- nexa - server-nexa.polito.it

Paper: "100,000 false positives for every real terrorist: Why anti-terror algorithms don’t work"
by J.C. DE MARTIN Sept. 6, 2017

Sept. 6, 2017

*100,000 false positives for every real terrorist: Why anti-terror algorithms don’t work *by Timme Bisgaard Munk /First Monday/, Volume 22, Number 9 - 4 September 2017 http://journals.uic.edu/ojs/index.php/fm/article/view/7126/6522 doi: http://dx.doi.org/10.5210/fm.v22i19.7126 / //Abstract/ Can terrorist attacks be predicted and prevented using classification algorithms? Can predictive analytics see the hidden patterns and data tracks in the planning of terrorist acts? According to a number of IT firms that now offer programs to predict terrorism using predictive analytics, the answer is yes. According to scientific and application-oriented literature, however, these programs raise a number of practical, statistical and recursive problems. In a literature review and discussion, this paper examines specific problems involved in predicting terrorism. The problems include the opportunity cost of false positives/false negatives, the statistical quality of the prediction and the self-reinforcing, corrupting recursive effects of predictive analytics, since the method lacks an inner meta-model for its own learning- and pattern-dependent adaptation. The conclusion is algorithms don’t work for detecting terrorism and is ineffective, risky and inappropriate, with potentially 100,000 false positives for every real terrorist that the algorithm finds.

2 2

il controllo delle mail in ambito lavorativo sentenza CEDU, Grand Chamber, caso Barbulescu c. Romania del 5 settembre 2017
by Mauro Alovisio Sept. 6, 2017

Sept. 6, 2017

Gent.me/msegnalo sentenza di interesse: sentenza CEDU, Grand Chamber, caso Barbulescu c. Romania del 5 settembre 2017 <http://www.dirittoegiustizia.it/allegati/PP_INTERN_CEDU_milizia_s_4.pdf> e relativo articolo di analisi dell'Avv. Giulia Milizia ad oggetto "*Il datore non può spiare la mail e la chat del dipendente, nemmeno se aziendali: viola la sua privacy* " http://www.dirittoegiustizia.it/news/17/0000085620/Il_datore_non_puo_spiare… Cordiali saluti Mauro Alovisio

1 0

Il nuovo regolamento europeo in materia di protezione dei dati personali - 22 settembre 2017 (Torino)
by Francesco Ruggiero Sept. 5, 2017

Sept. 5, 2017

Gentilissimi, venerdì *22 settembre* *2017,* dalle ore 16:30 alle 20:30, si svolgerà un seminario dal titolo: *Il nuovo regolamento europeo in materia di protezione dei dati personali. Quale impatto per le imprese?* L'incontro, organizzato dalla Fondazione dell’Ordine degli Ingegneri della Provincia di Torino in collaborazione con l’Ordine Ingegneri e il Centro Studi di Informatica Giuridica di Ivrea – Torino, si terrà presso la *Sala Consiglio dell’ Ordine degli Ingegneri di Torino*, in via Giolitti, 1. Parteciperanno *Francesco PIZZETTI*, *Mauro ALOVISIO*, *Maria Grazia D’AMICO, Nicola DI MARE e Fabrizio Mario VINARDI.* Introduce e modera *ing. Alberto Maria BARBERIS*. Maggiori informazioni sul programma sono disponibili alla pagina: http://www.privacyitalia.eu/event/seminario-regolamento-europeo-materia-pro… Cordiali saluti, Francesco -- Francesco Ruggiero Communication Manager Nexa Center for Internet & Society Politecnico di Torino - DAUIN Corso Duca degli Abruzzi, 24 - 10129 Torino web: http://nexa.polito.it mail: francesco.ruggiero(a)polito.it tel: 011 090 7219

1 0

Why this woman strikes fear into the net’s big boys | John Naughton
by Alberto Cammozzo Sept. 4, 2017

Sept. 4, 2017

https://www.theguardian.com/commentisfree/2017/sep/03/woman-strikes-fear-in… ... The McCaskill bill proposes to “clarify” section 230 so that websites that knowingly facilitate sex trafficking can be held liable for what they publish. It’s cleverly drafted to target a particularly odious use of the immunity provided by section 230. But the internet companies are deeply alarmed by it, because they know it could be the thin end of a powerful wedge – a wedge that could, effectively, destroy their business models (and change the internet). But they’re in a bind because – after all – they’re all opposed to sex slavery. Just like they were once all in favour of the first amendment. This is soon going to get really interesting. Stay tuned.

1 0

How to Regulate Artificial Intelligence
by Alberto Cammozzo Sept. 3, 2017

Sept. 3, 2017

How to Regulate Artificial Intelligence By OREN ETZIONI <https://www.nytimes.com/2017/09/01/opinion/artificial-intelligence-regulati…> The technology entrepreneur Elon Musk recently urged the nation’s governors <http://thehill.com/policy/technology/342345-elon-musk-we-need-to-regulate-a…> to regulate artificial intelligence “before it’s too late.” Mr. Musk insists that artificial intelligence represents an “existential threat to humanity,” an alarmist view that confuses A.I. science with science fiction. Nevertheless, even A.I. researchers like me recognize that there are valid concerns about its impact on weapons, jobs and privacy. It’s natural to ask whether we should develop A.I. at all. I believe the answer is yes. But shouldn’t we take steps to at least slow down progress on A.I., in the interest of caution? The problem is that if we do so, then nations like China will overtake us. The A.I. horse has left the barn, and our best bet is to attempt to steer it. A.I. should not be weaponized, and any A.I. must have an impregnable “off switch.” Beyond that, we should regulate the tangible impact of A.I. systems (for example, the safety of autonomous vehicles) rather than trying to define and rein in the amorphous and rapidly developing field of A.I. I propose three rules for artificial intelligence systems that are inspired by, yet develop further, the “three laws of robotics” that the writer Isaac Asimov introduced in 1942: A robot may not injure a human being or, through inaction, allow a human being to come to harm; a robot must obey the orders given it by human beings, except when such orders would conflict with the previous law; and a robot must protect its own existence as long as such protection does not conflict with the previous two laws. These three laws are elegant but ambiguous: What, exactly, constitutes harm when it comes to A.I.? I suggest a more concrete basis for avoiding A.I. harm, based on three rules of my own. []

1 0

programma di e-privacy 2017 autumn edition, Venezia 13 e 14 ottobre
by Mauro Alovisio Sept. 2, 2017

Sept. 2, 2017

Gent.me/mi segnalo da altra lista interessanti informazioni e programma di e-privacy 2017 "E' stato pubblicato il programma di e-privacy 2017 autumn edition, che si svolgera' a Venezia il 13 e 14 ottobre, nella sede storica del Tribunale a Rialto. Il tema e' "I dati perduti - Processo al Cloud". L'argomento Cloud viene affrontato da e-privacy per la seconda volta, a motivo dell'impatto che la "virtualizzazione" dei dati edi vincoli sulla loro disponibilta' ha ed avra' sulla Rete. Considerando l'argomento senz'altro controverso e la sede di quest'anno, nella giornata di sabato le presentazioni saranno organizzate in forma di vero e proprio "Processo", con testimonianze, arringhe, sentenza ed ampia partecipazione del pubblico e delle parti "in causa". Marco Calamari Trovate il programma ed il link per l'iscrizione, come sempre gratuita, sul sito istituzionale http://e-privacy.winstonsmith.org/ Vi aspettiamo. Marco Calamari" Buon fine settimana Mauro Alovisio <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…> Mail priva di virus. www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

1 0

Lavorare gratis, addestrare intelligenze artificiali
by Antonio Casilli Sept. 2, 2017

Sept. 2, 2017

Per un anno, gli utilizzatori dell'applicazione Google Quick Draw (un gioco che consiste a disegnare entro 20 secondi una figura data) hanno "addestrato" gratis il sistema di intelligenza artificiale SketchRNN. Ecco il paper che presenta i risultati, e spiega come il training dataset per questa rete neurale è stato estratto: "We constructed QuickDraw , a dataset of vector drawings obtained from Quick, Draw! [ 11 ], an online game where the players are asked to draw objects belonging to a particular object class in less than 20 seconds. QuickDraw consists of hundreds of classes of common objects. Each class of QuickDraw is a dataset of 70K training samples, in addition to 2.5K validation and 2.5K test samples." Ha, David & Eck, Douglas (2017). "A Neural Representation of Sketch Drawings" https://arxiv.org/pdf/1704.03477.pdf E un articolo un po' meno tecnico qui: https://www.theatlantic.com/technology/archive/2017/06/google-drawing/52947… -- Antonio A. Casilli Associate Professor, Telecom ParisTech Research Fellow, Edgar-Morin Center (EHESS) Member, Interdisciplinary Institute for Innovation (i3 UMR 9217 CNRS) Faculty Fellow, Nexa Center for Internet & Society

1 0

CFP: Recent Ethical Challenges in Social Network Analysis (Paris, 5-6 December 2017)
by Antonio Casilli Sept. 1, 2017

Sept. 1, 2017

Dear all, Co-organizzo un convegno interdisciplinare sull'etica della ricerca sulle reti sociali. Si svolgerà a Parigi, il 5 e 6 dicembre 2017 (in inglese). Per iscriversi o sottomettere una comunicazione, l'indirizzo web è: [ https://recsna17.sciencesconf.org/ | https://recsna17.sciencesconf.org/ ] Cheers, Antonio Casilli Call for papers Recent Ethical Challenges in Social Network Analysis (RECSNA17) Paris, 5-6 December 2017 Interdisciplinary research on social networks is experiencing unprecedented growth, fuelled by the consolidation of the field of social network analysis and the increasing availability of data from digital networking platforms. However, it raises formidable ethical issues that often fall outside existing regulations and guidelines. New tools to collect, treat, store personal data expose both research participants and practitioners to specific risks. Issues surrounding political instrumentalization or economic takeover of scientific results transcend standard research concerns. Legal and social ramifications of studies on personal ties and human networks surface at an unprecedented pace. The aim of this workshop is to bring together researchers in the social sciences, economics and management, statistics, computer science, law and philosophy, as well as other stakeholders , to further advance the ethical reflection in the face of new research challenges. Organization, dates, and venues This two-day workshop is hosted by the MSH (Maison des Sciences de l'Homme) Paris Saclay (*) and organized by the [ https://britsoc.co.uk/groups/study-groups/social-network-analysis-study-gro… | Social Network Analysis Group of British Sociological Association ] (BSA-SNAG), in collaboration with: the European Network on Digital Labor (ENDL) and the Social Networks group of the French Sociological Association (AFS RT 26 – Réseaux Sociaux). (*) Financial support from the Maison des Sciences de l’Homme Paris-Saclay is gratefully acknowledged. The workshop will take place in Paris on Tue. 5 and Wed. 6 December 2017. The venues are: * On 5 December 2017, it is organized as a full-day, small-attendance workshop bringing together presenters, discussants and chairs to discuss contributed papers in depth. It takes place at the Cachan campus of the [ http://ens-paris-saclay.fr/en/campus | ENS (École normale supérieure) Paris Saclay ] . * On 6 December 2017 (morning only), it consists in a keynote session and a roundtable in which highlights from the previous day will be summarized and presented to a larger public. It takes place at the [ http://www.paris-iea.fr/en/presentation-of-the-institute/the-hotel-de-lauzun | Hôtel de Lauzun ] , a splendid 17th century palace situated in the heart of historic Paris, home to the Paris IAS (Institute for Advanced Study). Background The study of social networks long predates the advent of the Internet. Since the 1930s, sociologists have used surveys to detect relationships between individuals and groups, for example advice between employees of a company, or friendship between pupils in a school class. In all these cases, there are ethical issues regarding informed consent (as relationships reported by participants may concern non-participants, possibly unaware of the research) and anonymization (which cannot be achieved at the data collection stage, which must include personal identifiers, but only ex post ). While researchers have already engaged with these issues (notably with a special issue of the leading journal Social Networks , 2005), new challenges arise today with the increased availability of relational data from digital platforms such as Facebook and Twitter. The boundaries between “public” and “private” spheres blur, platforms’ algorithms affect users’ behaviours in ways that are not always transparent to researchers, and information sources are often owned by commercial firms unwilling to share them. Issues surrounding the ownership and accessibility of personal data via proprietary apps and APIs are the center of huge controversies as well as of regulatory efforts such as the new European GDRP (General Data Protection Regulation). Paid “crowdsourcing” (recruiting people through platforms such as Amazon Mechanical Turk to perform tasks that may range from answering questionnaires to downloading their full online contact lists) enables researchers to produce extra data, but raises formidable issues in terms of contributors’ working conditions. In short, the data do not speak for themselves, and we must consider the conditions of their use, production and extraction as well as the software architecture imposed by commercial platforms, scientific publishers, tool developers and institutions. Possible consequences include barriers to data access, inequalities between researchers (with a potential advantage for corporate R&D over publicly-funded research), and a general sense of uncertainty which may hamper otherwise beneficial social studies. Themes The following is a non-exhaustive list of themes: * Social network research and individual/fundamental rights * Ethics and codes of conduct of innovative fields of research in social network analysis * Culture and values in international and comparative social network research * Respect of participants in research on sensitive populations, problematic behaviors, covert networks * Anonymity, privacy, and security in social network analysis * Impact and risks of publishing network data * Psychological impact and other risks for research participants * Dialogue between public and private sector researchers * Complying with regulatory efforts in social network analysis * Dealing with risks and threats for the researcher * Personal and institutional responsibility in research The aim of this workshop is to offer a space for researchers in social network analysis to discuss these issues on the basis of reflective accounts of their experiences, in consultation with other stakeholders in research ethics committees, regulatory bodies, and businesses. We invite scholars and other professionals to submit papers that critically engage with ethics in research related to social networks, preferably on the basis of one or more case studies (which may be problems they encountered in their own research activity), taken as concrete illustrations of the general principles at stake, the attitudes and behaviours of stakeholders, or the legal and institutional constraints. We are particularly interested in novel, original answers to some unprecedented ethical challenges, or the need to re-interpret norms in ambiguous situations. Target audience We welcome contributions from academics (at both senior and early-career levels, also including postgraduate research students) as well as researchers in corporate R&D services, business leaders, representatives of platform companies, policy-makers and members of research ethics committees. We are open to diverse disciplinary backgrounds (such as the social sciences, economics, digital humanities, computer science, statistics, philosophy / ethics), research approaches (qualitative / quantitative) and empirical settings (online / face-to-face social networks). The workshop welcomes the participation of stakeholders, general public, and non-presenting attendees, who can attend both days (5-6 December 2017: see below) or day 2 only (6 December 2017). Deadline for submissions * 15 October 2017: [ https://recsna17.sciencesconf.org/user/submit | submission of a 300-word abstract ] . * 20 October 2017: expression of interest to be panel discussants and chairs (send to: [ mailto:recsna17@msh-paris-saclay.fr | recsna17(a)msh-paris-saclay.fr ] ). * 31 October 2017: acceptance notifications. * 15 November 2017: deadline for speaker registration (paper presenters, panel discussants, chairs) * 30 November 2017: deadline for general attendee registration. You may contact [ mailto:paola.tubaro@inria.fr | paola.tubaro(a)inria.fr ] with questions about scholarly contributions to this conference. Registration * Registration is free but mandatory, and includes: venues, lunch on day 1, and coffee breaks. * However, registration does not cover travel and accommodation costs for participants. * [ https://recsna17.sciencesconf.org/registration/index | Registration form ] (open from 15 October 2017 to 30 November 2017): Keynote speakers [ http://pagines.uab.cat/joseluismolina/content/presentaci%C3%B3n | José Luis Molina ] , Autonomous University of Barcelona, HyperEthics: A Critical Account [ https://www.oii.ox.ac.uk/people/bernie-hogan/ | Bernie Hogan ] , Oxford Internet Institute, Privatising the personal network: Ethical challenges for social network site research Scientific committee Antonio A. Casilli (Telecom ParisTech, FR), Alessio D’Angelo (Middlesex University, UK), Guillaume Favre (University of Toulouse Jean-Jaurès, FR), Bernie Hogan (Oxford Internet Institute, UK), Elise Penalva-Icher (University of Paris Dauphine, FR), Louise Ryan (University of Sheffield, UK), Paola Tubaro (CNRS, FR). Contact us Email: [ mailto:recsna17@msh-paris-saclay.fr | recsna17(a)msh-paris-saclay.fr ] Webpage: [ http://recsna17.sciencesconf.org/ | http://recsna17.sciencesconf.org ] Twitter: [ http://twitter.com/recsna17 | @recsna17 ] _______________________________________________ Diffz-ses mailing list Diffz-ses(a)listes.telecom-paristech.fr [ https://listes.telecom-paristech.fr/mailman/listinfo/diffz-ses | https://listes.telecom-paristech.fr/mailman/listinfo/diffz-ses ]

1 0

AI Programs Are Learning to Exclude Some African-American Voices
by Antonio Vetro' Aug. 31, 2017

Aug. 31, 2017

Questo breve articolo fa 2-3 piccoli ma chiari esempi di bias di alcuni algoritmi usati nel riconoscimento vocale e di sentimenti : https://www.technologyreview.com/s/608619/ai-programs-are-learning-to-exclu… Un caro saluto, Antonio

1 0

Corso di Perfezionamento in Criminalità Informatica e Investigazioni Digitali
by Silvia Martinelli Aug. 31, 2017

Aug. 31, 2017

Cari tutti, solo per segnalare il Corso di Perfezionamento dell'Università degli Studi di Milano in Criminalità Informatica e Investigazioni Digitali, Prof. Giovanni Ziccardi, che si terrà a Milano a partire dal mese di ottobre. Il corso si rivolge ai possessori di laurea e laurea magistrale e intende fornire ai partecipanti nozioni al fine di gestire le emergenze connesse alla valutazione, acquisizione e gestione del rischio, della prova digitale e dell’informazione elettronica in generale sia a fini processuali (penali, civili, giuslavoristici, amministrativi) sia per lo svolgimento di investigazioni interne e/o aziendali nonché alla gestione del rischio, della prova digitale in contesti transnazionali e alla comprensione dell’aspetto sostanziale e processuale dei principali e più comuni reati informatici. Il percorso formativo è strutturato in 56 ore di didattica frontale: - stalking e bullismo; - crimini informatici a sfondo sessuale; - odio domestico; - ambito finanziario e aziendale; - diritto di famiglia; - normativa 231, privacy, misure di sicurezza e trattamento illecito dei dati; - reputazione dell’individuo online e sua tutela/violazione. Le lezioni si terranno un giorno a settimana, di giovedì (12 e 26 ottobre, 9,16,23 e 30 novembre, 14 dicembre. Dalle 9 alle 18) Contributo di iscrizione: € 706,00 Per ulteriori informazioni: infomaster.giurisprudenza(a)unimi.it o giovanni.ziccardi(a)unimi.it Iscrizioni aperte da oggi! http://www.unimi.it/studenti/corsiperf/111497.htm Un caro saluto, S.

1 0