AI and Data Science Academic Programs in U.S. Universities (Normalized Relational Dataset)
Description
This dataset contains organized information about universities in the United States that offer programs in Artificial Intelligence and Data Science. I created it as part of my DATA 6200 Data Management course to support database design, analysis, and research on academic programs in these fields. All of the information in the dataset comes from publicly available sources. These include NCES College Navigator (https://nces.ed.gov/collegenavigator/ (nces.ed.gov in Bing)), U.S. News Best Colleges (https://www.usnews.com/best-colleges (usnews.com in Bing)), DataScienceGraduatePrograms.com (https://www.datasciencegraduateprograms.com/), individual university program pages, GradSchools.com (https://www.gradschools.com/), NCES IPEDS (https://nces.ed.gov/ipeds/), and university faculty directories. These sources provide reliable details about universities, degree programs, admissions requirements, and yearly institutional statistics. The dataset is structured into five tables: University, DegreeProgram, AdmissionRequirements, YearlyUniversityStats, and Faculty. Each table represents a different part of the overall topic. The University table includes basic institutional information such as state, type, enrollment, and website. The DegreeProgram table lists program names, degree levels, tuition, and duration. AdmissionRequirements includes minimum GPA, test requirements, deadlines, and notes. YearlyUniversityStats contains time‑based data such as tuition, enrollment, and national rankings for multiple years. The Faculty table includes faculty names, departments, research areas, and email addresses. The goal of this dataset is to provide a clean, normalized structure that can be used for academic analysis, database modeling, or research on trends in AI and Data Science education. Because the data is organized in Third Normal Form, it can be easily imported into a relational database and used for queries, visualizations, or further study. This dataset is intended for educational and research purposes.
Files
Steps to reproduce
To reproduce this dataset, the first step is to gather information from the same public sources used in this project. These include NCES College Navigator, U.S. News Best Colleges, DataScienceGraduatePrograms.com, university program pages, GradSchools.com, NCES IPEDS, and university faculty directories. These websites provide details about universities, their academic programs, admissions requirements, and yearly statistics. A researcher should start by selecting at least twenty universities that offer programs related to Artificial Intelligence or Data Science. It is helpful to include a mix of public and private institutions from different states. Next, collect the basic university information such as name, state, type, total enrollment, and website. Then gather program‑level information, including the program name, degree level, tuition, and duration. Admission requirements such as minimum GPA, test requirements, and application deadlines can be found on university admissions pages or GradSchools.com. Yearly statistics like tuition, enrollment, and national rankings can be collected from NCES IPEDS and U.S. News Rankings. Faculty information can be taken from university faculty directories, which usually list names, departments, research areas, and email addresses. After collecting the raw data, the next step is to design the database structure. The dataset should include five tables: University, DegreeProgram, AdmissionRequirements, YearlyUniversityStats, and Faculty. Each table should have a primary key, and foreign keys should be used to connect related tables. For example, DegreeProgram should reference UniversityID, and AdmissionRequirements should reference ProgramID. The structure should follow Third Normal Form to avoid redundancy and ensure that each attribute depends only on its primary key. Once the design is complete, create the tables in a relational database system such as MySQL. Insert the collected data into the tables using SQL INSERT statements. After the data is inserted, run SELECT queries to verify that the tables contain the correct information. When everything looks correct, export each table as a CSV file. These CSV files form the final dataset. To reproduce the dataset, another researcher would simply follow the same steps: gather data from the same public sources, create the same table structure, insert the data, and export the tables as CSV files. This ensures that the dataset can be recreated accurately and consistently.
Institutions
- Wentworth Institute of TechnologyMassachusetts, Boston