Four homework projects for the Advanced Information Systems and Big Data course (academic year 2025-2026), by Filippo Camossi. Each one runs the same kind of analysis on a different data model: document, time series, graph and MapReduce.
| Project | Dataset | What it does |
|---|---|---|
| MongoDB | Amazon Fine Food Reviews, about 568k reviews | Aggregation pipelines, VADER sentiment analysis with grid search, and a hybrid recommender (collaborative, TF-IDF content and sentiment) |
| InfluxDB | Los Angeles crime data, about 900k records | Batch ingestion, Flux queries on time and area, and online clustering of locations with River |
| Neo4j | ArXiv AI papers | Graph of papers, authors and topics, Cypher analytics, co-authorship paths and weighted Jaccard (Ruzicka) topic similarity |
| MapReduce | Synthetic highway tolls, 100k records | Average toll per vehicle class in 2015 and 2025, implemented with mrjob and with PySpark |
Every folder is independent. It has its own README, requirements.txt, the assignment PDF and the written report (*_Camossi.pdf).
You need Python 3.8+, Jupyter, and the database for the project you want to run (MongoDB, InfluxDB 2.x or Neo4j). The MapReduce project needs Java 8+ for Spark.
git clone https://github.com/SickCiQuattro/SEBD_homework.git
cd SEBD_homework/MongoDB # or InfluxDB, Neo4j, MapReduce
pip install -r requirements.txt
jupyter notebookNote
The MongoDB and InfluxDB datasets are too large for the repository. Each README links to where you can download them.