Lecture Notes in Artificial Intelligence 7689

Size: px

Start display at page:

Download "Lecture Notes in Artificial Intelligence 7689"

Moris Jennings
6 years ago
Views:

1 Lecture Notes in Artificial Intelligence 7689 Subseries of Lecture Notes in Computer Science LNAI Series Editors Randy Goebel University of Alberta, Edmonton, Canada Yuzuru Tanaka Hokkaido University, Sapporo, Japan Wolfgang Wahlster DFKI and Saarland University, Saarbrücken, Germany LNAI Founding Series Editor Joerg Siekmann DFKI and Saarland University, Saarbrücken, Germany

2 Bernardo Magnini Francesco Cutugno Mauro Falcone Emanuele Pianta (Eds.) Evaluation of Natural Language and Speech Tools for Italian International Workshop, EVALITA 2011 Rome, January 24-25, 2012 Revised Selected Papers 13

3 Series Editors Randy Goebel, University of Alberta, Edmonton, Canada Jörg Siekmann, University of Saarland, Saarbrücken, Germany Wolfgang Wahlster, DFKI and University of Saarland, Saarbrücken, Germany Volume Editors Bernardo Magnini Fondazione Bruno Kessler, Via Sommarive 18, Povo (TN), Italy Francesco Cutugno University of Naples, Via Cinthia, Napels, Italy Mauro Falcone Fondazione Ugo Bordoni, Viale del Policlinico 147, Rome 161, Italy Emanuele Pianta CELCT, Via alla Cascata 56/C, Povo (TN), Italy Managing Editors Rachele Sprugnoli, CELCT, Povo (TN), Italy Valentina Bartalesi Lenzi, CELCT, Povo (TN), Italy ISSN e-issn ISBN e-isbn DOI / Springer Heidelberg Dordrecht London New York Library of Congress Control Number: CR Subject Classification (1998): I.2.7, I.2.6, H.2.4, I.5.4, H.3.6 LNCS Sublibrary: SL 7 Artificial Intelligence Springer-Verlag Berlin Heidelberg 2013 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Typesetting: Camera-ready by author, data conversion by Scientific Publishing Services, Chennai, India Printed on acid-free paper Springer is part of Springer Science+Business Media (

4 Preface EVALITA ( is the reference evaluation campaign of both Natural Language Processing and Speech Technologies for the Italian language. The objective of the shared tasks proposed at EVALITA is to promote the development of language technologies for Italian, providing a common framework where different systems and approaches can be evaluated and compared in a consistent manner. This volume collects the final and extended contributions presented at EVALITA 2011, the third edition of the evaluation campaign. EVALITA 2011 is organized collectively by the Italian community on Human Language Technologies, and it is supported both by the NLP Special Interest Group of the Italian Association for Artificial Intelligence (AIxIA) and by the Italian Association of Speech Science (AISV). This EVALITA edition culminated in a workshop organized in Rome during January 24 25, 2012, held jointly with the 8th AISV Congress. Both academic institutions and industrial organizations were invited to participate at EVALITA We received 87 expressions of interest in the 11 tasks proposed (8 for NLP and 3 for speech), with the final number of participants being 30, distributed as follows: Text tasks: Lemmatization (4), Parsing (5), Domain Adaptation (2), Entity Recognition on Transcriptions (4), Cross-Document Coreference (1), Anaphora Resolution (1), Super Sense Tagging (2), Frame Labeling (3). Speech tasks: Automatic Speech Recognition (2), Forced Alignment (3), Student Contest (3). Overall, 23 different organizations participated in EVALITA 2011, out of which five were not Italian and three were nonacademic, confirming the interest in the proposed tasks. After the campaign, most of the datasets were made freely available for research purposes, with the aim of encouraging further improvement of the technologies. We thank all the people and the institutions involved in the organization of EVALITA A special thanks to Manuela Speranza, who took care of most of the managing work at FBK, and to Valentina Bartalesi Lenzi and Rachele Sprugnoli, who managed the EVALITA website and the realization of the proceedings at CELCT. July 2012 Bernardo Magnini Franco Cutugno Mauro Falcone Emanuele Pianta

5 Reviewers The editors express their gratitude to the colleagues listed below for their assistance in reviewing the papers in this volume: Valentina Bartalesi Lenzi, CELCT, Italy Roberto Basili, University of Rome Tor Vergata, Italy Luisa Bentivogli, FBK, Italy Cristina Bosco, University of Turin, Italy Fabio Brugnara, FBK, Italy Felice DellOrletta, ILC-CNR, Italy Roberto Gretter, FBK, Italy Alessandro Lenci, University of Pisa, Italy Alessandro Marchetti, CELCT, Italy MarcoMatassoni,FBK,Italy Alessandro Mazzei, University of Turin, Italy Simonetta Montemagni, ILC-CNR, Italy Antonio Origlia, University of Naples Federico II, Italy Roberto Pieraccini, International Computer Science Institute, USA Massimo Poesio, University of Essex, UK Dino Seppi, University of Leuven, Belgium Maria Simi, University of Pisa, Italy Manuela Speranza, FBK, Italy Rachele Sprugnoli, CELCT, Italy Fabio Tamburini, University of Bologna, Italy Olga Uryupina, University of Trento, Italy

6 EVALITA 2011 Coordination The scientific coordination of EVALITA 2011 was in the charge of Fondazione Bruno Kessler (Bernardo Magnini), the University of Naples Federico II (Franco Cutugno), Fondazione Ugo Bordoni (Mauro Falcone), and CELCT (Emanuele Pianta). The following institutions contributed to the organization of the different tasks of the EVALITA 2011 campaign: CELCT, Italy FBK, Italy ILC-CNR, Italy International Computer Science Institute, USA ITTIG-CNR, Italy Nuance Communications Inc., Italy University of Bologna, Italy University of Essex, UK University of Leuven, Belgium University of Naples Federico II, Italy University of Pisa, Italy University of Rome Tor Vergata, Italy University of Turin, Italy University of Trento, Italy

7 EVALITA 2011 Steering Committee Tommaso Agnoloni, ITTIG-CNR, Italy Valentina Bartalesi Lenzi, CELCT, Italy Roberto Basili, University of Rome Tor Vergata, Italy Luisa Bentivogli, FBK, Italy Cristina Bosco, University of Turin, Italy Fabio Brugnara, FBK, Italy Paolo Coppo, Nuance Communications Inc., Italy Diego De Cao, University of Rome Tor Vergata, Italy Stefano Dei Rossi, University of Pisa, Italy Felice DellOrletta, ILC-CNR, Italy Giulia Di Pietro, University of Pisa, Italy Enrico Francesconi, ITTIG-CNR, Italy Roberto Gretter, FBK, Italy Alessandro Lenci, University of Pisa, Italy Alessandro Marchetti, CELCT, Italy Simone Marchi, ITTIG-CNR, Italy MarcoMatassoni,FBK,Italy Alessandro Mazzei, University of Turin, Italy Simonetta Montemagni, ILC-CNR, Italy Alessandro Moschitti, University of Trento, Italy Antonio Origlia, University of Naples Federico II, Italy Roberto Pieraccini, International Computer Science Institute, USA Massimo Poesio, University of Essex, UK Dino Seppi, University of Leuven, Belgium Maria Simi, University of Pisa, Italy Manuela Speranza, FBK, Italy Rachele Sprugnoli, CELCT, Italy Fabio Tamburini, University of Bologna, Italy Olga Uryupina, University of Trento, Italy Giulia Venturi, ILC-CNR, Italy Sponsoring Institution Loquendo, Turin, Italy

8 Table of Contents Parsing - Dependency Parsing Track The EVALITA Dependency Parsing Task: From 2007 to Cristina Bosco and Alessandro Mazzei Use of Semantic Information in a Syntactic Dependency Parser Leonardo Lesmo Parsit at Evalita 2011 Dependency Parsing Task Matteo Grella and Marco Nicola An Ensemble Model for the EVALITA 2011 Dependency Parsing Task Alberto Lavelli Tuning DeSR for Dependency Parsing of Italian Giuseppe Attardi, Maria Simi, and Andrea Zanelli Parsing - Constituency Parsing Track Looking Back to the EVALITA Constituency Parsing Task: Cristina Bosco, Alessandro Mazzei, and Alberto Lavelli Domain Adaptation for Dependency Parsing Domain Adaptation for Dependency Parsing at Evalita Felice Dell Orletta, Simone Marchi, Simonetta Montemagni, Giulia Venturi, Tommaso Agnoloni, and Enrico Francesconi Experiments in Newswire-to-Law Adaptation of Graph-Based Dependency Parsers Barbara Plank and Anders Søgaard Domain Adaptation by Active Learning Giuseppe Attardi, Maria Simi, and Andrea Zanelli Named Entity Recognition on Transcribed Broadcast News Named Entity Recognition on Transcribed Broadcast News at EVALITA Valentina Bartalesi Lenzi, Manuela Speranza, and Rachele Sprugnoli

9 XII Table of Contents A Simple Yet Effective Approach for Named Entity Recognition from Transcribed Broadcast News Md. Faisal Mahbub Chowdhury A Combination of Classifiers for Named Entity Recognition on Transcription Firoj Alam and Roberto Zanoli The Tanl Tagger for Named Entity Recognition on Transcribed Broadcast News at Evalita Giuseppe Attardi, Giacomo Berardi, Stefano Dei Rossi, and Maria Simi Cross-Document Coreference Resolution of Named Person Entities The News People Search Task at EVALITA 2011: Evaluating Cross-Document Coreference Resolution of Named Person Entities in Italian News Luisa Bentivogli, Alessandro Marchetti, and Emanuele Pianta Exploiting Background Knowledge for Clustering Person Names Roberto Zanoli, Francesco Corcoglioniti, and Christian Girardi Anaphora Resolution Evalita 2011: Anaphora Resolution Task Olga Uryupina and Massimo Poesio UNIPI Participation in the Evalita 2011 Anaphora Resolution Task Giuseppe Attardi, Stefano Dei Rossi, and Maria Simi SuperSense Tagging Description and Results of the SuperSense Tagging Task Stefano Dei Rossi, Giulia Di Pietro, and Maria Simi Super-Sense Tagging Using Support Vector Machines and Distributional Features Pierpaolo Basile SuperSense Tagging with a Maximum Entropy Markov Model Giuseppe Attardi, Luca Baronti, Stefano Dei Rossi, and Maria Simi

10 Table of Contents XIII Frame Labeling over Italian Texts EvalIta 2011: The Frame Labeling over Italian Texts Task Roberto Basili, Diego De Cao, Alessandro Lenci, Alessandro Moschitti, and Giulia Venturi Two Level Approach to SRL Luca Dini, Milen Kouylekov, Marcella Testa, and Marco Trevisan Generative and Discriminative Learning in Semantic Role Labeling for Italian Cristina Giannone Structured Kernel-Based Learning for the Frame Labeling over Italian Texts Danilo Croce, Emanuele Bastianelli, and Giuseppe Castellucci Lemmatisation The Lemmatisation Task at the EVALITA 2011 Evaluation Campaign Fabio Tamburini Italian Lemmatization by Rules with Getaruns Rodolfo Delmonte Data Driven Lemmatization and Parsing of Italian Djamé Seddah, Joseph Le Roux, and Benoît Sagot The Tanl Lemmatizer Enriched with a Sequence of Cascading Filters Giuseppe Attardi, Stefano Dei Rossi, and Maria Simi The AnIta-Lemmatiser: A Tool for Accurate Lemmatisation of Italian Texts Fabio Tamburini Automatic Speech Recognition - Large Vocabulary Transcription Evalita 2011: Automatic Speech Recognition Large Vocabulary Transcription Marco Matassoni, Fabio Brugnara, and Roberto Gretter The Vocapia Research ASR Systems for Evalita Julien Despres, Lori Lamel, Jean-Luc Gauvain, Bianca Vieru, Cécile Woehrling, Viet Bac Le, and Ilya Oparin The FBK ASR System for Evalita Ronny Ronny, Aamir Shakoor, Fabio Brugnara, and Roberto Gretter

11 XIV Table of Contents Forced Alignment on Spontaneous Speech EVALITA 2011: Forced Alignment Task Francesco Cutugno, Antonio Origlia, and Dino Seppi The SPPAS Participation to the Forced-Alignment Task of Evalita Brigitte Bigi SAD-Based Italian Forced Alignment Strategies Giulio Paci, Giacomo Sommavilla, and Piero Cosi UNINA System for the EVALITA 2011 Forced Alignment Task Bogdan Ludusan Author Index

This volume is published and copyrighted by: Roberto Basili Fabio Crestani Marco Pennacchiotti ISSN XXXXX Copyright c 2014 for the individual papers

. This volume is published and copyrighted by: Roberto Basili Fabio Crestani Marco Pennacchiotti ISSN XXXXX Copyright c 2014 for the individual papers by the papers authors. Copying permitted only for