Authors:

Anna Kendzior

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

This project focuses on how to develop software, specifically R packages, that can most reliably extract references to different government agencies from a long and cluttered text. The aim of this project is to prevent repetition of tedious tasks related to sorting through text data by producing a natural language algorithm that is capable of returning references to any alias of a government agency (for example, both "DOJ" and "Department of Justice" return the same result). | tested and improved a hand-built agency lookup table provided to me by my mentor, listing government agencies and all known aliases/references and tested several approaches for pulling the data into an R environment and formatting it in a way that can be easily understood. After cleaning that data, | was able to use an R package developed by another member of my lab to develop my own package that extracts all references to government agencies from a text. Next steps for this project include submitting it for publication on CRAN and testing by researchers from other fields to determine whether the software works for other research tasks. This project will reduce the time and effort required for researchers to do their work, allowing them to cut out time spent on menial tasks and focus on the meaningful parts of their work.

Source:

Chicago Area Undergraduate Research Symposium

Topics:

No topics listed

Co-authors:

Anna Kendzior