Skip to main content

About this book

These transactions publish research in computer-based methods of computational collective intelligence (CCI) and their applications in a wide range of fields such as the semantic Web, social networks, and multi-agent systems. TCCI strives to cover new methodological, theoretical and practical aspects of CCI understood as the form of intelligence that emerges from the collaboration and competition of many individuals (artificial and/or natural). The application of multiple computational intelligence technologies, such as fuzzy systems, evolutionary computation, neural systems, consensus theory, etc., aims to support human and other collective intelligence and to create new forms of CCI in natural and/or artificial systems. This twentieth issue contains 11 carefully selected and revised contributions.

Table of Contents


Developing Embodied Agents for Education Applications with Accurate Synchronization of Gesture and Speech

Embodied agents have great potential for education field, which are promising to maximize the learner’s learning gains and enjoyment. In many education applications, multimodal representation of embodied agents is a powerful approach for obtaining the above benefit, which requires accurate synchronization of gesture and speech. For this purpose, we investigate the important issues in synchronization as a practical guideline for our algorithm design through a precedent case study and propose a two-step synchronization method. Our case study reveals that two issues (i.e. duration and timing) play an important role in synchronizing of gesture with speech. Considering the synchronization problem as a motion synthesis problem instead of a behavior scheduling problem used in the conventional methods, we employ a motion graph technique with constraints on gesture structure for coarse synchronization in a first step and refine this further by shifting and scaling the gesture in a second step. Subjective evaluation has demonstrated that the proposed method achieves more accurate synchronization with respect to both duration and timing, and higher motion quality than the state-of-the-art methods.
Furthermore, we have implemented the proposed synchronization method in an authoring tool for education applications. We have conducted several experiments in a university, whose results have demonstrated that our system makes the creation of attractive animations easier and faster (only about 10 % operation time) than manual creation of equal quality, and it is effective to use embodied agents in education applications.
Jianfeng Xu, Yuki Nagai, Shinya Takayama, Shigeyuki Sakazawa

Abstraction of Heterogeneous Supplier Models in Hierarchical Resource Allocation

Resource allocation problems such as finding a production schedule given a set of suppliers’ capabilities are generally hard to solve due to their combinatorial nature, in particular beyond a certain problem size. Large-scale instances among them, however, are prominent in several applications relevant to smart grids including unit commitment and demand response. Decomposition constitutes a classical tool to deal with this increasing complexity. We present a hierarchical “regio-central” decomposition based on abstraction that is designed to change its structure at runtime. It requires two techniques: (1) synthesizing several models of suppliers into one optimization problem and (2) abstracting the direct composition of several suppliers to reduce the complexity of high-level optimization problems. The problems we consider involve limited maximal and, in particular, minimal capacities along with on/off constraints. We suggest a formalization termed supply automata to capture suppliers and present algorithms for synthesis and abstraction. Our evaluation reveals that the obtained solutions are comparable to central solutions in terms of cost efficiency (within 1 % of the optimum) but scale significantly better (between a third and a half of the runtime) in the case study of scheduling virtual power plants.
Alexander Schiendorfer, Gerrit Anders, Jan-Philipp Steghöfer, Wolfgang Reif

Shape Recognition Through Tactile Contour Tracing

A Simulation Study
We present Contour-net, a bio-inspired model for tactile contour-tracing driven by an Hopf oscillator. By controlling the rhythmic movements of a simulated insect-like feeler, the model executes both wide searching and local sampling movements. Contour-tracing is achieved by means of contact-induced phase-forwarding of the oscillator. To classify the shape of an object, collected contact events can be directly fed into machine learning algorithms with minimal pre-processing (scaling). Three types of classifiers were evaluated, the best one being a Support Vector Machine. The likelihood of correct classification steadily increases with the number of collected contacts, enabling an incremental classification during sampling. Given a sufficiently large training data set, tactile shape recognition can be achieved in a position-, orientation- and size-invariant manner. The suitability for robotic applications is discussed.
André Frank Krause, Nalin Harischandra, Volker Dürr

Real-Time Tear Film Classification Through Cost-Based Feature Selection

Dry eye syndrome is an important public health problem, and can be briefly defined as a symptomatic disease which affects a wide range of population and has a negative impact on their daily activities. In clinical practice, it can be diagnosed by the observation of the tear film lipid layer patterns, and their classification into one of the Guillon categories. However, the time required to extract some features from tear film images prevents the automatic systems to work in real time. In this paper we apply a framework for cost-based feature selection to reduce this high computational time, with the particularity that it takes the cost into account when deciding which features to select. Specifically, three representative filter methods are chosen for the experiments: Correlation-Based Feature Selection (CFS), minimum-Redundancy-Maximum-Relevance (mRMR) and ReliefF. Results with a Support Vector Machine as a classifier showed that the approach is sound, since it allows to reduce considerably the computational time without significantly increasing the classification error.
Verónica Bolón-Canedo, Beatriz Remeseiro, Noelia Sánchez-Maroño, Amparo Alonso-Betanzos

Scalarized and Pareto Knowledge Gradient for Multi-objective Multi-armed Bandits

A multi-objective multi-armed bandit (MOMAB) problem is a sequential decision process with stochastic reward vectors. We extend knowledge gradient (KG) policy to the MOMAB problem, and we propose Pareto-KG and scalarized-KG algorithms. The Pareto-KG trades off between exploration and exploitation by combining KG policy with Pareto dominance relations. The scalarized-KG makes use of a linear or non-linear scalarization function to convert the MOMAB problem into a single-objective multi-armed bandit problem and uses KG policy to trade off between exploration and exploitation. To measure the performance of the proposed algorithms, we introduce three regret measures. We compare empirically the performance of the KG policy with UCB1 policy on a test suite of MOMAB problems with normal distributions. The Pareto-KG and scalarized-KG are the algorithms with the best empirical performance.
Saba Yahyaa, Madalina M. Drugan, Bernard Manderick

Extensibility Based Multiagent Planner with Plan Diversity Metrics

Coordinated sequential decision making of a team of cooperative agents is described by principles of multiagent planning. In this work, we extend the MA-Strips formalism with the notion of extensibility and reuse a well-known initiator–participants scheme for agent negotiation. A multiagent extension of the Generate-And-Test principle is used to distributively search for a coordinated multiagent plan. The generate part uses a novel plan quality estimation technique based on metrics often used in the field of diverse planning. The test part builds upon planning with landmark actions by compilation to classic planning. We designed a new multiagent planning domain which illustrates the basic properties of the proposed multiagent planning approach. Finally, our approach was experimentally evaluated on four classic IPC benchmark domains modified for multiagent settings. The results show (1) which combination of plan quality estimation and (2) which diversity metrics provide the best planning efficiency.
Jan Tožička, Jan Jakubův, Karel Durkota, Antonín Komenda

Concurrent and Distributed Shortest-Path Searches in Multiagent-Based Transport Systems

The Fourth Industrial Revolution and the consequent integration of the Internet of Things and Services into industrial processes increase the requirements of transport processes. Customer demanding same-day deliveries, shorter transit-times, individual qualities of shipments, and higher amounts of small size orders raise the complexity and dynamics in logistics. In these highly dynamic environments, multiagent systems (MAS) and multiagent-based simulation (MASB) offer a suitable approach to handle the complexity and to provide the required flexibility, robustness, as well as customized behavior. This article focuses on the impact and the relevance of shortest-path queries in MAS and MABS. It compares the application of state-of-the-art algorithms and investigates different modeling approaches for efficient and concurrent shortest-path searches. The results prove that the application of a highly efficient algorithm such as hub labeling with contraction hierarchies is an essential key component in the agent-based control of dynamic transport processes. Moreover, the results reveal that choosing a modeling approach which slightly restricts the agents’ autonomy increases significantly the runtime performance without losing the advantages of multiagent systems. This allows for applying MAS to solve large scale real-world transport problems and for performing MABS with low hardware requirements.
Max Gath, Otthein Herzog, Maximilian Vaske

SAJaS: Enabling JADE-Based Simulations

Multi-agent systems (MAS) are widely acknowledged as an appropriate modelling paradigm for distributed and decentralized systems, where a (potentially large) number of agents interact in non-trivial ways. Such interactions are often modelled defining high-level interaction protocols. Open MAS typically benefit from a number of infrastructural components that enable agents to discover their peers at run-time. On the other hand, multi-agent-based simulations (MABS) focus on applying MAS to model complex social systems, typically involving a large agent population. Several MAS development frameworks exist, but they are often not appropriate for MABS; and several MABS frameworks exist, albeit sharing little with the former. While open agent-based applications benefit from adopting development and interaction standards, such as those proposed by FIPA, MABS frameworks typically do not support them. In this paper, a proposal to bridge the gap between MAS simulation and development is presented, including two components. The Simple API for JADE-based Simulations (SAJaS) enhances MABS frameworks with JADE-based features. While empowering MABS modellers with modelling concepts offered by JADE, SAJaS also promotes a quicker development of simulation models for JADE programmers. In fact, the same implementation can, with minor changes, be used as a large scale simulation or as a distributed JADE system. In its current version, SAJaS is used in tandem with the Repast simulation framework. The second component of our proposal consists of a MAS Simulation to Development (MASSim2Dev) tool, which allows the automatic conversion of a SAJaS-based simulation into a JADE MAS, and vice-versa. SAJaS provides, for certain kinds of applications, increased simulation performance. Validation tests demonstrate significant performance gains in using SAJaS with Repast when compared with JADE, and show that the usage of MASSim2Dev preserves the original functionality of the system.
Henrique Lopes Cardoso

Strategic Negotiation and Trust in Diplomacy – The DipBlue Approach

The study of games in Artificial Intelligence has a long tradition. Game playing has been a fertile environment for the development of novel approaches to build intelligent programs. Multi-agent systems (MAS), in particular, are a very useful paradigm in this regard, not only because multi-player games can be addressed using this technology, but most importantly because social aspects of agenthood that have been studied for years by MAS researchers can be applied in the attractive and controlled scenarios that games convey. Diplomacy is a multi-player strategic zero-sum board game, including as main research challenges an enormous search tree, the difficulty of determining the real strength of a position, and the accommodation of negotiation among players. Negotiation abilities bring along other social aspects, such as the need to perform trust reasoning in order to win the game. The majority of existing artificial players (bots) for Diplomacy do not exploit the strategic opportunities enabled by negotiation, focusing instead on search and heuristic approaches. This paper describes the development of DipBlue, an artificial player that uses negotiation in order to gain advantage over its opponents, through the use of peace treaties, formation of alliances and suggestion of actions to allies. A simple trust assessment approach is used as a means to detect and react to potential betrayals by allied players. DipBlue was built to work with DipGame, a MAS testbed for Diplomacy, and has been tested with other players of the same platform and variations of itself. Experimental results show that the use of negotiation increases the performance of bots involved in alliances, when full trust is assumed. In the presence of betrayals, being able to perform trust reasoning is an effective approach to reduce their impact.
André Ferreira, Henrique Lopes Cardoso, Luís Paulo Reis

Overcoming Limited Onboard Sensing in Swarm Robotics Through Local Communication

In swarm robotics systems, the constituent robots are typically equipped with simple onboard sensors of limited quality and range. In this paper, we propose to use local communication to enable sharing of sensory information between neighboring robots to overcome the limitations of onboard sensors. Shared information is used to compute readings for virtual, collective sensors that, to a control program, are indistinguishable from a robot’s onboard sensors. We evaluate two implementations of collective sensors: one that relies on sharing of immediate sensory information within a local frame of reference, and another that relies on sharing of accumulated sensory information within a global frame of reference. We compare performance of swarms using collective sensors with: (i) swarms in which robots only use their onboard sensors, and (ii) swarms in which the robots have idealized sensors. Our experimental results show that collective sensors significantly improve the swarm’s performance by effectively extending the capabilities of the individual robots.
Tiago Rodrigues, Miguel Duarte, Margarida Figueiró, Vasco Costa, Sancho Moura Oliveira, Anders Lyhne Christensen

A Question of Balance

The Benefits of Pattern-Recognition When Solving Problems in a Complex Domain
The dual-process theory of human cognition proposes the existence of two systems for decision-making: a slower, deliberative, problem-solving system and a quicker, reactive, pattern-recognition system. We alter the balance of these systems in a number of computational simulations using three types of agent equipped with a novel, hybrid, human-like cognitive architecture. These agents are situated in the stochastic, multi-agent Tileworld domain, whose complexity can be precisely controlled and widely varied. We explore how agent performance is affected by different balances of problem-solving and pattern-recognition, and conduct a sensitivity analysis upon key pattern-recognition system variables. Results indicate that pattern-recognition improves agent performance by as much as 36.5 % and, if a balance is struck with particular pattern-recognition components to promote pattern-recognition use, performance can be further improved by up to 3.6 %. This research is of interest for studies of expert behaviour in particular, and AI in general.
Martyn Lloyd-Kelly, Fernand Gobet, Peter C. R. Lane


Additional information

Premium Partner

    Image Credits