Fabrizio Cominetti

Logo

LinkedIn | GitHub

Typing SVG
๐Ÿ“Œ Data Science at University of Milano-Bicocca
๐Ÿ“Œ Digital Editor at AC Milan

Portfolio


SELECT * FROM sections

University

print("Explore University Projects")

Masterโ€™s Degree in Data Science | University of Milano-Bicocca | GitHub

Collection of projects realized for the university courses of the MS at UniMiB (2021-2023).

Social Media Analytics

My complete implementation of assignments and projects in Social Media Analytics by UniMiB (2023).

Open Research Poster View on GitHub

Smart Working - Social Network and Content analysis: social media analytics project realized with .py

This project focuses on the analysis of the Twitter community and sentiment regarding the enunciated topic. Through the API, 2667 tweets related to the selected keywords were collected from December 26, 2022 to January 5, 2023. After an initial preprocessing phase based on text mining techniques, a Social Network Analysis (Nodes Degree, Assortativity, Community Detection) and a Social Content Analysis (Sentiment Analysis, Emotion Recognition) were performed.

SMAn

Financial Market Analytics

My complete implementation of assignments and projects in Financial Market Analytics by UniMiB (2022).

Open Research Poster View on GitHub

Portfolio Analysis: financial market analytics project realized with .py

In this group work, we want to better understand the structural characteristics that risk brings to real investment portfolios. In order to understand this empirically, we need to build real portfolios that are concentrated/tilted with respect to a specific level and kind of risk.

The results showed that the portfolios created over-performed compared to the index during the period under consideration. However, the same results also showed higher volatility than that of the index.

finm


Machine Learning

My complete implementation of assignments and projects in Machine Learning and Decision Models by UniMiB (2022).

Open Research Poster View on GitHub

Climate Change - Temperature Prediction: machine learning classification project realized with Knime

The project consists in the analysis of various regression models applied to a dataset containing environmental information referred to the global temperature from January 1750 to November 2015.

wML

Data Management

My complete implementation of assignments and projects in Data Management by UniMiB (2022).

Open Research Poster View on GitHub

Marvel Graph Database: data management project realized with Neo4j

The project realized aims to create a graph-database containing the relationships between the various Marvel products. For the realization of this project we have chosen a non-relational graph-database, built through the use of Neo4j. We have realized this project starting from two data sources, API and web scraping, that we have then integrated and on which we have then performed a quality check.

The final database is complete of the various relationships between characters, movies and comics of the Marvel world, moreover each node contains several information about its nature.

dman


Data Visualization

My complete implementation of assignments and projects in Data Visualization by UniMiB (2022).

Open in Browser Open Research Poster View on GitHub

Film - Rating and Duration Time: data visualization project realized with Bokeh

In today's attention economy and society is it true that people also prefer shorter films? This consideration guided us in the realization of this project and allowed us to determine the following research question: do shorter films generally receive better ratings?

We have therefore used the datasets provided by IMDb to answer the research question through an interactive visualization.

dviz

My complete implementation of assignments and projects in Text Mining and Search by UniMiB (2023).

Open Research Poster View on GitHub

IMDB Reviews - Text Classification and Clustering: text mining and search project realized with .py

In this project, user reviews from the IMDB platform were analyzed through the use of text mining techniques. After carrying out an initial phase of text processing and text representation, the project continued with the classification of the reviews, through some text classification techniques - such as Support Vector Machines (SVM), Multilayer Perceptron (MLP), and Logistic Regression. Next, a text clustering phase was carried out through the use of two algorithms: DBSCAN and k-means.

TMS

Data Science Lab on Smart Cities

My complete implementation of assignments and projects in Data Science Lab on Smart Cities by UniMiB (2023).

Open Research Poster View on GitHub

Mobility in Northern Sardinia: ds lab for smart cities project realized with .py

In this project, we're diving into how people move around in the northern part of Sardinia, starting from the arrivals by airports and ports and focusing mainly on public transport stops to reach key facilities. Several indicators are calculated, accompanied by data visualizations. As for airports and ports, the project aims to examine the flows both on a seasonal basis, differentiating by the airport of arrival and departure, and by checking the number of domestic and foreign tourists. The analysis then focuses in particular on the situation of public transportation for tourists, analyzing their current conditions and the possibility of reaching popular destinations such as beaches, as well as for residents and connections to more populated areas.

LSC

Data Science

print("Explore Data Science Projects")

Data Science | Data Analytics | GitHub

Collection of various data science projects (2021-now).

FantaSanremo

View on GitHub

FantaSanremo Trend and Sentiment Analysis project; realized with .py

The "Sanremo Festival" is becoming increasingly social. And so, companies and content creators also follow the events of the kermesse to stay up-to-date on any trends to be exploited and opportunities for growth. In recent years, "FantaSanremo" - a game that involves viewers but also the singers in the competition - has become very popular among young and old enthusiasts. This project focuses on analyzing tweets related to the 2023 edition of #FantaSanremo, with the goal of analyzing content, the most relevant hashtags and sentiment related to the topic of interest.

fantasanremo

fantasanremo

Billboard Hot-100 EDA

View on GitHub

Exploratory Data Analysis of Billboardโ€™s Hot-100 Weekly Charts; realized with SQLite and Tableau

The Billboard Hot 100 is the music industry standard record chart in the United States for songs, published weekly by Billboard magazine. I've performed an exploratory data analysis of the Billboard dataset containing all the charts from 1958 to today. Finally, I've pictured some results obtained in a dashboard.

billboard


Serie A xG

View on GitHub Open Notebook

Web Scraping and Exploratory Data Analysis of FBrefโ€™s Serie A Expected Goals Standing; realized with .py

Web scraping of data from the FBref website to analyze the current state-of-doing of the Serie A Italian championship of football with requests and BeautifulSoup. At the time of the project, the break for the national teams leaves the league as is, with 8-9 games to play and with many situations to be determined. These last matches could make the difference between an extraordinary season and an undertone season, or a season where the team achieves its goals.

The dataset contains info on classic football data with the addition of xG, a metric used to determine the expected goals, for and against, a team.

serieA


The Office

View on GitHub View on Kaggle

Web Scraping of every lines from all seasons of the TV show The Office followed by an Exploratory Data Analysis and a Sentiment Analysis of the extracted data; realized with BeautifulSoup, PowerBI and VADER

The Office is an American mockumentary sitcom television series that depicts the everyday work lives of office employees in the Scranton, Pennsylvania branch of the fictional Dunder Mifflin Paper Company. In this project, I've firstly scraped all the lines of the TV show in a CSV file, then I've performed in two different jupyter notebooks an exploratory data analysis and a sentiment analysis of all the lines.

theOffice

Morbius

View on GitHub

Morbius & Moon Knight Sentiment Analysis project; realized with Tweepy and VADER

Morbius is a film of the Marvel catalogue. I've collected tweets containing the keyword 'Morbius' with the Twitter API to analyze the sentiment that followed the release of the film. The data has been collected for the subsequent 4 days and then compared in a kdeplot.

I've also collected data for the Moon Knight TV series, also from the Marvel Cinematic Universe. The data has been collected for two days after the release of the series and then compared to the first two days after the release of Morbius.

morbius


Football Analytics

print("Explore Football Analytics Projects")

Football Analytics | GitHub

Collection of various football analytics projects and visualizations (2021-now).

xG

Serie A - xG Lollipop: football analytics project realized with .py and matplotlib

xG

xG Rolling Plot: football analytics project realized with .py and matplotlib

teamNemesis

Serie A - Team Nemesis: football analytics project realized with .py and matplotlib


ยฉ Fabrizio Cominetti