Search

In public media as well as in scientific publications, the term \emph{bias} is used in conjunction with machine learning in many different contexts, and with many different meanings. This paper proposes a taxonomy of these different meanings, terminology, and definitions by surveying the, primarily scientific, literature on machine learning. In some cases, we suggest extensions and modifications to promote a clear terminology and completeness. The survey is followed by an analysis and discussion on how different types of biases are connected and depend on each other. We conclude that there is a complex relation between bias occurring in the machine learning pipeline that leads to a model, and the eventual bias of the model (which is typically related to social discrimination). The former bias may or may not influence the latter, in a sometimes bad, and sometime good way.

International audience

International audience

The rapid dynamics of COVID-19 calls for quick and effective tracking of virus transmission chains and early detection of outbreaks, especially in the phase 2 of the pandemic, when lockdown and other restriction measures are progressively withdrawn, in order to avoid or minimize contagion resurgence. For this purpose, contact-tracing apps are being proposed for large scale adoption by many countries. A centralized approach, where data sensed by the app are all sent to a nation-wide server, raises concerns about citizens' privacy and needlessly strong digital surveillance, thus alerting us to the need to minimize personal data collection and avoiding location tracking. We advocate the conceptual advantage of a decentralized approach, where both contact and location data are collected exclusively in individual citizens' "personal data stores", to be shared separately and selectively, voluntarily, only when the citizen has tested positive for COVID-19, and with a privacy preserving level of granularity. This approach better protects the personal sphere of citizens and affords multiple benefits: it allows for detailed information gathering for infected people in a privacy-preserving fashion; and, in turn this enables both contact tracing, and, the early detection of outbreak hotspots on more finely-granulated geographic scale. Our recommendation is two-fold. First to extend existing decentralized architectures with a light touch, in order to manage the collection of location data locally on the device, and allow the user to share spatio-temporal aggregates - if and when they want, for specific aims - with health authorities, for instance. Second, we favour a longer-term pursuit of realizing a Personal Data Store vision, giving users the opportunity to contribute to collective good in the measure they want, enhancing self-awareness, and cultivating collective efforts for rebuilding society., Revised text. Additional authors

Despite recent progress in plan-based narrative generation, one major limitation is that systems tend to produce a single plotline whose progression entirely determines the narrative experience. However, for certain narrative genres such as serial dramas and soaps, multiple interleaved subplots are expected by the audience, as this tends to be the norm in real-world, human-authored narratives. Current narrative generation techniques have overlooked this important requirement, something which could improve the perceived quality of generated stories. To this end, we have developed a flexible plan-based approach to multiplot narrative generation, that successfully generates narratives conforming to different subplot profiles, in terms of the number of subplots interleaved and the relative time spent on each presentation. We have identified specific challenges such as: distribution of virtual characters across subplots; length of each subplot presentation; and transitioning between subplots.\ud \ud In this paper, we overview this approach and describe its operation in a prototype Interactive Storytelling (IS) System set in the serial drama genre. Results of experiments with the system demonstrate its usability. Furthermore, results of a user study highlight the potential of the approach, with clear user preference for presentations that feature interleaved multiple subplots.

We introduce the formalism of infinite-state reactive modules to reason about the strategic behaviour of autonomous agents in a setting where data are explicitly exhibited in the systems description and in the specification language. Technically, we endow reactive modules with an infinite domain of interpretation for individual variables, and introduce FO-ATL, a first-order version of alternating time temporal logic, for the specification of properties of interest. We show that their verification is decidable for classes of data types of interest. This result is proved by defining a first-order version of alternating bisimulations and finite bisimilar abstractions. We illustrate the formal machinery by applying it to English and sealed bid auctions. In particular, we show that strategic properties of agents in auctions, including manipulability and collusion, can be expressed and verified in this framework.

In the past, research on learning language models mainly used syntactic information during the learning process but in recent years, researchers began to also use semantic information. This paper presents such an approach where the input of our learning algorithm is a dataset of pairs made up of sentences and the contexts in which they are produced. The system we present is based on inductive logic programming techniques that aim to learn a mapping between n-grams and a semantic representation of their associated meaning. Experiments have shown that we can learn such a mapping that made it possible later to generate relevant descriptions of images or learn the meaning of words without any linguistic resource. ispartof: pages:1764-1765 ispartof: Proceedings of the 22nd European Conference on Artificial Intelligence vol:285 pages:1764-1765 ispartof: European Conference on Artificial Intelligence location:The Hague, The Netherlands date:30 Aug - 3 Sep 2016 status: published

We study the query evaluation problem in probabilistic databases in the presence of probabilistic existential rules. Our focus is on the Datalog± family of languages for which we define the probabilistic counterpart using a flexible and compact encoding of probabilities. This formalism can be viewed as a generalization of probabilistic databases, as it allows to generate new facts from the given ones, using so-called tuple-generating dependencies, or existential rules. We study the computational cost of this additional expressiveness under two different semantics. First, we use a conventional approach and assume that the probabilistic knowledge base is consistent and employ the standard possible world semantics. Thereafter, we introduce a probabilistic inconsistency-tolerant semantics, which we call inconsistency-tolerant possible world semantics. For both of these cases, we provide a thorough complexity analysis relative to different languages, drawing a complete picture of the complexity of probabilistic query answering in this family.

We investigate the parameterised model checking problem for specifications expressed in alternating-time temporal logic. We introduce parameterised concurrent game structures representing infinitely many games with different number of agents. We introduce a parametric variant of ATL to express properties of the system irrespectively of the number of agents present in the system. While the parameterised model checking problem is undecidable, we define a special class of systems on which we develop a sound and complete counter abstraction technique. We illustrate the methodology here devised on the prioritised version of the train-gate-controller.

Using Punch and Judy as a story domain, we describe an interactive puppet show, where the flow and content of the story can be influenced by the actions of the audience. As the puppet show is acted out, the audience reacts to events by cheering or booing the characters. This affects the agents’ emotional state, potentially causing them to change their actions, altering the course of the narrative. An institutional normative model is used to constrain the narrative so that it remains consistent with the Punch and Judy canon. Through this vignette of a socio-technical system (STS), comprising human and software actors, an institutional model – derived from narrative theory – and (simplistic) technological interaction artifacts, we begin to be able to explore some of the issues that can arise in STS through the prism of the World-Institution-Technology (WIT) model.

Minimum sum-of-squares clustering (MSSC) is a widely studied task and numerous approximate as well as a number of exact algorithms have been developed for it. Recently the interest of inte- grating prior knowledge in data mining has been shown, and much attention has gone into incorporating user constraints into clustering algorithms in a generic way. Exact methods for MSSC using integer linear programming or constraint programming have been shown to be able to incorporate a wide range of constraints. However, a better performing method for unconstrained exact clustering is the Repetitive Branch-and-Bound Algorithm (RBBA) algorithm. In this paper we show that both ap- proaches can be combined. The key idea is to replace the internal branch-and-bound of RBBA by a constraint programming solver, and use it to compute tight lower and upper bounds. To achieve this, we integrate the computed bounds into the solver using a novel con- straint. Our method combines the best of both worlds, and is generic as well as performing better than other exact constrained meth- ods. Furthermore, we show that our method can be used for multi- objective MSSC clustering, including constrained multi-objective clustering. ispartof: pages:462-470 ispartof: Proceedings of the 22nd European Conference on Artificial Intelligence (ECAI16) vol:285 pages:462-470 ispartof: European Conference on Artificial Intelligence (ECAI16) location:The Hague, The Netherlands date:29 Aug - 2 Sep 2016 status: published

285 1722 1723 Sponsors: City of The Hague Essence ITN Network PricewaterhouseCoopers Taylor and Francis Group Vrije Universiteit Amsterdam Conference code: 126200 Cited By :1

Clustering is an underspecified task: there are no universal criteria for what makes a good clustering. This is especially true for relational data, where similarity can be based on the features of individuals, the relationships between them, or a mix of both. Existing methods for relational clustering have strong and often implicit biases in this respect. In this paper, we introduce a novel similarity measure for relational data. It is the first measure to incorporate a wide variety of types of similarity, including similarity of attributes, similarity of relational context, and proximity in a hypergraph. We experimentally evaluate how using this similarity affects the quality of clustering on very different types of datasets. The experiments demonstrate that (a) using this similarity in standard clustering methods consistently gives good results, whereas other measures work well only on datasets that match their bias; and (b) on most datasets, the novel similarity outperforms even the best among the existing ones. ispartof: pages:1674-1675 ispartof: Frontiers in Artificial Intelligence and Applications vol:285 pages:1674-1675 ispartof: European Conference on Artificial Intelligence location:NETHERLANDS, Hague date:29 Aug - 2 Sep 2016 status: published

Coordination is essential to the effective operation of multi-agent systems. Convention emergence offers a low-cost and decentralised method of ensuring compatible actions and behaviour, without requiring the imposition of global rules. This is of particular importance in environments with no centralised control or where agents belong to different, possibly conflicting, parties. The timely emergence of robust conventions can be facilitated and manipulated via the use of fixed strategy agents, who attempt to influence others into adopting a particular strategy. Although fixed strategy agents have previously been investigated, they have not been considered in dynamic networks. In this paper, we explore the emergence of conventions within a dynamic network, and examine the effectiveness of fixed strategy agents in this context. Using established placement heuristics we show how such agents can encourage convention emergence, and we examine the impact of the dynamic nature of the network. We introduce a new heuristic, Life-Degree, to enable this investigation. Finally, we consider the ability of fixed strategy agents to manipulate already established conventions, and investigate the effectiveness of placement heuristics in this domain.\ud \ud

For many applications it is important to be able to detect what a human is currently doing. This ability is useful for applications such as surveillance, human computer interfaces, games and healthcare. In order to recognize a human action, the typical approach is to use manually labeled data to perform supervised training. This paper aims to compare the performance of several supervised classifiers trained with manually labeled data versus the same classifiers trained with data automatically labeled. In this paper we propose a framework capable of recognizing human actions using supervised classifiers trained with automatically labeled data in RGB-D videos. info:eu-repo/semantics/publishedVersion

In this paper we present a rich decision model for intelligent agents involved in multi issue negotiations. The model, grounded on information theory, takes into account the aspects of trust and preference to devise mechanisms to manage dialogues. The model supports the design of agents that aim to take informed decisions taking into account that which they have actually observed.

This paper describes how cognitive modeling can be exploited in the design of software agents that support naval training sessions. The architecture, specifications, and embedding of the cognitive agent in a simulation environment are described. Subsequently, the agent's functioning was evaluated in complex, real life training situations for naval officers.

Emergent processes are non-routine, collaborative business processes whose execution is guided by the knowledge that emerges during a process instance. In so far as the process goal gives direction to conventional business processes, the continually evolving process knowledge gives direction to emergent processes. Emergent processes may involve informal interaction, and so there is a limit to the extent to which the processes can be "managed". The collaboration however can be managed. Managing collaboration needs an intelligent agent that is guided not by a process goal, but by observing the performance of the other agents. Each agent has process knowledge --- that is information either generated by the individual users or is extracted from the environment, and performance knowledge --- that describes how the other agents, together with their 'owners', perform --- including how reliable they are. The integrity of the information derived from past observations decays in time, and so they have an inference mechanism that can cope with information of decaying integrity. An agent is described that achieves this by using ideas from information theory. The agents' internal representation language is probabilistic first-order logic. They derive models of the other agents using entropy-based inference that is based on random worlds. Maximum entropy inference is used to construct these models that are then refreshed as new information is received using minimum relative entropy inference.

[This corrects the article DOI: 10.1007/s11023-020-09527-6.]., (© Springer Nature B.V. 2021.)