We have released the video above which demonstrates the practical applications of our research in emergency care situations. In this case study, nurses or student nurses wearing a Smart-Assistant (in this example Smart-Eye-Ware) are attempting to resuscitate a patient (doll). Besides offering information on demand in their HMD (e.g. instant feedback, regulations or quick-check information, hints), the Smart-Assistant also detects specific activities like performing chest compressions and provides feedback if the activity is not performed to required standards.

This research expands on work presented in the award winning papers: Smart-Watch Life Saver: Smart-Watch Interactive-Feedback System for Improving Bystander CPR and Recognizing Hospital Care Activities with a Pocket Worn Smartphone (award details here).

Abstract: Healthcare Information Systems typically fall into the group of systems in which the need of data sharing conflicts with the privacy. A myriad of these systems have to, however, constantly communicate among each other. One of the ways to address the dilemma between data sharing and privacy is to use data obfuscation by lowering data accuracy to guarantee patient’s privacy while retaining its usefulness. Even though many obfuscation methods are able to handle numerical values, the obfuscation of non-numerical values (e.g., textual information) is not as trivial, yet extremely important to preserve data utility along the process. In this paper, we preliminary investigate how to exploit ontologies to create obfuscation mechanism for releasing personal and electronic health records (PHR and EHR) to selected audiences with different degrees of obfuscation. Data minimisation and access control should be supported to enforce different actors, e.g., doctors, nurses and managers, will get access to no more information than needed for their tasks. Besides that, ontology-based obfuscation can also be used for the particular case of data anonymisation. In such case, the obfuscation has to comply with a specific criteria to provide anonymity, so that the data set could be safely released. This research contributes to: state the problems in the area; review related privacy and data protection legal requirements; discuss ontology-based obfuscation and anonymisation methods; and define relevant healthcare use cases. As a result, we present the early concept of our Ontology-based Data Sharing Service (O-DSS) that enforces patient’s privacy by means of obfuscation and anonymisation functions.

Citation: Iwaya, Leonardo H. and Giunchiglia, Fausto and Martucci, Leonardo A. and Hume, Alethia and Fischer-H{\”u}bner, Simone and Chenu-Abente, Ronald, “Ontology-Based Obfuscation and Anonymisation for Privacy”, In “Privacy and Identity Management. Time for a Revolution? 10th IFIP WG 9.2, 9.5, 9.6/11.7, 11.4, 11.6/SIG 9.2.2 International Summer School, Edinburgh, UK, August 16-21, 2015, Revised Selected Papers”, 2016, Springer International Publishing, Cham, pages 343–358, isbn 978-3-319-41763-9, doi 10.1007/978-3-319-41763-9_23, http://dx.doi.org/10.1007/978-3-319-41763-9_23. New York, USA, July 2016.

Download: http://bit.ly/2iTQvzT

Abstract: The flâneur is the urban vagabond in search of experiences and inspirations from serendipitously exploring a city environment. This construct is put beside post-modern stances about the suburban areas built and populated after the Second World War industrialization, along with considerations about ecological psychology, cultural materialism, and sound theory. The main concept is to provide those places with a communication level that would be pleasant to discover while wandering without a destination. Therefore it is desirable to conceive a meta-design tool able to incorporate creativity, ownership, and conviviality.

Citation: Torsi, Silvia. “City of Beats: Analysing Flânerie as a Practice for Living the Physical Space.” Cultural Influences on Architecture. IGI Global, 2017. 157-180. Web. 11 Jan. 2017. doi:10.4018/978-1-5225-1744-3.ch006

Download: http://bit.ly/2jIi9nO

Abstract: In computing science, much attention has been paid to generic methods for sharing data in secure infrastructures. These sorts of methods and infrastructures are, of course, necessary for sharing healthcare data. The authors are, however, a long way away from being able to realise the potential of medical and healthcare data to support the sorts of extensive, data-intensive experiments being demanded by precision and stratified medicine. A key architectural problem remaining to be solved is how to maintain control of patient data within the governance of local data jurisdictions, while also allowing these jurisdictions to engage with experiment designs that (because of the need to scale to large population sizes) may require analyses across several jurisdictions. This study provides a snapshot of architectural work underway to provide a clear, effective structure of data safe havens within jurisdictions. It then describes how formally specified experiment designs can be used to enable jurisdictions to work together on experiments that no single jurisdiction could tackle alone. The authors’ current work relates to two jurisdictions (in Scotland and in Italy), but the architecture and methods are general across similar jurisdictions.

Citation: David Robertson, Fausto Giunchiglia, Stephen Pavis, Ettore Turra, Gabor Bella, Elizabeth Elliot, Andrew Morris, Malcolm Atkinson, Gordon McAllister, Areti Manataki, Petros Papapanagiotou, and Mark Parsons (2016). Healthcare data safe havens: towards a logical architecture and experiment automation. The Journal of Engineering, Institution of Engineering and Technology, October, 2016. This is an open access article published by the IET under the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0/), http://digital-library.theiet.org/content/journals/10.1049/joe.2016.0170.

Download: http://bit.ly/2j3khFT

Abstract: Public institutions, such as universities, maintain data in several information silos, each of them engineered to serve a specific vertical application. Data about key entities—such as people, publications, courses, projects—is scattered across them and difficult to correlate due to the diversity in format, metadata, conventions, and terminology used. In such a scenario, nowadays it is practically impossible to correlate data and support advanced search and analytics facilities, in turn vital to identify institutional priorities and support institutional strategic goals, as well as to offer effective data visualization and navigation services to their users (e.g., researchers, students, alumni, companies).

Citation: Vincenzo Maltese and Fausto Giunchiglia. 2016. Search and Analytics Challenges in Digital Libraries and Archives. J. Data and Information Quality 7, 3, Article 10 (August 2016), 3 pages. DOI: http://dx.doi.org/10.1145/2939377

Download: http://bit.ly/2j3vczf

Abstract: As the European Union (EU) funded SmartSociety project aims to create a toolset for rapidly and systematically engineering collective intelligence systems to support daily living, it simultaneously wants to ameliorate the risks to individuals of participating in these types of hyper-connected digital systems. This paper reports on a panel session at the close at of the 2015 IFIP summer school that reflected upon a keynote speech covering SmartSociety concepts, technologies and ethical dilemmas. The panel session was conceived as a consultative exercise as part of the ongoing Responsible Research and Innovation (RRI) approach embedded within the SmartSociety project. In this chapter we present an analysis of the panel session discussion, which touched on several key issues, including the relationships between technology and society, what we should expect from a ‘SmartSociety’, barriers and horizons in managing ethical issues, and brokerage as a methodological approach to weaving multiple perspectives into design.

Citation: Hartswood, Mark, and Marina Jirotka. “SmartSociety: Collaboration Between Humans and Machines, Promises and Perils.” In Privacy and Identity Management. Time for a Revolution?, Aspinall, D., Camenisch, J., Hansen, M., Fischer-Hübner, S.,Raab, C. (Eds.) pp. 30-48. Springer.

Download: http://bit.ly/2k9xR89

Imaginary has released a series of videos on our work here at SmartSociety. The videos offer an everyday-human perspective of the challenges faced by a society entering an era marked by an increased digital presence. As our 3 years FP7 EU project nears its end, we are reminded of the diverse and impactful work delivered by all SmartSociety members at the University of Trento, the University of Edinburgh, U-hopper s.r.l., the German Research Centre for Artificial Intelligence, the University of Oxford, Ben-Gurion University of the Negev, Imaginary s.r.l., the University of Karlstad, the Vienna University of Technology, and the University of Southampton.

Download: http://bit.ly/2jsmRp9

Download: http://bit.ly/2jCx7ZQ

Download: http://bit.ly/2iuIHJw

Download: http://bit.ly/2jCCuZ5

Download: http://bit.ly/2jsdztp

Abstract: Plan recognition algorithms need to maintain all candidate hypotheses which are consistent with the observations, even though there is only a single hypothesis that is the correct one. Unfortunately, the number of possible hypotheses can be exponentially large in practice. This paper addresses the problem of how to disambiguate between many possible hypotheses that are all consistent with the actions of the observed agent. One way to reduce the number of hypotheses is to consult a domain expert or the acting agent directly about its intentions. This process can be performed sequentially, updating the set of hypotheses during the recognition process. The paper specifically addresses the problem of how to minimize the number of queries made that are required to find the correct hypothesis. It adapts a number of probing techniques for choosing which plan to query, such as maximal information gain and maximum likelihood. These approaches were evaluated on a domain from the literature using a well known plan recognition algorithm. The results showed that the information gain approach was able to find the correct plan using significantly fewer queries than the maximum likelihood approach as well as a baseline approach choosing random plans. Our technique can inform the design of future plan recognition systems that interleave the recognition process with intelligent interventions of their users.

Citation: Reuth Mirsky, Ya’akov Gal, Roni Stern, Meir Kalech. Sequential Plan Recognition. International Joint Conference on Artificial Intelligence (IJCAI), New York, USA, July 2016.

Download: http://bit.ly/2iR3wLf

Abstract: Plan Recognition algorithms require to recognize a complete hierarchy explaining the agent’s actions and goals. While the output of such algorithms is informative to the recognizer, the cost of its calculation is high in run-time, space, and completeness. Moreover, performing plan recognition on-line requires the observing agent to reason about future actions that have not yet been seen and maintain a set of hypotheses to support all possible options. This paper presents a new and efficient algorithm for online plan recognition called SLIM (Semi-Lazy Inference Mechanism). It combines both a bottom-up and top-down parsing processes, which allow it to commit only to the minimum necessary actions in real-time, but still provide complete hypotheses post factum. We show both theoretically and empirically that although the computational cost of this process is still exponential, there is a significant improvement in run-time when compared to a state of the art of plan recognition algorithm.

Citation: Reuth Mirsky, Ya’akov Gal. SLIM: Semi-Lazy Inference Mechanism for Plan Recognition. International Joint Conference on Artificial Intelligence (IJCAI), New York, USA, July 2016.

Download: http://bit.ly/2jh7HTZ

Abstract: Volunteer-based crowdsourcing depend critically on maintaining the engagement of participants. We explore a methodology for extending engagement in citizen science by combining machine learning with intervention design. We first present a platform for using real-time predictions about forthcoming disengagement to guide interventions. Then we discuss a set of experiments with delivering different messages to users based on the proximity to the predicted time of disengagement. The messages address motivational factors that were found in prior studies to influence users’ engagements. We evaluate this approach on Galaxy Zoo, one of the largest citizen science application on the web, where we traced the behavior and contributions of thousands of users who received intervention messages over a period of a few months. We found sensitivity of the amount of user contributions to both the timing and nature of the message. Specifically, we found that a message emphasizing the helpfulness of individual users significantly increased users’ contributions when delivered according to predicted times of disengagement, but not when delivered at random times. The influence of the message on users’ contributions was more pronounced as additional user data was collected and made available to the classifier.

Citation: Avi Segal, Ya’akov Gal, Ece Kamar, Eric Horvitz, Alex Bower, Grant Miller. Intervention Strategies for Increasing Engagement in Volunteer-Based Crowdsourcing. International Joint Conference on Artificial Intelligence (IJCAI), New York, USA, July 2016.

Download: http://bit.ly/2jh9fgG

Abstract: What is a fair way to assign rooms to several housemates, and divide the rent between them? This is not just a theoretical question: many people have used the Spliddit website to obtain envy-free solutions to rent division instances. But envy freeness, in and of itself, is insufficient to guarantee outcomes that people view as intuitive and acceptable. We therefore focus on solutions that optimize a criterion of social justice, subject to the envy freeness constraint, in order to pinpoint the “fairest” solutions. We develop a general algorithmic framework that enables the computation of such solutions in polynomial time. We then study the relations between natural optimization objectives, and identify the maximin solution, which maximizes the minimum utility subject to envy freeness, as the most attractive. We demonstrate, in theory and using experiments on real data from Spliddit, that the maximin solution gives rise to significant gains in terms of our optimization objectives. Finally, a user study with Spliddit users as subjects demonstrates that people find the maximin solution to be significantly fairer than arbitrary envy-free solutions; this user study is unprecedented in that it asks people about their real-world rent division instances. Based on these results, the maximin solution has been deployed on Spliddit since April 2015.

Citation: Ya’akov Gal, Moshe Mash, Ariel D. Procaccia, Yair Zick. Which Is the Fairest (Rent Division) of Them All? ACM Conference on Economics and Computation (EC), July, Maasticht, The Netherlands. Best Paper Award.

Download: http://bit.ly/2iQYwGz

Abstract: Researchers that study Human-Like computing mainly aim to understand how systems can emulate human cognitive performance. However, when Human-Like systems are designed for sharing economy applications in which humans have to collaborate in order to achieve a desired task, there are several problems that needs to be addressed before asking how cognitive performance of a single human can be emulated. In this paper, we highlight these problems, provide examples related to the ridesharing scenario, and introduce how we approach these problems and which techniques we are using to tackle them.

Citation: S.Ceppi. The Dilemma of Human-Like Collective Systems. In S. Muggleton et al, editor, Twentieth Workshop on Machine Intelligence (MI 20), Windsor Park, UK, 23-25 October, 2016. In press.

Download: http://bit.ly/2iivPpQ

Abstract: Vehicle-to-grid (V2G) is a promising approach whereby electric vehicles (EVs) are used to store excess electricity supply (e.g., from renewable sources), which is sold back to the grid in times of scarcity. In this paper we consider the setting of a smart car park, where EVs come and go, and can be used for V2G while parked. We develop novel allocation and payment mechanisms which truthfully elicit the EV owners’ preferences and constraints, including arrival, departure, required charge, as well as the costs of discharging due to loss of efficiency of the battery. The car park will schedule the charging and discharging of each EV, ensuring the constraints of the EVs are met, and taking into consideration predictions about future electricity prices. Optimally solving the global problem is intractable, and we present three novel heuristic online scheduling algorithms. We show that, under certain conditions, two of these satisfy monotonicity and are therefore truthful. We furthermore evaluate the algorithms using simulations, and we show that some of our algorithms benefit significantly from V2G, achieving positive benefit for the car park even when agents do not pay for using it.

Citation: Enrico H. Gerding, Sebastian Stein, Sofia Ceppi and Valentin Robu. Online Mechanism Design for Vehicle-to-Grid Car Parks. In Proceedings of the 25th International Joint Conference on Artificial Intelligence (IJCAI). New York, USA, July 2016.

Download: http://bit.ly/2jpbxYL

Abstract: While AI has recently produced impressive systems that achieve human-like performance at challenging tasks, these systems tell us very little about how human intelligence works. In particular, they do not address the problem of composing knowledge and behaviour incrementally – a phenomenon that is pervasive in individual and collective human intelligence. We argue that achieving more human-like AI requires focusing on diversity in reasoning and behaviour among humans and artificial agents, and that developing systems capable of dealing with such diversity is key to achieving more human-like AI. In these systems intelligence should not only be measured in terms of how a system performs at a certain task, but also in terms of the properties of the process by which each component combines its knowledge and behaviour with that of others, just like humans do.

Citation: M. Rovatsos. Diversity-Awareness – The Key to Human-Like Computing? In S. Muggleton et al, editor, Twentieth Workshop on Machine Intelligence (MI 20), Windsor Park, UK, 23-25 October, 2016. In press.

Download: http://bit.ly/2jd0oMI

Abstract: Sharing economy applications need to coordinate humans, each of whom may have different preferences over the provided service. Traditional approaches model this as a resource allocation problem and solve it by identifying matches between users and resources. These require knowledge of user preferences and, crucially, assume that they act deterministically or, equivalently, that each of them is expected to accept the proposed match. This assumption is unrealistic for applications like ridesharing and house sharing (like airbnb), where user coordination requires handling of the diversity and uncertainty in human behaviour.
We address this shortcoming by proposing a diversity-aware recommender system that leaves the decision-power to users but still assists them in coordinating their activities. We achieve this through taxation, which indirectly modifies users’ preferences over options by imposing a penalty on them. This is applied on options that, if selected, are expected to lead to less favourable outcomes, from the perspective of the collective. The framework we used to identify the options to recommend is composed by three optimisation steps, each of which has a mixed integer linear program at its core. Using a combination of these three programs, we are also able to compute solutions that permit a good trade-off between satisfying the global goals of the collective and the individual users’ interests. We demonstrate the effectiveness of our approach with two experiments in a simulated ridesharing scenario, showing: (a) significantly better coordination results with the approach we propose, than with a set of recommendations in which taxation is not applied and each solution maximises the goal of the collective, (b) that we can propose a recommendation set to users instead of imposing them a single allocation at no loss to the collective, and (c) that our system allows for an adaptive trade-off between conflicting criteria.

Citation: P. Andreadis, S. Ceppi, M. Rovatsos, and S. Ramamoorthy. Diversity-Aware Recommendation for Human Collectives. In Proceedings of the 1st International Workshop on Diversity-Aware Artificial Intelligence (DIVERSITY 2016), The Hague, The Netherlands, 2016

Download: http://bit.ly/2jp8rUr

Abstract: Human participation in socio-technical systems is overgrowing conventional crowdsourcing where humans solve simple, independent tasks. Novel systems are attempting to leverage humans for more intellectually challenging tasks, involving longer lasting worker engagement and complex collaboration patterns. Controllability of such systems requires different direct and indirect methods of influencing the participating humans. Conventional human organizations, such as companies or institutions, have been using incentives for decades to align the interests of workers and organizations. With the collaborations managed by the socio-technical platforms growing ever more complex and resembling, or even surpassing in complexity, the conventional ones, there is a need to apply advanced incentivizing techniques in the virtual environment as well. However, existing incentive management techniques in use in crowdsourcing/sociotechnical platforms are not suitable for the described (complex or intellectually-challenging) tasks. In addition, existing platforms currently use custom-developed solutions. This approach is not portable, and effectively prevents reuse of common incentive logic and reputation transfer. Consequently, this prevents workers from comparing different platforms, hindering the competitiveness of the virtual labor market and making it less attractive to skilled workers.
This research presents a complete set of models and tools for programmable incentive management for social computing platforms. In particular, it introduces:
(i) A comprehensive, multidisciplinary review of existing literature on incentives as well as an extensive survey of real-world incentive practices in social computing milieu,
(ii) A low-level model of incentives suitable for use in socio-technical systems
(iii) princ – a model and framework for execution of programmable incentive mechanisms, allowing the offering of incentives through a service model.
(iv) pringl – a high-level domain-specific language for encoding complex incentive strategies for socio-technical systems, encouraging a modular approach in building
incentive strategies, cutting down development and adjustment time and creating a basis for development of standardized but tweakable incentives.
The tools are meant to allow system and incentive designers a complete environment for modeling, administering/executing and adjusting a whole spectrum of realistic incentive mechanisms in a privacy-preserving manner. No known comparable systems were known to exist at the time of writing of the thesis.

Citation: PhD Thesis: Ognjen Scekic: Automated Incentive Management for Social Computing – Foundations, Models, Tools and Algorithms, TU Wien, March 2016.

Download: http://bit.ly/2jJKIBR

Abstract: Recent developments of computing systems allow humans to participate not only as service consumers but also as service providers. The interweaving of human-based computing into machine-based computing systems becomes apparent in smart city settings, where human-based services together with software-based services and thing-based services (e.g., sensor-as-a-service) are orchestrated for solving complex problems, leading to the creation of the so-called Cyber-PhySical-Social Systems (CPSSs). Monitoring such CPSSs is essential for system planning, management, and governance. However, due to the diversity of the involved building blocks, it is challenging to monitor such systems. In this paper, we present metric models and the associated Quality of Data (QoD) to elastically monitor the execution metrics of a centralized coordinated CPSS. We develop a monitoring framework for capturing and analyzing runtime metrics occurring on various facets of the coordinated CPSS. Furthermore, we present the implementation of our monitoring framework, and showcase monitoring features in a simulated system using real world infrastructure maintenance scenarios.

Citation: Z.C. Muhammad Candra, Hong-Linh Truong, Schahram Dustdar: On Monitoring Cyber-Physical-Social Systems. SERVICES 2016: 56-63.

Download: http://bit.ly/2iMVeUP

However, to reap these benefits, new thinking is required around the regulation and governance of smart online platforms to ensure balanced interests, and to promote fair and safe forms of participation, particularly relating to the role played by algorithms and data in driving economies of scale.

In order to address the issues this raises, we are organising a policy event on smart online platforms on the 5th of December 2016, at the Scotland Europa Conference Centre in Brussels.

Please visit our page on the event for further details and registration.

09:15-09:30 Welcome
09:30-10:00 Towards Building Ontologies with the Wisdom of the Crowd. Paula Chocron, Dagmar Gromann and Francisco José Quesada Real
10:00-10:30 A Methodology to Take Account of Diversity in Collective Adaptive Systems. Heather S. Packer and Luc Moreau
10:30-11:00 Coffee break
11:00-11:30 Diversity-Aware Recommendation for Human Collectives. Pavlos Andreadis, Sofia Ceppi, Michael Rovatsos and Subramanian Ramamoorthy
11:30-12:00 Industry talk: Democracy by Design. Marcel van Hest
12:00-13:00 Invited talk by Antonella de Angeli
13:00-14:00 Lunch
14:00-14:20 A Semantic Distance based Architecture for a Guesser Agent in ESSENCE’s Location Taboo Challenge. Kemo Adrian, Aysenur Bilgin and Paul Van Eecke
14:20-14:40 Interdisciplinarity as an Indicator of Diversity in a Corpus of Artificial Intelligence Research Articles. Bilge Say
14:40-15:00 Managing human diversity in diverse multi-agent collaborative intelligence systems. Mark Hartswood, Kevin Page, Avi Segal, Kobi Gal and Marina Jirotka
15:00-15:20 Analysing communicative diversity via the Stag Hunt. Robert van Rooij and Katrin Schulz
15:20-15:40 Domain-Based Sense Disambiguation in Multilingual Structured Data. Gabor Bella, Alessio Zamboni and Fausto Giunchiglia
15:40-16:10 Coffee break
16:15-17:15 Panel discussion
17:15-17:30 Wrap-up

Objective

Collective intelligence platforms, where large numbers of people contribute their skills, resources, and knowledge in a collaborative way, have emerged in recent years as a new type of socio-technical systems that use the creativity of their users to solve complex problems. However, digital infrastructures that support such collective intelligence are only successful and sustainable in the long term if they manage to adapt to the evolution of the ways in which they are used, and co-evolve meaningfully with their user base. The overall objective of this atelier is to build a prototypical Web-based collective intelligence system that has such adaptive capabilities. Participants will be trained in the fundamental theoretical and technological building blocks of such systems using the SmartSociety software platform, which contains many of the key components to engineering collective adaptive systems successfully.

Format

The atelier will involve daily lectures on each of the key technologies used in the SmartSociety platform, and the rest of the day group work implementation of an actual case study. The theoretical introductions will focus only very briefly on general material, and then go into aspects specific to collective adaptive systems for the most part.

Prerequisites

Participants are expected to have general mathematical literacy (background in mathematical notation, discrete mathematics, probability theory), and good programming skills. Experience with web programming (REST, Javascript, etc) is strongly recommended, as this will be extensively used in the practical part of the atelier.

Programme

Monday 5th September
15:30-17:00 Lecture: Introduction to Collective Adaptive Systems
Michael Rovatsos, University of Edinburgh

Tuesday 6th September
11:00-13:00 Lecture: Overview of SmartSociety Platform
Tommaso Schiavinotto, U-Hopper Srl
14:30-17:00 Lab: Experimentation with Orchestration Manager

Wednesday 7th September
11:00-13:00 Lecture: Introduction to semantic privacy technologies
Ronald Chenu-Abente, University of Trento
14:30-17:00 Lab: Experimentation with Peer Manager

Thursday 8th September
11:00-13:00 Lecture: Introduction to incentive mechanisms
Michael Rovatsos, University of Edinburgh
14:30-17:00 Lab: Experimentation with Incentives Manager

Friday 9th September
11:00-13:00 Lecture: Value-sensitive design
Mark Hartswood, University of Oxford
14:30-17:00 Lab: Completion of implementation
17:30-19:30 Final presentation

The workshop seeks to explore diversity as a phenomenon that both poses a challenge for AI in terms of dealing with and managing diversity in an intelligent system (or ecosystem of intelligent human and/or artificial agents) and presents an opportunity in terms of leveraging diversity (for example through processes like crowdsourcing and collaborative knowledge production) to achieve human-like (and human-friendly) capabilities in more open-ended, incrementally evolving, and interactive AI systems.

The workshop seeks to explore diversity as a phenomenon that both poses a challenge for AI in terms of dealing with and managing diversity in an intelligent system (or ecosystem of intelligent human and/or artificial agents) and presents an opportunity in terms of leveraging diversity (for example through processes like crowdsourcing and collaborative knowledge production) to achieve human-like (and human-friendly) capabilities in more open-ended, incrementally evolving, and interactive AI systems. For more details, please see the original announcement, here.

*** Daily Agenda is now out ***

Organisers

Michael Rovatsos, The University of Edinburgh, mrovatso@inf.ed.ac.uk
Ronald Chenu-Abente, University of Trento, chenu@disi.unitn.it

Background

Diversity is pervasive in human nature and culture, and is deeply rooted in the variation of natural traits and experience among individuals, the collectives they form, and the environments they inhabit. When humans reason individually, they maintain different representations, conceptualisations, and theories, and apply different rules of inference, learning, and decision making. When they interact with each other to combine their skills or resources, to coordinate their activities, and to resolve conflicts between their individual objectives, they exchange information and knowledge, negotiate and align their individual views, and adapt to each other’s behaviour dynamically. Arguably, diversity is not only a phenomenon that humans have to deal with, but it is also the vehicle for achieving some of the most impressive products of human intelligence.

Artificial Intelligence, on the other hand, has so far largely relied on a certain degree of homogeneity, not necessarily in terms of the components involved in a method or system, but in terms of the process that combines them. While various areas within AI have already developed methods that can cope with and/or exploit diversity to some extent, for example

• electronic markets where individual agents have different goals and aim to maximise their own profit,
• hybrid robot architectures that involve different layers of representation and reasoning,
• knowledge sharing infrastructures where different agents use different domain ontologies, and
• machine learning systems that combine different sources of data and/or learning units,

more often than not, these systems still involve a “monolithic”, global approach to integration. This usually derives from a global task context, a common intermediate representation layer, or a global output to be produced by the integrated system.

We believe that there is a huge potential in bringing the insights from work on problems that involve diversity – like those listed in the examples above – together to gain a deeper understanding of the phenomenon of diversity, as well as to develop principled methodological approaches that will enable us to better utilise diversity in future AI systems.

Workshop Description

The workshop seeks to explore diversity as a phenomenon that both poses a challenge for AI in terms of dealing with and managing diversity in an intelligent system (or ecosystem of intelligent human and/or artificial agents) and presents an opportunity in terms of leveraging diversity (for example through processes like crowdsourcing and collaborative knowledge production) to achieve human-like (and human-friendly) capabilities in more open-ended, incrementally evolving, and interactive AI systems.

We aim to bring together researchers from different communities that have each addressed diversity in different ways, such as

• hierarchical and hybrid inference systems (combining representation and reasoning mechanisms),
• semantic web and ontologies (interoperability of information sources, ontology alignment),
• non-monotonic and defeasible reasoning (reasoning about conflicting and changing information),
• mechanism design and social choice (reaching agreement in the presence of conflict of interest),
• language evolution and emergent semantics (evolving shared symbol and concept spaces),
• cross-lingual approaches to natural language understanding (integrating different natural languages),
• teamwork and collaborative multiagent systems (integrating heterogeneous knowledge/behaviours),
• human-AI/human-robot collaboration (aligning agents’ views and objectives with those of humans),
• crowdsourcing and human computation (managing diverse contributions of large human collectives).

The workshop will provide an open forum for researchers from these (and other) areas to contribute their insights on diversity in order to develop a shared agenda for the future study of diversity in AI. We welcome submissions on all aspects of diversity, ranging from theoretical foundations to practical applications, case studies, and surveys. The workshop will be heavily discussion-based, with relatively short paper presentations and a focus on formulating key research questions and a longer-term research agenda for the area. To enable high-quality discussion and debate, a key evaluation criterion will be the focus of papers contributed to the workshop on the diversity “angle“ of the research reported. Specifically, papers should clearly identify

• what type of diversity or aspects of diversity the reported research investigates or accommodates,
• the methods the paper proposes to deal with and/or exploit diversity,
• how the proposed method combines and/or exceeds existing diversity-oriented capabilities, and
• what key challenges in terms of diversity it leaves open for future research.

Beyond this key requirement, we deliberately impose no restrictions on methodological approach, or maturity of the research. In particular, the workshop aims to be inclusive with regard to the types of diversity considered, including (but not limited to) diversity of representations, algorithms, systems infrastructures, datasets, agent behaviours, skills and capabilities, preferences and objectives, but also users, user populations, cultures, contexts of use, application domains, user interfaces, etc.

Also, in keeping with the Special Topic of ECAI 2016 Artificial Intelligence for Human Values, we particularly invite papers that address the ethics and social impact of AI applications related to diversity, for example addressing issues related to the social dynamics of diversity in systems comprising of humans and artificial agents, the emergence of “digital divides“ and the implications of diversity on the cohesiveness of these systems, diversity-aware accountability and privacy methods, or the potential risks and benefits of diversity-aware AI in terms of promoting human diversity in various domains.

Paper submission

We invite full (8-12 pages) and short (4-6 pages) papers for presentation at the workshop, to be submitted through the workshop’s Easychair web site using the ECAI format (which can be downloaded together with instructions from this page). Each paper will be peer-reviewed by at least two Programme Committee members, and authors will be expected to produce final versions of their papers in good time before the workshop.

All accepted papers will be made available online prior to the workshop, and distributed to all participants in hardcopy. If a sufficiently high number of high-quality papers is received, we will aim to produce a special issue in a high-quality journal where revised versions of the papers will be published alongside invited papers.

Important dates

The following is a (tentative) timeline of key dates:

• Paper submission deadline – 14th June 2016 extended to 17th June 2016
• Author notification – 28th June 2016 30th June 2016
• Camera-ready versions – 15th July 2016
• Workshop – 29th or 30th August 2016

Agenda

The following is the workshop agenda for the 29th of August:

09:15-09:30 Welcome
09:30-10:00 Towards Building Ontologies with the Wisdom of the Crowd. Paula Chocron, Dagmar Gromann and Francisco José Quesada Real
10:00-10:30 A Methodology to Take Account of Diversity in Collective Adaptive Systems. Heather S. Packer and Luc Moreau
10:30-11:00 Coffee break
11:00-11:30 Diversity-Aware Recommendation for Human Collectives. Pavlos Andreadis, Sofia Ceppi, Michael Rovatsos and Subramanian Ramamoorthy
11:30-12:00 Industry talk: Democracy by Design. Marcel van Hest
12:00-13:00 Invited talk by Antonella de Angeli
13:00-14:00 Lunch
14:00-14:20 A Semantic Distance based Architecture for a Guesser Agent in ESSENCE’s Location Taboo Challenge. Kemo Adrian, Aysenur Bilgin and Paul Van Eecke
14:20-14:40 Interdisciplinarity as an Indicator of Diversity in a Corpus of Artificial Intelligence Research Articles. Bilge Say
14:40-15:00 Managing human diversity in diverse multi-agent collaborative intelligence systems. Mark Hartswood, Kevin Page, Avi Segal, Kobi Gal and Marina Jirotka
15:00-15:20 Analysing communicative diversity via the Stag Hunt. Robert van Rooij and Katrin Schulz
15:20-15:40 Domain-Based Sense Disambiguation in Multilingual Structured Data. Gabor Bella, Alessio Zamboni and Fausto Giunchiglia
15:40-16:10 Coffee break
16:15-17:15 Panel discussion
17:15-17:30 Wrap-up

Financial Support

The workshop is sponsored by the ESSENCE (www.essence-network.com) and SmartSociety (www.smart-society-project.eu) projects, which will provide extensive financial support to participants, in particular PhD students and junior researchers who wish to participate. To be eligible for such support, interested individuals should submit a short or full paper, and email Michael Rovatsos (mrovatso@inf.ed.ac.uk) with a one-page case for support, providing a short bio, describing their interest in the workshop, and specifying the requested amount together with a justification of the anticipated expenses.

Committees

Workshop Organisers

Michael Rovatsos, The University of Edinburgh, mrovatso@inf.ed.ac.uk
Ronald Chenu-Abente, University of Trento, chenu@disi.unitn.it

Steering Committee

Alan Bundy, University of Edinburgh, United Kingdom
Peter Gardenfors, University of Lund, Sweden
Fausto Giunchiglia, University of Trento, Italy
Asuncion Gomez Perez, Universidad Politecnica de Madrid, Spain
Ben Kuipers, University of Michigan, USA
Ariel Procaccia, Carnegie-Mellon University, USA
Carles Sierra, IIIA-CSIC Barcelona, Spain
Luc Steels, Vrije Universiteit Brussels, Belgium
Michael Wooldridge, University of Oxford, United Kingdom
Gerhard Weiss, University of Maastricht, The Netherlands

Programme Committee

Yoram Bachrach, Microsoft Research Cambridge, United Kingdom
Gabor Bella, University of Trento, Italy
Sofia Ceppi, University of Edinburgh, United Kingdom
Jerome Euzenat, INRIA Grenoble, France
Kobi Gal, Ben-Gurion University of the Negev, Israel
Fabien Gandon, INRIA Sophia-Antipolis, France
Mark Hartswood, University of Oxford, United Kingdom
Nick Hawes, University of Birmingham, United Kingdom
Catholijn Jonker, Technical University of Delft, The Netherlands
Ian Kash, Microsoft Research Cambridge, United Kingdom
Oliver Lemon, Heriot-Watt University Edinburgh, United Kingdom
Nicolas Maudet, Universite Pierre et Marie Curie Paris, France
Fiona McNeill, Heriot-Watt University Edinburgh, United Kingdom
Roberto Navigli, University of Rome “La Sapienza”, Italy
Luc Moreau, University of Southampton, United Kingdom
Iyad Rahwan, MIT, USA
Subramanian Ramamoorthy, University of Edinburgh, United Kingdom
Katharina Reinecke, University of Washington, USA
Robert van Rooij, ILLC University of Amsterdam, The Netherlands
Carlos Ruiz, TAIGER S.A., Spain
Marco Schorlemmer, IIIA-CSIC Barcelona, Spain
Onn Shehory, IBM Haifa Labs, Israel
Pavel Shvaiko, Informatica Trentina, Italy
Remi van Trijp, Sony Computer Science Labs Paris, France