rundstedtzz / mpac_project1 Goto Github PK

Maine Prisoner Advocacy Coalition: Newspaper Text Analysis

Executive Summary

Overview

The Maine Prisoner Advocacy Coalition is planning to campaign on behalf of the reinstatement of parole and organize policy action to support the new parole initiative. In working with the MPAC, we used data from three Maine newspapers to evaluate local public perception toward parole by utilizing text analysis.

The Problem

The bill to reinstate parole in Maine has yet to be signed by the governor. Therefore, the MPAC is planning to campaign on behalf of the bill to increase resident support for the initiative and to assist legislators in drafting an amended bill that will pass once the investigator has been conducted.

Our Research

Our research of opinion mining categorizes language into positive, negative, or neutral sentiments and generates a series of visualizations to highlight trends. These trends identified by our sentimental analysis will help the MPAC make informed decisions when organizing campaigns so they can maximize saliency to the public.

Next Steps

For further research on the public perspective of parole and probation using text analysis, we have the following suggestions. Further analyses can implement stemming to minimize repeatedly counting words of the same stem Further application of machine learning techniques to text analysis Refinement of word association analysis

Summary

There are several limitations to our study, primarily that we assume sentiments articulated in the newspaper articles are representative of general public opinion in Maine. However, we believe that our results can give context to some of the public’s reactions to issues around parole. This report can be vital to campaigning strategy and is a great foundation for future research.

The technical report can be accessed here

The R-shiny app can be accessed here