- Working on projects related to cyber crime investigation activities, with particular focus on technologies and techniques enabling the extended analysis of data coming from BIG DATA sources, specifically social networks. In this context, I am: 1) Designing complex BIG DATA architectures which include both batch processing and streaming processing software modules. 2) Using the Hadoop software stack (HDFS, HBase, etc.). 3) Using the most important tools currently available on the market for BIG DATA processing: Apache Spark (both in batch mode and in streaming mode), Spark MLlib, Spark SQL and Dataframe, Apache Kafka and Elasticsearch. 4) Using various machine learning methods, specifically targeted for the textual analysis. - Working on deep learning techniques specifically targeted to textual processing (word embeddings, doc embeddings, LSTM, etc) and usable together with classic NLP techniques and algorithms like POS tagging, stemming, etc. These set of skills are used in the context of social media data analysis in order to perform various types of activities like tweets clustering and classification, textual topic modeling, textual language identification, textual similarity, hate speech detection, prediction of user political leaning, etc. .

- Working on development, testing and maintenance of commercial software VCS, developed at NeMIS lab of ISTI-CNR in Pisa- Designing and developing a new software for automatic web sites classification and built around technologies based on machine learning, text mining and statistical analysis.- Working on native language identification task for next "NLI Shared Task 2014"​ (see https://sites.google.com/site/nlisharedtask2013/home for 2013 edition).- Working on designing and developing a new distributed computing platform. This core technology will be used to build a new machine learning platform for BIG DATA with capabilities for batch and online processing- Testing and working with in BIG DATA contexes with several data processing distributed framework: Apache Hadoop, Apache Spark, Apache Flink.- Developed distributed implementations of AdaBoost.MH and MP-Boost machine learning algorithms using Apache Spark. The software is available at https://github.com/tizfa/sparkboost.- Working on a library of Apache Spark ML components specialized for Natural Language Processing (NLP) tasks. The code is available at https://github.com/tizfa/nlp4sparkml .

IT consultant involved in several activities:- PQuadro S.r.l., Navacchio (PI)Design and development of software for the footwear world.- Omnia Group Srl, Sesto Fiorentino (FI)Software analyst/developer at Coop Italia, Prato (PO): analysis and evolution of the informative processes used internally at Coop for data invoice generation. Improvement of existent procedures and rewrite in Java of old software code originally written in Cobol.- ISTI - CNR, PisaFor High Performance Computing (HPC) Lab, design and development of a software prototype for mobile devices (specifically Android) which allows the users to access multimedia archives stored locally on the device, in absence of Internet connectivity, and with the possibility to search the archive content by using a compact "real" search engine directly available on the mobile device. Other activities- Design and development of a CRM (Customer Relationship Management) for an insurance company. The software is a multiuser Web application with ACL fine-grain access type to its features, and it has been developed using the software stack Java/J2EE for the backend and Vaadin for the frontend.- Teacher training for various .NET courses targeting IT companies. The main technologies touched were C#, Ado.NET, LINQ, Entity Framework, WCF, Asp.NET MVC, Windows Forms, etc.- Developed several Android apps using Android SDK and Eclipse as development tools.

Software engineer at PQuadro s.r.l., Santa Maria a Monte (PI)Design and development of a CAD 2D software (in Java/Swing) for the design and manufacturing of shoes models (fashion sector). Involved also in other minor software projects always related to fashion sector and based on Java/Swing/JDBC technologies.

Research fellow at NeMIS lab (Network Multimedia Information Systems) with research topics related to Text Categorization, Machine Learning and Information Retrieval.Worked in the following projects:- VCS (Verbatim Coding System) project (http://nemis.isti.cnr.it/product/verbatim-coding-system)Involved in the design and implementation of a CRM for the english bank EGG. In particular, I made the software module of opinion extration of EGG customers related to products/services offered by the bank. The bank periodically submits open-text surveys to customers to get opinions about the customer satisfaction/unsatisfaction. Automatically monitoring this aspect is crucial to improve the quality of the offered services/products or not to lose customers. - NeP4B "Network of Peers for Business" project1. Design and development of an automatic classification algorithm for images in a custom taxonomy. The algorithm can use any one of the 7 MPEG-7 visual descriptors to learn a classification model to be used to automatically classify unlabeled images.2. Design and development of an automatic classification algorithm for images based on committee principle. Each internal committee classifier can be of type as described in point 1.3. Experimentation on several strategies to use while selecting negative examples in hierarchical text classification problems.4. In the context of text categorization, usage of clustering techniques in order to reorganize hierarchical taxonomies generated by humans. The main idea is to change the original structure of the taxonomy into a new one which statistically better represents the input data and therefore can be easily analyzed by a machine learning algorithm.

Research fellow at ISTI-CNR, Pisa in the lab Network Multimedia Information System (NEMIS) with research interests in Text Categorization, Machine Learning e Information Retrieval.Worked in the project OPTEX "Opinion Mining from Online Text", doing the following activities:1. Design and development of a modular and extensible Java software for the generation of automatic text classifiers suitable for topic or genre classification.2. Design, implementation and experimentation of a new machine learning algorithm based on boosting (named MP-Boost) for the generation of automatic text classifiers suitable for flat taxonomies.3. Design, implementation and experimentation of a new machine learning algorithm based on boosting (named TreeBoost) for the generation of automatic text classifiers suitable for hierarchical taxonomies.

Research fellow at High Performance Computing lab of ISTI-CNR with main research topics inweb search engine caching, information retrieval e grid computing.In project PNR 2001-2003 "Piattaforme abilitanti per griglie computazionali a elevate prestazioni orientate a organizzazioni virtuali scalabili", I was involved in several activities:1. Analysis and study of the current state of the art technologies used in grid computing.2. Analysis and study of the main software libraries available to write applications grid-oriented.3. Analysis and study of Globus (versions 2.x.and 3.x), the reference platform to implement computational grids.4. Extended analysis of XCat, a grid computing environment component-oriented used to write distributed applications.5. Analysis and design of a software module able to handle the computational resources available in a computing grid.

Research fellow at High Performance Computing lab of ISTI-CNR: extension of the work done in Master's Degree thesis (Caching on web search engines).

Academic year 2004-2005: teaching assistant for the course "Laboratory of concurrent programming" of the Master's Degree in Computer Science at University of Pisa.

Academic year 2003-2004: teaching assistant for the course "Laboratory of concurrent programming" of the Master's Degree in Computer Science at University of Pisa.

During the university path, I did the following technical exams:- Exam "Information Processing Systems: Network of Computers 2"Development in C of a network protocol based both on TCP/IP and UDP/IP and running on Linux systems. - Exam "Operating Systems 2"Design and development in C of a distributed I/O library to be used in clusters composed by Linux machines. - Exam "Programming laboratory: concurrent and object-orient programming"Design and development in Java of a distributed system for hotels reservation.- Exam "Information Processing Systems: parallel applications"Design and development in C of a scientific application using the parallel computing library PVM.- Exam "Computing architectures: laboratory of massive parallel systems"Design and development in C of a scientific application using the library MPI and tested on a Cray T3 supercalculator.- Exam "Human-machine interfaces: user interfaces programming"Design and development in C++ of a 3D modeling software using the library OpenGL.