Part-Time Research Internship Opportunity at NIST
Title: Word Embeddings of Maintenance Work Order Data
Location: NIST, Gaithersburg MD
Salary: $15-$30 per hour, 10-40 hours/week, in the period May 1 2020 to 30 April 2021
The Knowledge Extraction and Application for Manufacturing Operations project (https://www.nist.gov/programs-projects/knowledge-extraction-and-application-manufacturing-operations (new window)) is planning to create GloVe (Global Vectors for Word Representation) and BERT (Bidirectional Encoder Representations from Transformers)-style embeddings specialized for the analysis of technical datasets, containing short sentences, informal language, abbreviations, and technical jargon (e.g., maintenance logs). This effort will begin with the curation of maintenance-related text in the form of technical specifications, maintenance manuals, and maintenance logs. It will then continue with the creation of embeddings from the collected text and their evaluation for use in specific maintenance-related analyses. The goal of this work is to learn differences and similarities between NLP techniques for “normal”, non-technical documents and technical, informal documents. The desired student will have the following attributes:
- Working on or completed BS or MS in Computer Science, Information Systems, Computational Logistics, or related field with a focus on text analysis, natural language processing, data science, and machine learning.
- Demonstrated experience with Python and familiarity with its NLP, data analysis, and machine learning packages.
- Comfortable working with command-line Linux on large scientific server platforms.
- Willingness to work across the full data science lifecycle from data acquisition and curation through analysis and publication.
Application process: Send a short CV, including courses taken and relevant experience, to email@example.com (new window), and fill out an expression of interest form (new window) at https://softmatter.georgetown.edu/nist-professional-research-experience-program-prep/ (new window)