Gaetano Rossiello

PhD Student

University of Bari "Aldo Moro"
Department of Computer Science
Via E.Orabona, 4 - 70126 BARI, Italy
Phone: +39 080 5442298
e-mail: gaetano.rossiello[_AT_]uniba.it

Research Interests | Short CV | Summer Schools | Publications | Conferences | Tools

Main Research Interests

• Natural Language Processing
• Information Retrieval
• Recommender System
• Machine Learning
• Deep Learning

Short CV

• NOVEMBER 2015 - Start the Ph.D. in Computer Science at University of Bari. My supervisor is Prof. Giovanni Semeraro.
• APRIL 2015 - I received a Master's Degree with full marks and honors in Computer Science from University of Bari. I discussed a thesis titled: ″Semantic Query Suggestion: a probabilistic query auto-completion model based on dependencies between concepts″.
• MARCH 2005 - I received a Bachelor's Degree with full marks and honors in Computer Science from University of Bari. I discussed a thesis titled: ″Personalized Relevance Feedback: a technique to introduce the user preferences into a information retrieval model″.

Summer Schools

• BigDat 2017. International Winter School on Big Data - Bari, Italy
• LxMLS 2016. Lisbon Machine Learning School - Lisbon, Portugal
• RegML 2016. Regularization Methods for Machine Learning - Genoa, Italy

Publications

• G. Rossiello, P. Basile, G. Semeraro. Centroid-based Text Summarization through Compositionality of Word Embeddings. MultiLing 2017 Workshop in EACL 2017. Summarization and summary evaluation across source types and genres.
• G. Rossiello. Neural Abstractive Text Summarization. Doctoral Consortium at International Conference of the Italian Association for Artificial Intelligence. AI*IA 2016.
• G. Rossiello, P. Basile, G. Semeraro, M. Di Ciano and G. Grasso. Improving Neural Abstractive Text Summarization with Prior Knowledge. 1st Workshop on Deep Understanding and Reasoning: A challenge for Next-generation Intelligent Agents. URANIA 2016.
• C. Greco, A. Suglia, P. Basile, G. Rossiello and G. Semeraro. Iterative Multi-document Neural Attention for Multiple Answer Prediction. 1st Workshop on Deep Understanding and Reasoning: A challenge for Next-generation Intelligent Agents. URANIA 2016.
• P. Basile, A. Caputo, G. Rossiello and G. Semeraro. Learning to Rank Entity Relatedness through Embedding-Based Features. Proceedings of the 21st International conference on the Application of Natural Language to Information Systems. NLDB 2016
• P. Basile, A. Caputo, M. Di Ciano, G. Grasso, G. Rossiello and G. Semeraro. SEPIR: a SEmantic and Personalised Information Retrieval Tool for the Public Administration based on Distributional Semantics. International Journal of Electronic Governance (IJEG)
• P. Basile, A. Caputo, M. Di Ciano, G. Grasso, G. Rossiello and G. Semeraro. Distributional Semantics for Intelligent Information Access of Documents within Public Administration. I-Cities 2016
• A. Caputo, P. Basile, M. de Gemmis, P. Lops, G. Semeraro and G. Rossiello. SABRE: a Sentiment Aspect-Based Retrieval Engine. Information Filtering and Retrieval. DART 2014: Revised and Invited Papers.

Conferences

• AI*IA 2016. The 15th International Conference of the Italian Association for Artificial Intelligence. 29 November 2016 - 1 December, Genoa (IT)
• URANIA 2016. 1st Workshop on Deep Understanding and Reasoning: A challenge for Next-generation Intelligent Agents, AI*IA 2016. 28 November 2016, Genoa (IT)
• NLDB 2016. 21st International conference on the Application of Natural Language to Information Systems. 22-24 June 2016 at the University of Salford, MediaCityUK Campus, Manchester (UK).
• FINREC 2016. 2nd International Workshop on Personalization and Recommender Systems in Financial Services. 16 June 2016, Bari (IT).
• IIR 2012. 3nd Italian Information Retrieval Workshop. 26-27 Junuary 2012, Bari (IT).

Tools

• Text Summarizer - A centroid-based method for extractive text summarization which exploits the compositional capability of word embeddings. GitHub
• PROMETHEUS - Prometheus is a tool for auto-completion where the queries are dynamically generated from a document collection. It is based on a Probabilistic Graphical Model, called Factor Graph, able to model the dependencies between concepts. It is useful in contexts (eg. enterprise environments) where query logs are not available.