Michael Rossetti

Text Embeddings for User Classification in Social Networks

We use text embeddings, including those produced by large language models from OpenAI, to classify users based on the content of their social media posts. Social media companies can use these methods to identify bots and monitor users that are likely to spread disinformation.

• Code
• Website
• Conference Talk (INFORMS 2023)

Bots, Disinformation, and the First Trump Impeachment

With co-author Tauhid Zaman from Yale University, we analyze the role of automated accounts called "bots" in spreading disinformation across social networks.

• Paper (PLOS ONE)
• Pre-print (Arxiv)
• Code
• Data
• Website
• Conference Talk (INFORMS 2021)

Adjunct Professor

Design and teach courses in computer programming, software development, and data science — for the Operations and Analytics area, as well as the Computer Science department.

• Georgetown Faculty Profile
• Data Science Research Group

Adjunct Assistant Professor

Design and teach summer courses in software development — for the department of Technology, Operations, and Statistics.

Lecturer

Taught courses in databases and management information systems — for the department of Information Systems and Technology Management.

AlphaVantage API

Thanks to AlphaVantage for providing my students with academic access to their amazing financial data API!

• MSB Electives Focus on Data, Georgetown McDonough News, September 2018.
• GSA Hackathon 2: twice as big and open to govvies, Federal Computer Week (FCW), October 2015.

Open Source Software Developer and Evangelist

Promote adoption of open source software and reproducible research across the university, as a member of the Open Source Program Office (OSPO).

Presidential Innovation Fellow

Created prototype web applications and wrote technical documentation to help Federal government developers adopt a shared identity management platform. Developed web applications in languages like Ruby, Node.js, and Go.

• Presidential Innovation Fellowship Profile

Data Engineer + Director of Analytics

Directed hiring and operations of a small team of data analysts, engineers, and scientists. Implemented strategic vision for a company-wide business intelligence (BI) and analytics program, including the identification, prioritization, measurement, and ongoing self-service monitoring of key performance indicator (KPI) metrics. Developed a data warehouse application in Ruby on Rails, to extract, transform, and load (ETL) data from production databases and external APIs into a single, integrated SQL database environment. Developed interactive data dashboards in Tableau and JavaScript.

Media Data Analyst + Engineer

Analyzed political media and polling data, including over $750M in ad buys, on behalf of US Presidential and Congressional campaign clients (i.e. Obama for America, DSCC, DCCC). Identified qualitative and quantitative trends in opponents' media and messaging strategies. Compared media expenditures against polling results to provide data-driven media-buying insights. Directed and oversaw operations of an internal campaign video blogging system to inform campaign leadership and staff across the country about media and messaging trends in their respective localities. Contributed to a Ruby on Rails application, wrote and maintained a library of SQL scripts, and developed interactive data dashboards in Tableau and JavaScript.

Lead Polling Analyst

Analyzed quantitative and qualitative results of campaign-sponsored polls and focus groups. Designed question language for US nationwide and "battleground state" surveys. Wrote SQL scripts to integrate internal and external polling data into a unified dataset for analysis. Produced campaign polling insights in the form of Tableau data dashboards.

Technology Risk Consultant

Advised clients at the Internal Revenue Service (IRS) on projects to mitigate their risk of cyber threats and ensure their systems’ compliance with Federal information security requirements. Created and maintained an MS Access database of systems and security metrics. Streamlined reporting processes by developing automated data dashboards in Crystal Reports.

Business Intelligence Analyst

Identified and prioritized key performance indicators for measuring workforce performance and productivity across the department. Wrote and maintained SQL scripts, and created automated reporting dashboards in MS SQL Server.

The George Washington University

Complete courses in data science and machine learning. Teach technical skills as a graduate teaching assistant and adjunct faculty member. Conduct research in applied machine learning, focusing on natural language processing (NLP) and large language models (LLMs).

The George Washington University

Completed advanced courses in databases, systems design, computer programming, and human-computer interaction. Taught technical skills as a graduate teaching assistant.