Home


Hello, my name is

Matteo Ciniselli


I'm a PhD student at the Università della Svizzera Italiana in Lugano.

I'm part of the SEART research group led by Gabriele Bavota and I'm working on the DEVINTA project.

My research aims at using Deep Learning Models to help developer write code.

Education and Professional Experience

Education

PhD in Computer Science

May 2020 - Present

Università della Svizzera Italiana, Lugano

During my PhD, I'm investigating the usage of Deep Learning models to automatically generate code.

Master in High Apprenticeship in Analytics and Business Intelligence

July 2015 - July 2017

Polytechnic of Milan, Milan

During these two years, I studied topics related to Big Data and Machine Learning. I released a tool using MondoDB and NodeJS able to suggest the choice of an order for a tender

Master Degree in Mathematical Engineering, Statistical Address

October 2012 - April 2015

Polytechnic of Milan, Milan

During these years, I studied statistics at 360 degrees, enriching my knowledge with Optimization, Data Analysis and Programming courses.

Bachelor's Degree in Mathematical Engineering

September 2009 - September 2012

Polytechnic of Milan, Milan

During my Bachelor's Degree, I studied Mathematical Analysis, Statistics, Computer Science, Physics and Probability.

Professional Experience

Data Scientist @ Machine Learning Center Of Excellence

March 2019 - April 2020

Accenture S.p.A., Milan

I realized different tools for Computer Vision released in portable services, using Flask with HTML and CSS. Creation of a NodeJS server that uses DialogFlow API to understand the intent of the sentences and allows the user to interact vocally with the washing machine. Creation of a chatbot deployed in different Docker containers.

.NET Developer

October 2017 - February 2019

Aglea S.r.L., Milan

Realization of internal tools in C# and study of authorization security in SAP. Production of a tool to manage activities in my company (purchasing, invoicing, ..) based on a SQL database.

Business Intelligence Consultant

January 2015 - September 2017

NextInt S.a.s., Milan

Use of RapidMiner and R to solve different optimization problems. Generation of dashboards with QlikView to help customers seeing their data in order to take business decisions.

Publications

An Empirical Study on the Usage of BERT Models for Code Completion

M. Ciniselli, N. Cooper, L. Pascarella, D. Poshyvanyk, M. Di Penta, G. Bavota

Mining Software Repositories (MSR 2021), 12 pages

An Empirical Study on the Usage of Transformer Models for Code Completion

M. Ciniselli, N. Cooper, L. Pascarella, A.Mastropaolo, E. Aghajani, D. Poshyvanyk, M. Di Penta, G. Bavota

IEEE Transactions on Software Engineering (TSE 2021), 20 pages

To What Extent do Deep Learning-based Code Recommenders Generate Predictions by Cloning Code from the Training Set?

M. Ciniselli, L. Pascarella, G. Bavota

Mining Software Repositories (MSR 2022), 12 pages

Source Code Recommender Systems: The Practitioners' Perspective

M. Ciniselli, L. Pascarella, E Aghajani, S Scalabrino, R Oliveto, G Bavota

International Conference on Software Engineering (ICSE 2023), to appear

To What Extent do Deep Learning-based Code Recommenders Generate Predictions by Cloning Code from the Training Set?

A Mastropaolo, L Pascarella, E Guglielmi, M. Ciniselli, S Scalabrino, R Oliveto, G Bavota

International Conference on Software Engineering (ICSE 2023), to appear

Contact Me





Istitutional Email: matteo.ciniselli[at]usi[dot]ch

Office: Campus EST, Section D, Office D3.10 (Level 3), Via la Santa 1, 6962 Viganello