Normalized Relational Dataset of AI & Data Science University Programs (2020-2024)
Description
This dataset contains a fully normalized relational database describing university level characteristics, AI/Data Science academic programs, admission requirements, and annual institutional statistics for the years 2020-2024. The goal of the dataset is to provide a clean, structured, and reproducible representation of higher education program information that can be used for analysis, querying, and relational modeling. The dataset includes four tables: University, Program, AdmissionRequirements, and AnnualUniversityStats. Each table represents a distinct entity group and is structured according to relational database normalization principles. The University table contains static institutional attributes such as name, state, institution type, campus size, and website. The Program table contains program-level attributes including degree level, year first offered, focus area, and tuition. The AdmissionRequirements table stores minimum GPA, standardized test requirements, and minimum test scores for each program. The AnnualUniversityStats table contains year-specific institutional metrics such as national ranking, total student population, number of programs, and average tuition. All data was collected from publicly available sources including IPEDS College Navigator, university factbooks, institutional research dashboards, US News Education, and official program websites. Because universities do not consistently publish program-level outcomes or historical data for every year, the dataset focuses on attributes that are reliably available across institutions.
Files
Steps to reproduce
The dataset was created through a multi-stage process involving data collection, cleaning, normalization, schema design, SQL implementation, and export. The following steps describe the full workflow so that another researcher can reproduce the dataset or rebuild the database from the original sources. 1. Data Collection: University-level information was collected from IPEDS College Navigator, including institution type, state, campus size, and total enrollment. Program-level information such as degree level, program name, year first offered, and tuition was collected from a combination of official university program websites and the IPEDS College Navigator. Admission requirements were gathered from program admissions pages and the IPEDS College Navigator, including minimum GPA, standardized test requirements, and minimum test scores. Annual university statistics (2020-2024) were collected from IPEDS, university factbooks, and US News Education, including national ranking, total student population, number of programs, and average tuition. 2. Data Cleaning: All collected data was transferred into Excel and cleaned to ensure consistency. Missing values were documented and tuition values were converted into numeric format. University names and program names were standardized to avoid duplicates. 3. Database Design: Four entities were identified: University, Program, AdmissionRequirements, and AnnualUniversityStats. Functional dependencies were analyzed to ensure each table satisfied 3NF. Primary keys and foreign keys were assigned: university_id links University to Program and AnnualUniversityStats, while program_id links program to AdmissionRequirements. 4. SQL Implementation: A MySQL database named ai_ds_programs was created. Four tables were created using CREATE TABLE statements with primary and foreign key constraints. Data from Excel was inserted using INSERT INTO statements. A final joined dataset was generated using a SELECT query joining all four tables. 5. Export: Each normalized table was exported from MySQL as a CSV file. These CSV files were uploaded to Mendeley Data as the final dataset. The above workflow ensures that any researcher can reproduce the dataset by following the same steps and using the same publicly available data sources.