DSAA 2018 in Turin, Italy, is the 5th annual installment of the conference. This year brings the collaboration of the American Statistical Association, to complement the IEEE Computational Intelligence Society and the ACM SIGKDD. This year DSAA also adds the support of the ISI Foundation, which has worked for 35 years to break down traditional silos in the sciences of complexity and data.

Important Dates

http://users.dsic.upv.es/~prosso/

Keynote Speech
Stance and Misogyny: Analysing Cases of Hate Speech

Abstract. Once upon a time we believed in the wisdom of crowds and that we could learn from what others discussed in social media. Nowadays, we are not so sure about it. Comments in social media often lack of aurgumentation and users do not try to persuade others to agree with their claim by presenting evidence. Communication is mostly with users belonging to the same community with the risk of an intellectual isolation (filter bubble), where beliefs may be reinforced by a repeated message inside the closed community (echo chamber).  When users communicate with someone who disagrees with their viewpoints, they often do it spewing hateful comments behind a veil of anonymity. This may only increase political and social polarization and
extremism. In this keynote, we will show some cases of hate speech in Twitter we came across in the datasets of two shared tasks we organized at IberEval on stance detection and misogyny identification. We will also comment how some participants detect them.

Short bio. Farah Benamara (https://www.irit.fr/~Farah.Benamara/) is Associate Professor in computer science at Toulouse University since 2005. She has been designated to be a member of The French University Council (CNU) during 4 years (2015-2019) and has been awarded by the French Higher Education and Research Ministry with a research excellence fellowship for the 4 years (2015-2019). She works on the development of semantic and pragmatic models for language understanding, focusing in particular on five main research areas: question answering, sentiment analysis, figurative language processing, discourse processing and Arabic processing. She has published more than 70 papers in peer-reviewed international conference and journal papers. She has supervised 4 PhD theses and 4 projects, one of them an industrial project on opinion analysis which resulted in a technology transfer. Farah Benamara is a member of various program committees of major international conferences and journals (ACL, EMNLP, COLING, *SEM) and serves on the editorial board of the Journal of Dialogue and discourse. She has co-organised several national and international conferences and workshops, such as KRAQ workshops associated to IJCAI 2005 and EACL 2016 on advanced reasoning in question answering, the national workshop ColTal@TALN 2016 on NLP tools for social media, the French evaluation initiative DEFT@TALN 2017 on sentiment analysis and figurative language on tweets, and finally the international workshop SentIsData@DSAA 2016 on Emotion and Sentiment in Intelligent Systems and Big Social Data Analysis. She has been co-editor of a special issue of the Journal of Computational Linguistics on Language in Social Media.

Cristina Bosco, University of Turin, Italy

Short bio. Cristina Bosco is Assistant Professor in Computer Science, PhD in Computer Science, with a background in philosophy, computer science, and linguistics. Her research focus is on computational linguistics and precisely on the development of resources for morphological and syntactic analysis of language, machine translation, and sentiment analysis, with a special interest for hate speech, irony and sarcasm detection. She published more than 90 papers in peer-reviewed international conferences and journals. She is principal investigator on projects funded for detecting hate speech in social media, and organized several scientific events: special session about Emotion and Sentiment in Intelligent Systems and Big Social Data Analysis at the IEEE-DSAA in 2015 and 2016, Workshop Emotion and Sentiment in Social and Expressive Media in 2013, 2015 and 2017, Second Conference for Italian Computational Linguistics in 2015, Workshop Evalita in 2014 and several evaluation tasks in Evalita from 2007. Founding member of the Italian Association for Computational Linguistics (AILC), she is in program committees of several conferences (LREC, COLING, TLT) and reviewer of journals.

Elisabetta Fersini, University of Milano-Bicocca, Italy

Short bio. Elisabetta Fersini, PhD in Computer Science, is currently Aassistant Professor at the University of Milano-Bicocca (Department of Informatics, Systems and Communications). Her research is mainly focused on propositional and relational machine learning, with a particular interest in probabilistic discriminative and generative models. Her activity finds application in several natural language processing tasks, among which text classification and clustering, information extraction and sentiment analysis both for well-formed text and social media contents. She is the principal investigator in projects funded for data analytics and the organizer of the special session about Emotion and Sentiment in Intelligent Systems and Big Social Data Analysis at the IEEE-DSAA in 2015 and 2016. She is a member of the editorial board of the journal titled Information Processing and Management and editor of the book “Sentiment Analysis in Social Networks”.

Gabriella Pasi, University of Milano-Bicocca, Italy

Short bio. Gabriella Pasi is Full Professor at the Department of Informatics, Systems, and Communication (DISCo) of the University of Milano Bicocca, Milano, Italy. Within DISCo she leads the Information Retrieval Lab. Her main research activities are related to Information Retrieval and Information Filtering. In recent years, she has addressed the issues of contextual search and user modeling. She is also conducting research activities related to the analysis of user-generated content on social media. She has published more than 200 papers in International Journals and Books, and on the Proceedings of International Conferences. She is involved in several activities for the evaluation of research; in particular, she was appointed as an expert of the Computer Science panel for the Starting Grants (till 2011), and Consolidator Grants (2012) of the Programme Ideas at the European Research Council. From 2013 to 2017 she has been the President of the European Society for Fuzzy Logic and Technologies (EUSFLAT). She is a member of the Editorial Board of several international journals, and she has delivered several keynote talks/plenary lectures at international conferences related to her research interests. She has participated in the organization of several International events, in both roles of organization and program chair.

Viviana Patti, University of Turin, Italy

Short bio. Viviana Patti received her PhD in Computer Science from the University of Torino, Italy, where she currently works as Associate Professor. Her recent research interests include sentiment analysis, affective computing, irony detection, semantic web ontologies. She is recently applying her research in the field of hate speech monitoring. In 2013 she co-founded the series of international workshops on Emotion and Sentiment in Social and Expressive media (ESSEM: 2013@AI*IA, 2015@AAMAS and 2017@ACII). She has been guest editor of the special issue on Emotion and Sentiment in Social and Expressive Media in the Information Processing & Management Journal (January, 2016) and of the special section on Affect and Interaction in Agent-based Systems and Social Media in ACM Transactions on Internet Technology (January 2017). She has (co-)authored over almost 90 peer-reviewed-publications and serves on multiple conference committees in top conferences (IJCAI 2017, ACII 2017, AAAI18, LREC18). Co-chair of the 2016 and 2018 editions of the EVALITA evaluation campaign for NLP tools for Italian. She contributed to the development of Twitter corpora for sentiment analysis in Italian, French and Spanish, English and co-organized the 1st Italian shared task on Sentiment Analysis in Twitter (SENTIPOLC), 2014 and 2016 editions. She co-organized the shared task `Stance and Gender Detection in Tweets on Catalan Independence’ at Ibereval 2017 and the shared task on ‘Multilingual Emoji Prediction’ at Semeval 2018. She is also co-chair of several special tracks at international conferences, e.g., SentIsData@DSAA (16-17), AI4BigData@FLAIRS (16-18), Emotions@AISB18 and of the workshop series PEOPLES (@COLING2016, @NAACL2018). She has been area co-chair for the track on ‘Pragmatics and creativity’ at the 4th Italian Conference on Computational Linguistics (CLiC 2017). 

Marco Viviani, University of Milano-Bicocca, Italy

Short bio. Marco Viviani is Senior Research Fellow at the Department of Informatics, Systems, and Communication (DISCo) of the University of Milano-Bicocca (Milan, Italy). He graduated in Computer Science and he obtained the Ph.D. in Informatics from the University of Milan (Italy). In 2009 he obtained a one-year postdoctoral Research Fellow at the University of Burgundy / Le2i (Dijon, France), in 2010 a two-year postdoctoral Research Fellow at the Institut National des Sciences Appliquées (INSA) / Liris (Lyon, France), and from 2011 to 2013 he held a postdoctoral Research Fellow at the University of Insubria / DiSTA (Varese, Italy). He is currently working on the issue of assessing information credibility in social media.  His topics of interest include knowledge representation and management, social media analytics, trust and reputation modeling, information retrieval. On these topics, he has published several papers in International Journals, Books, and Conferences.

• Subjectivity, sentiment, and emotion detection in social media and big data
• Sentiment-based indexing, search, and retrieval in social media
• Sentiment topic detection & trend discovery
• Multimodal emotion and sentiment detection in social media
• Big data for multimodal affective interaction (e.g., chatbots)
• Time evolving opinion & sentiment analysis
• Emotions, sentiment, geographic locations and places
• Using sentiment and affect for social media predictive analysis (including genre identification, political preference, etc.)
• Irony and sarcasm detection
• Emotion models and ontologies of emotions
• Affect in natural language
• Hate speech detection
• Information fusion for affective computing
• Summarization and visualization of emotions, sentiment and affective data
• Applications of sentiment analysis and emotion detection in social media to education, entertainment, health, e-government, games
• Ethical issues in affect and opinion detection in user-generated contents
• Opinion spam, group spam, fake news detection
• Credibility/reliability of health-related information
• Fact-checking
• Multimedia content credibility
• Information/misinformation diffusion
• Trust and reputation in virtual communities
• Retrieval of credible information
• Gold standard datasets generation with respect to the credibility of information
• Crowdsourcing credibility

On the one hand, the focus of the Special Session will be given to the study and the application of affective computing and sentiment analysis to social data, which can impact on monitoring, analyzing and counteracting discrimination and hate speech, which are increasingly spreading phenomena in our countries also in combination with the pervasiveness of social media. Furthermore, also the applications of sentiment analysis and emotion detection in social media for the development of education, entertainment, health, e-government, and games will be considered as an interesting object of investigation.

On the other hand, by considering the process of “disintermediation” that affects social media, the Special Session will also investigate the problem of assessing the credibility of information spreading among and across virtual communities. The diffusion of fake news, hoaxes, rumors, fake reviews, inaccurate health information, can have a negative impact on society with respect to different aspects, from influencing political elections, producing harmful effects if connected to the health of patients, to generating hate and discrimination phenomena. For all these reasons, the study and the development of approaches that can help people in automatically assess the level of credibility of information is a fundamental research issue in the last years.

The aim of this Special Session is, therefore, to cover different aspects related to Data Analysis applied to social data, by addressing to a heterogeneous community of researchers who has data science as a common denominator.

Please visit the SeCredISData page on the DSAA’18 Website for the detailed program of the session.