/jpl-academic-divisions

Primary LanguagePythonMIT LicenseMIT

JPL Academic Divisions

Objective

This project aims to demonstrate Natural Language Processing by answering the following problem:

Can we identify the scientific field a JPL author belongs to?

For this project, we will use Caltech's academic divisions as a way to define a scientific field. Caltech has 6 academic divisions:

  • Biology and Biological Engineering
  • Chemistry and Chemical Engineering
  • Engineering and Applied Science
  • Geological and Planetary Sciences
  • Humanities and Social Science
  • Physics, Mathematics and Astronomy

Thus, the problem changes to the following:

Can we determine which Caltech academic division a JPL author most likely belongs to?

General Tasks

Gathering Publication Data

We must source publication data from Caltech and JPL to accomplish this task. We obtained articles and conference proceedings from Web of Science with the following filter:

  • Organization-Enhanced: Califonia Institute of Technology or NASA Jet Propulsion Lab
  • Dates: 5 date range (2015, 2016, 2017, 2018, 2019)

Here are the number of records found:

  • 21175

We will store all of this data into MongoDB

Data Cleaning and Processing

Determine which fields from the Web of Science output that will be used for the analysis

Calculate TF-IDF vector representation of each document

Create Author document

Create Division document

Perform similarity calculation between JPL Author and Division

Create dashboard to visualize similarity of JPL Author to all Divisions