Description: Create a better search functionality of govt contracts using the data available through is the publicGSA website that hosts all open source govt contracting information.

The SIG team performed the following activities:

Data Collection and Feature Extraction

Data Loading, Cleaning, and Accessing Using API Calls

Natural Language Processing and Model Input Creation

API Development to Access the Most Up-to-Date Contracts/Opportunities

Machine Learning Application Development

Building and Testing Clusters with the KMeans algorithm

Recommender Creation and New Contract Cluster Classification Using K Nearest Neighbors

Cloud Hosting and Model Runs

GUI Creation to Wrap Models


Results: Team delivered a graphical user interface (GUI) for a new search engine that delivered a refined list of results using machine learning to identify contracts based on language as opposed to just keywords or codes. The enhancement significantly improved alignment to key contract/opportunity search criteria. The team was able to leverage cloud hosting for data and models significantly enhancing the ability to process and analyze the large data set.