguistic-Driven

Online-Driven

Social-Driven

Alpha Test

Data

Observations

Future Work and Directions

Support for Additional Platforms

Obtain Additional Information

Add Additional Functionality

Analyze User Data to Improve SHIHbot

13 Ethics and Artificial Intelligence in Public Health Social Work

Introduction.

Case Study: Adapting TND Network for Homeless Youth

Beneficence Problems

Moral Duties and Beneficence Problems

A Framework for Resolving Conflicts

Operationalizing the Framework

Notes

Glossary

Index.

Notes:

Title from publisher's bibliographic system (viewed on 16 Nov 2018).

Includes bibliographical references and index.

ISBN:

1-108-69146-3

1-108-66901-8

1-108-64522-4

OCLC:

1060524972