Avvisi dell'associazione

Manifesto per la rinascita di una nazione

posted Apr 12, 2018, 3:17 AM by Amministratore Sito Web

Lavoro e Intelligenza Artificiale

posted Mar 27, 2018, 3:45 AM by Amministratore Sito Web

In memoria di Stephen William Hawking

posted Mar 14, 2018, 9:33 AM by Amministratore Sito Web


Il problema e’ ….. la nostra intelligenza (?)

posted Feb 17, 2018, 6:40 AM by Amministratore Sito Web

Il valore dell’Intelligenza Artificiale

posted Jan 29, 2018, 4:10 AM by Amministratore Sito Web

Qualche giorno fa è apparsa sulla rete la notizia, ripresa da molte agenzie di stampa, che, grazie all’Intelligenza Artificiale è possibile, in modo molto semplice, sostituire il volto di una persona con quella di un’altra in un video.

L’articolo citava l’esempio di video porno, ma è chiaro che la tecnologia è applicabile anche ad altri contesti, più o meno decenti o appropriati ed in generale apre una nuova prospettiva nella generazione di notizie false (le temute fake news). Si pensi semplicemente alla possibilita’ di accompagnare ad un testo falso il video di un personaggio famoso che dice cose “sconvenientemente false” in modo estremamente realistico.  La barriera tra vero e falso avrebbe subito un ulteriore duro colpo.

Non stiamo parlando di un evento nuovo. La maldicenza è in uso nella società da moltissimo tempo e costituisce un fenomeno di lapidazione sociale fra i più pericolosi (in alcuni casi sfocia in episodi di violenza estrema o suicidio).

In passato però questi fenomeni, seppur deleteri in qualsiasi epoca, non avevano a disposizione strumenti così potenti come i social network, i video registrati (veri o falsi che siano) e persino le tecniche di intelligenza artificiale.

Gli strumenti tecnologici di cui ci stiamo dotando hanno grandi potenzialità e come tutti gli strumenti potenti vanno usati con cautela. Oggi queste applicazioni sono nelle mani di tutti (così come in molti paesi lo sono le armi automatiche) ed è facile che vengano usati da malintenzionati, persone con scarso livello etico, delinquenti oppure immaturi.

Inoltre viviamo in un contesto storico in cui il disagio sociale sta aumentando, insieme a livelli di disoccupazione, guerre, violenza più o meno gratuita, sentimenti di odio verso minoranze sociali, e così via.  Abbiamo purtroppo già visto in passato dove queste situazioni possano portarci.

Come Associazione Italiana per l’Intelligenza Artificiale sentiamo da tempo il dovere di approfondire gli studi sui possibili effetti indesiderati derivanti dalle applicazioni della scienza di cui ci occupiamo.

Riteniamo importante che la ricerca e gli utilizzi dell’Intelligenza Artificiale vengano indirizzati verso il benessere dell’uomo; denunciamo e respingiamo con vigore gli impieghi deleteri sia in campo economico (riduzione dei posti di lavoro per la sola massimizzazione dei profitti), militare (le armi autonome ad esempio), di controllo sociale, così come l’uso improprio da parte di soggetti devianti.

Riteniamo che l’Intelligenza Artificiale possa costituire un enorme potenziale per l’umanità se indirizzata verso l’educazione, la sanità, la crescita economica del benessere, la salvaguardia dell’ambiente, la democrazia, il miglioramento delle condizioni di lavoro e la prevenzione del crimine.

Condanniamo in modo deciso qualsiasi strumentalizzazione e utilizzo perverso degli strumenti che contribuiamo a realizzare e intendiamo profondere tutto il nostro impegno per indirizzarli verso il benessere dell’umanità e del pianeta.

II post sulla pagina AIxIA:

Italian AI ecosystem, by area

posted Nov 20, 2017, 11:16 AM by Riguzzi Fabrizio   [ updated Mar 21, 2018, 12:21 PM ]





Per essere inclusi nel landscape, compilare questo form https://goo.gl/forms/HqO7GatSSN5lsmiC3 

Italian AI ecosystem, by region

posted Nov 20, 2017, 11:14 AM by Riguzzi Fabrizio   [ updated Mar 21, 2018, 12:22 PM ]





Per essere inclusi nel landscape, compilare questo form https://goo.gl/forms/HqO7GatSSN5lsmiC3

EurAI Bulletin #2, 2017

posted Oct 4, 2017, 2:28 AM by Riguzzi Fabrizio

EurAI Bulletin #2, 2017

Highlights from the General Assembly, Melbourne

Here are some issues and topics discussed during the GA in Melbourne:

  • EurAI is financially in good shape. The final 2016 budget and 2018 budget proposal were approved.

  • A proposal for an ACAI 2018 on "Statistical Relational AI", to be held on August 27-31, 2018 in Ferrara was presented and approved.

  • IJCAI/ECAI 2018: Jeff Rosenschein (Gemeral Chair) and Fredrik Heintz (Local Chair) presented the current status of the first joint IJCAI/ECAI to be held next year in Stockholm. The event will be co-located with AAMAS and ICML, with joint sessions and workshops.

  • EurAI is planning to increase its engagement in European programmes.

  • During an open discussion on the future of ECAI, it was felt that ECAI should remain as is, that is, happen every two years, but that we may want to consider making ACAI annual, perhaps co-located with a national AI conference of one of our member societies.

ACAI Summer School 2017

7-14 October 2017, Nieuwpoort, Belgium

We are happy to announce that this year’s ACAI summer school will be solely devoted to Reinforcement Learning (basics and recent developments), a branch of machine learning that focuses on learning from interaction with an uncertain environment and potential peers. The school will be organized from 7th until 14th October 2017 in Nieuwpoort, Belgium.

Recent progress and successes, such as e.g. AlphaGo, have revived strong interest in Reinforcement Learning (RL) and have pushed the boundaries of the state-of- the-art. Unfortunately, in typical machine learning courses and text books only the very basics of RL are covered. To bridge the gap between basic RL algorithms, such as for instance Q-learning, and the RL architecture that allowed to beat some of the best human Go players, this summerschool will introduce various advanced RL techniques, such as: batch and policy search techniques, shaping and inverse RL, RL for multi-agent systems and robotics, and Deep Reinforcement Learning. These topics will be taught by top scientists and experts in RL.

Interested in learning more about RL? but unsure about your basic RL knowledge? No worries, we got you covered: you can opt for the additional two days before the actual start of the school to make sure you have the necessary background and skills before diving into the more advanced topics.

We have chosen for an on-site approach, where all participants and lectures stay at the same location and spend a full week together to ‘learn' RL and jointly participate in social activities. If you want to know more about the school and the registration process: check out the website at https://ai.vub.ac.be/ssrl2017

EurAI fellows 2017

The EurAI board is delighted to announce the list of the 2017 EurAI fellows:

- Vicent Botti (Spain)

- Gal Kaminka (Israel)

- Alessio Lomuscio (UK)

- Paola Mello (Italy)

- Ioannis Vlahavas (Greece)

The committee was comprised by Giuseppe de Giacomo, Helen Fargier, Ilka Niemela, Luc de Raedt, and Hector Geffner (as Chair).

Dissertation Award 2016

We are happy to report that the Dissertation Award Committee selected the winners for 2016 with presentations given at a special session at IJCAI.

The Best-of-n Problem in Robot Swarms Gabriele Valentini

Nominated by: Marco Dorigo

with an Honourable Mention to:

Probabilistic reasoning and Learning for the Semantic Web Riccardo Zese

Nominated by: Fabrizio Riguzzi, Evelina Lamma

EurAI conference sponsorship programme

We recall that EuRAI sponsors invited talks at international conferences. The guidelines for eligibility and duties of sponsored conferences can be read at https://www.eurai.org/apply_for/conference_sponsorship.

Conference organizers who want to apply to funding for a conference that takes place before the end of July 2018 should apply before November 10, 2017.

The submission Web site for EurAI Sponsorship spring 2018 is https://easychair.org/conferences/?conf=euraisponsorship2018.

There is no need to upload a pdf: you may simply provide the required information in the abstract.

There will be another call later in spring of next year for conferences that take place in the second half of 2018.

The contact person on the EurAI board is Catholijn Jonker.

AI-related conferences

3rd Global Conference on Artificial Intelligence, GCAI 2017

Miami, USA 18-22 October 2017

http://easychair.org/smart-program/GCAI2017/

9th International Joint Conference on Computational Intelligence IJCCI 2017

Funchal, Madeira, Portugal | November 1-3, 2017

http://www.ijcci.org/

32nd AAAI Conference on Artificial Intelligence, AAAI-18

New Orleans, US | February 2-7, 2017

AAAI/ACM Conference on AI, Ethics, and Society

New Orleans, US | February 2-3, 2017

9th International Joint Conference on Automated Reasoning, IJCAR 2018

Oxford, UK| July 14-17, 2018

http://www.ijcar2018.org

Federated Logic Conference, FLoC 2018

Oxford, UK| July 6-19, 2018

http://www.floc2018.org

16th International Conference on Principles of Knowledge Representation and Reasoning, KR 2018

Tampa US | October 30-November 2, 2018

http://kr2018.org/

Magazines

AI Magazine Vol 38 n. 2 https://www.aaai.org/ojs/index.php/aimagazine/issue/view/218

AI Communications - Volume 30, issue 5

http://content.iospress.com/journals/ai-communications/30/5

Contact Information

This bullettin is edited by Michela Milano, DISI Università di Bologna michela.milano@unibo.it.

For advertising AI-related conferences, schools and journal issues, please contact the editor.


Post-Doctoral contracts within H2020 Project SPECIAL

posted Sep 30, 2017, 4:25 AM by Riguzzi Fabrizio

Post-Doctoral contracts within H2020 Project SPECIAL
Scalable Policy-aware Linked Data Architecture For Privacy, Transparency and Compliance

Expected duration: 2 years. 
Location: University of Naples Federico II (Italy). 
Under the supervision of P.A. Bonatti (2). 

SPECIAL is developing an infrastructure, based on Semantic and Big Data technologies, aimed at supporting compliance with the upcoming European General Data Protection Regulation (GDPR).

The ideal candidates are expected to carry out research on one of the following topics:
  1. Semantic policy encoding and reasoning about policies (e.g. compliance, policy comparison, policy factorization, policy explanation). 
  2. Fair exchange protocols for transparency logs.
  3. Privacy enhancing mechanism design
Activities (1) and (2) involve both theoretical work and experimental research. Therefore, candidates are expected to have a solid background in computer science.  Competence in knowledge representation and/or security and privacy are particularly appreciated. Activity (3) requires a background in game theory.

For more information and expressions of interest please contact Prof. Bonatti


    Funded under the H2020-ICT-2016-1 Big Data PPP call 
    (Privacy-preserving Big Data technologies, ICT-18-2016)

EurAI Bullettin #1, 2017

posted Mar 10, 2017, 12:45 AM by Riguzzi Fabrizio

ACAI Summer School 2017

7-14 October 2017, Nieuwpoort, Belgium

We’re happy to announce that this year’s ACAI summer school will be solely

devoted to Reinforcement Learning (basics and recent developments), a branch

of machine learning that focuses on learning from interaction with an uncertain

environment and potential peers. The school will be organized from 7th until

14th October 2017 in Nieuwpoort, Belgium.

Recent progress and successes, such as e.g. AlphaGo, have revived strong interest

in Reinforcement Learning (RL) and have pushed the boundaries of the state-ofthe-

art. Unfortunately, in typical machine learning courses and text books only

the very basics of RL are covered. To bridge the gap between basic RL

algorithms, such as for instance Q-learning, and the RL architecture that allowed

to beat some of the best human Go players, this summerschool will introduce

various advanced RL techniques, such as: batch and policy search

techniques, shaping and inverse RL, RL for multi-agent systems and robotics,

and Deep Reinforcement Learning. These topics will be taught by top scientists

and experts in RL.

Interested in learning more about RL? but unsure about your basic RL

knowledge? No worries, we got you covered: you can opt for the additional two

days before the actual start of the school to make sure you have the necessary

background and skills before diving into the more advanced topics.

We have chosen for an on-site approach, where all participants and lectures stay

at the same location and spend a full week together to ‘learn' RL and jointly

participate in social activities. If you want to know more about the school and the

registration process: check out the website at https://ai.vub.ac.be/ssrl2017

EurAI fellows 2017 – Call for Nominations

Deadline for nominations: Wednesday 15 March 2017

INTRODUCTION

Nominations are sought for EurAI Fellows 2017. The EurAI Fellows program was

initiated in 1999 to recognize individuals who have made significant, sustained

contributions to the field of artificial intelligence (AI) in Europe. The EurAI

Fellows Program honors only a very small percentage of the total membership of

all EurAI member societies (up to a maximum of 3%). Each Fellow will receive a

Certificate signed by the EurAI Chair. The names of the recently elected Fellows

are published each year on the EurAI Webpages and in the EurAI Bulletin. At

each ECAI conference a Fellows lunch will be

sponsored by EurAI to honor the recently elected Fellows.

WHO IS ELIGIBLE TO BE AN EurAI FELLOW?

To be eligible for consideration, candidates **must** be a current member of an

EurAI member organisation (see http://www.eurai.org/member/ for a list of

member societies). Usually only individuals who have made contributions to AI

for a decade or more after receiving their PhD (or are at an equivalent career

stage) will be selected.

WHO CAN MAKE A NOMINATION?

There are two ways that an individual can be nominated:

• A nomination can be made by an EurAI member society. See

http://www.eurai.org/member/ for a list of member societies. Please,

note that for each member society the number of nominations is limited

to a maximum of **three**.

• A nomination can be made by two EurAI fellows, who must be from

different countries.

WHAT IS THE NOMINATION PROCESS?

A nomination must be a **single** PDF document, including the following:

• a statement confirming which EurAI member society the candidate is

currently a member of;

• a comprehensive current CV of the candidate;

• three separate letters of reference, each of which should clearly make the

case why the individual is considered appropriate for EurAI fellowship.

The nomination document should be submitted via EasyChair:

https://easychair.org/conferences/?conf=fellows2017

(the system will ask for an abstract - this can be omitted). Please

include the name of the nominee in the title of the submission.

Nominations must be submitted no later than

***Wednesday March 15, 2017***.

We hope to complete the evaluation process by May 2017. The nominators will

be informed of the decision for any nomination(s) they made.

HOW WILL NOMINATIONS BE EVALUATED?

All nominations will be evaluated by the EurAI fellows committee. The key

criterion of evaluation will be whether the nominee has achieved exceptional

distinction in AI in Europe. Distinction may be evidenced in several ways --

fellows' accomplishments range from pioneering advances in the theory of AI, to

unusual accomplishments in AI technology and applications. Leadership in

EurAI or EurAI member societies, support of forums for the exchange of ideas,

and exceptional service for the international AI community may will also be

considered in the selection process.

Most fellowship nominations are for technical contributions, and for most of

these, the case is usually based on the nominee's scientific achievements as

documented in their publication record. In evaluating such a publication record,

the fellowship committee will look for clear evidence of scientific excellence,

which may, for example, come in the form of a citation record that indicates a

profound influence on the field, a publication record of exceptional quality, or

technical results of unusual depth, insight, and impact. Note that

other forms of technical contribution evidence will also be considered, such as

patent awards or statements of longstanding contribution to an industrial AI

group effort that has been influential in the field generally; again, clear evidence

should be provided of such influence.

The fellowship committee will in all cases ultimately make decisions based on

their professional judgment of the overall quality of the case.

EurAI: call for conference sponsorship

EuRAI sponsors invited talks at international conferences. The guidelines for

eligibility and duties of sponsored conferences can be read at

https://www.eurai.org/apply_for/conference_sponsorship

Conference organizers who want to apply to funding for a conference that takes

place before the end of December 2017 should apply **before April 12, 2017**.

The submission Web site for EurAI Sponsorship 2017 is:

https://easychair.org/conferences/?conf=euraisponsorship17

There is no need to upload a pdf: you may simply provide the required

information in the abstract.

There will be another call later in the autumn of this year for conferences that

take place in the first half of 2018.

EurAI Artificial Intelligence Dissertation Award 2016

Nominations are invited for the 2016 Artificial Intelligence Dissertation Award

sponsored by EurAI, the European Association of Artificial Intelligence. Please

take a moment to nominate your students, or to recommend to your colleagues

to nominate their students.

This Award includes a certificate signed by the EurAI President and 1500 Euros

(which includes the travel grant for the Award ceremony). Nominations are due

by May 1, 2017 by submitting a single PDF file as described below listing both

the NOMINATED candidate and the NOMINATOR as the AUTHORS through

EasyChair at: https://easychair.org/conferences/?conf=aidissertationaward2

Eligible doctoral dissertations are those defended after December 1, 2015 in

the general area of Artificial Intelligence. The dissertation must have been

defended at a European university and the author must be a personal member of

a EurAI member society. In case a thesis is nominated for another national or

international award before or during its review for the EurAI dissertation award,

the author is requested to inform the EurAI dissertation award committee about

this fact.

To be considered, a dissertation must be nominated by the thesis supervisor,

who must submit the following items as a SINGLE PDF file in the order:

• Nomination cover-page that includes: the name, affiliation, and contact

details of the nominator; the date of the defense of the dissertation; and

the names of the jury/examiners involved in its defense;

• nomination letters, either digitally signed or scanned original, from two

referees supporting the submission and stating their assessment of why

the thesis should win the award.

• if the thesis was not written in English the nomination must include an

English paper describing the core ideas of the thesis that has been

submitted for publication in an international journal or a prestigious

conference.

• the full nominated dissertation.

The selection will be based on the originality, impact, and written quality of the

work. Work that has been submitted to and/or accepted at workshops,

conferences, or journals will be considered more favourably. Work that is

primarily attributed to the student's own initiative will also be considered more

favourably. Finally, the quality of the written document will be considered.

Please address your queries to Barry O'Sullivan - b.osullivan[at]cs.ucc.ie

The deadline for receipt of submissions is May 1, 2017.

The outcome will be announced by the first week of July.

The Award will be presented during IJCAI 2017.

AI-related conferences

16th International conference on Autonomous Agents and Multi Agent

Systems, AAMAS 2017

San Paulo –Brasil May 8-12, 2017

http://www.aamas2017.org/

International Joint Conference on Artificial Intelligence IJCAI 2017

Melbourne, Australia | August 19-25, 2017

http://ijcai-17.org/

23rd International Conference on Principles and Practice of Constraint

Programming, CP 2017

Melbourne, Australia | August 28 to September 1, 2017

http://cp2017.a4cp.org/

33rd International Conference on Logic Programming ICLP 2017

Melbourne, Australia | August 28 to September 1, 2017

http://iclp17.a4lp.org/

20th International Conference on Theory and Applications of Satisfiability

Testing SAT 2017

Melbourne, Australia | August 28 to September 1, 2017

http://sat2017.gitlab.io/

27th International Conference on Automated Planning and Scheduling,

ICAPS 2017

Pittsburgh, USA - June 18 - 23, 2017

http://icaps17.icaps-conference.org/

Special Issues

Journal of Automated Reasoning

Special Issue On Automated Reasoning Systems

https://www.mpi-inf.mpg.de/JARS17/

Magazines

AI Magazine Vol 37 n. 4

http://www.aaai.org/ojs/index.php/aimagazine/issue/view/216

AI Magazine Vol 37 n. 3

http://www.aaai.org/ojs/index.php/aimagazine/issue/view/215

AI Communications Volume

Volume preprints

http://content.iospress.com/journals/ai-communications/Preprint/Preprint

Contact Information

This bullettin is edited by Michela Milano, DISI Università di Bologna

michela.milano@unibo.it.

For advertising AI-related conferences, schools and journal issues, please contact

the editor.

1-10 of 67