DYNAMIC INSTITUTION REPORT GENERATOR IN EDUCATION TODAY

Smith, Nathan

Permalink

https://hdl.handle.net/2142/124874 Copy

Description

Title

DYNAMIC INSTITUTION REPORT GENERATOR IN EDUCATION TODAY

Author(s)

Smith, Nathan

Issue Date

2022-05-01

Keyword(s)

Higher Education; Application Development; Relational Databases; Web Design

Date of Ingest

2024-10-15T10:23:48-05:00

Abstract

In the Education Today project, along with other higher education ranking/compare websites, comparisons between institutions and faculty require keywords as input–a slow process that comes with a loss of accuracy when comparing entities. Previous methods to determine similarity use a simple total using maximum sum over a graph using Dijkstra’s algorithm; however, we use a function of average and difference for higher accuracy. This new software development and study aims to compare institutions accurately and generate a meaningful report without the process of inputting keywords. Additionally, this software is to be scalable and extendable as new keywords enter the computer science ecosystem. Instead of having individual keywords as input, we use configurations of major field to subfield mappings. Grouping keywords into representative fields is another problem, so we develop a framework for clustering keywords in the database. To allow customization while reducing the time spent choosing keywords, we develop configurations of grouped fields as an input to compare two or more institutions. We use existing ranking functions, which take data-mined keywords for every publication and considers similar keywords based on the Normal Pointwise Mutual Information. By generating a configurable and extendable comparison, we allow quick yet extensive comparison between higher education.

Type of Resource

text

Language

eng

Owning Collections