Skip to main content

Über dieses Buch

This volume collects a selection of contributions which has been presented at the 23rd Italian Workshop on Neural Networks, the yearly meeting of the Italian Society for Neural Networks (SIREN). The conference was held in Vietri sul Mare, Salerno, Italy during May 23-24, 2013. The annual meeting of SIREN is sponsored by International Neural Network Society (INNS), European Neural Network Society (ENNS) and IEEE Computational Intelligence Society (CIS). The book – as well as the workshop- is organized in two main components, a special session and a group of regular sessions featuring different aspects and point of views of artificial neural networks, artificial and natural intelligence, as well as psychological and cognitive theories for modeling human behaviors and human machine interactions, including Information Communication applications of compelling interest.





Identifying Emergent Dynamical Structures in Network Models

The identification of emergent structures in dynamical systems is a major challenge in complex systems science. In particular, the formation of intermediate-level dynamical structures is of particular interest for what concerns biological as well as artificial network models. In this work, we present a new technique aimed at identifying clusters of nodes in a network that behave in a coherent and coordinated way and that loosely interact with the remainder of the system. This method is based on an extension of a measure introduced for detecting clusters in biological neural networks. Even if our results are still preliminary, we have evidence for showing that our approach is able to identify these “emerging things” in some artificial network models and that it is way more powerful than usual measures based on statistical correlation. This method will make it possible to identify mesolevel dynamical structures in network models in general, from biological to social networks.

Marco Villani, Stefano Benedettini, Andrea Roli, David Lane, Irene Poli, Roberto Serra

Experimental Guidelines for Semantic-Based Regularization

This paper presents a novel approach for learning with constraints called Semantic-Based Regularization. This paper shows how prior knowledge in form of First Order Logic (FOL) clauses, converted into a set of continuous constraints and integrated into a learning framework, allows to jointly learn from examples and semantic knowledge. A series of experiments on artificial learning tasks and application of text categorization in relational context will be presented to emphasize the benefits given by the introduction of logic rules into the learning process.

Claudio Saccà, Michelangelo Diligenti, Marco Gori

A Preliminary Study on Transductive Extreme Learning Machines

Transductive learning is the problem of designing learning machines that succesfully generalize only on a given set of input patterns. In this paper we begin the study towards the extension of Extreme Learning Machine (ELM) theory to the transductive setting, focusing on the binary classification case. To this end, we analyze previous work on Transductive Support Vector Machines (TSVM) learning, and introduce the Transductive ELM (TELM) model. Contrary to TSVM, we show that the optimization of TELM results in a purely combinatorial search over the unknown labels. Some preliminary results on an artifical dataset show substained improvements with respect to a standard ELM model.

Simone Scardapane, Danilo Comminiello, Michele Scarpiniti, Aurelio Uncini

Avoiding the Cluster Hypothesis in SV Classification of Partially Labeled Data

We propose a Support Vector-based methodology for learning classifiers from partially labeled data. Its novelty stands in a formulation not based on the

cluster hypothesis

, stating that learning algorithms should search among classifiers whose decision surface is far from the unlabeled points. On the contrary, we assume such points as specimens of uncertain labels which should lay in a region containing the decision surface. The proposed approach is tested against synthetic data sets and subsequently applied to well-known benchmarks, attaining better or at least comparable performance w.r.t. methods described in the literature.

Dario Malchiodi, Tommaso Legnani

Learning Capabilities of ELM-Trained Time-Varying Neural Networks

System identification in nonstationary environments surely represents a challenging problem. The authors have recently proposed an innovative neural architecture, namely Time-Varying Neural Network (TV-NN), which has shown remarkable identification capabilities in this kind of scenarios. It is characterized by time-varying weights, each being a linear combination of a certain set of basis functions. This inevitably increases the network complexity with respect to the stationary NN counterpart and in order to keep the training time low, an Extreme Learning Machine (ELM) approach has been proposed by the same authors for TV-NN learning, instead of Back-Propagation based techniques. However the learning capabilities of TV-NN trained by means of ELM have not been investigated in the literature and in this contribution such a lack is faced: the theoretical foundations of ELM usage for TV-NN are analytically discussed, by extending the corresponding results obtained in the stationary case study.

Stefano Squartini, Yibin Ye, Francesco Piazza

A Quality-Driven Ensemble Approach to Automatic Model Selection in Clustering

A fundamental limitation of the data clustering task is that it has an inherent, ill-defined model selection problem: the choice of a clustering technique also implies some a-priori decision on cluster geometry. In this work we explore the combined use of two different clustering paradigms and their combination by means of an ensemble technique. Mixing coefficients are computed on the basis of partition quality, so that the ensemble is automatically tuned so as to give more weight to the best-performing (in terms of the selected quality indices) clustering method.

Raffaella Rosasco, Hassan Mahmoud, Stefano Rovetta, Francesco Masulli

An Adaptive Reference Point Approach to Efficiently Search Large Chemical Databases

The ability to rapidly search large repositories of molecules is a crucial task in chemoinformatics. In this work we propose AOR, an approach based on adaptive reference points to improve state of the art performances in querying large repositories of binary fingerprints basing on the Tanimoto distance. We propose a unifying view between the context of reference points and the previously proposed hashing techniques. We also provide a mathematical model to forecast and generalize the results, that is validated by simulating queries over an excerpt of the ChemDB. Clustering techniques are finally introduced to improve the performances. For typical situations the proposed algorithm is shown to resolve queries up to 4 times faster than compared methods.

Francesco Napolitano, Roberto Tagliaferri, Pierre Baldi

A Methodological Proposal for an Evolutionary Approach to Parameter Inference in MURAME-Based Problems

In this paper we propose an evolutionary approach in order to infer the values of the parameters for applying the MURAME, a multicriteria method which allows to score/rank a set of alternatives according to a set of evaluation criteria. This problem, known as preference disaggregation, consists in finding the MURAME parameter values that minimize the inconsistency between the model obtained with those parameters and the true preference model on the basis of a reference set of decisions of the Decision Maker. In order to represent a measure of inconsistency of the MURAME model compared to the true preference one, we consider a fitness function which puts emphasis on the distance between the scoring of the alternatives given by the Decision Maker and the one determined by the MURAME. The problem of finding a numerical solution of the involved mathematical programming problem is tackled by using an evolutionary solution algorithm based on the Particle Swarm Optimization. An application is finally provided in order to give an initial assessment of the proposed approach.

Marco Corazza, Stefania Funari, Riccardo Gusso

Genetic Art in Perspective

Since the pioneer observations of Alan Turing, emotional and aesthetical capabilities have been considered as one of the fundamental element of a genuinely intelligent machine. Among the proposed approaches, genetic algorithms try to combine intuitively a generative impulse with a critical capacity that steers the production towards a valuable goal. The approach here presented is based on Karl Sim’s approach in which a set of possible primitives is defined and it represent the genotype of the system. Such expressions are combined using genetic algorithms rules to obtain more complex functions that describe new images. At each step, images are evaluated by the user and this implicitly drives the evolution process. Results can be impressive, however a clear understanding of the determinants of our aesthetic evaluation is presently beyond reach.

Rachele Bellini, N. Alberto Borghese

Signal Processing


Proportionate Algorithms for Blind Source Separation

In this paper we propose an extension of time-domain Blind Source Separation algorithms by applying the well known proportionate and improved proportionate adaptive algorithms. These algorithms, known in the context of adaptive filtering, are able to use the sparseness of acoustic impulse responses of mixing environments and give better performances than standard algorithms. Some preliminary experimental results show the effectiveness of the proposed approach in terms of convergence speed.

Michele Scarpiniti, Danilo Comminiello, Simone Scardapane, Raffaele Parisi, Aurelio Uncini

Pupillometric Study of the Dysregulation of the Autonomous Nervous System by SVM Networks

Pupil is controlled by the autonomous nervous system. Patients with temporomandibular disorders (TMD) and with obstructive sleep apnea syndrome (OSAS) are affected by a dysregulation of the autonomous system. Pupillometry is here used to investigate the state of the autonomous system in 3 groups: control, TMD and OSAS. Different indexes are extracted from the pupillogram to characterize pupil dynamics investigated in rest and under stationary stimulations. All possible sets of 3 and 4 indexes are used as features to train support vector machines (SVM) to identify the correct groups. The indexes providing optimal classification are identified.

Luca Mesin, Ruggero Cattaneo, Annalisa Monaco, Eros Pasero

A Memristor Circuit Using Basic Elements with Memory Capability

After the introduction of the memory–resistor (i.e. memristor), a fundamental two-terminal circuit element defined as a nonlinear relationship between the integral of the voltage and the integral of the current, the class of memristor-based systems was extended by L.O. Chua and S. Kang in 1976. In the literature, the research interest devoted to the discover of novel physical systems with memristor behavior is growing. In 2012, an elementary electronic circuit discovered by F. Corinto and A. Ascoli, based on a Graetz bridge loaded with a RLC filter, was classified to be a memrisor-based system without memory capability. In this paper, the possibility of adding memory on the memristor-based system proposed in 2012 is exploited. In fact, the circuit proposed by F. Corinto and A. Ascoli does not have the memory capability, one of the more important characteristics of the memristor. The memory is added to the system using elementary components in a transfer charge circuit. Results show the memristor-based nature and the memory capability of the presented electronic system.

Amedeo Troiano, Fernando Corinto, Eros Pasero

Effects of Pruning on Phase-Coding and Storage Capacity of a Spiking Network

Synaptic pruning is a crucial process during development. We study the imprinting and replay of spatiotemporal patterns in a spiking network, as a function of pruning degree. After a Spike Timing Dependent Plasticity-based learning of synaptic efficacies, the weak synapses are removed through a competitive pruning process. Surprisingly, after this pruning stage, the storage capacity for spatiotemporal patterns is relatively high also for very high diluition ratio.

Silvia Scarpetta, Antonio De Candia

Predictive Analysis of the Seismicity Level at Campi Flegrei Volcano Using a Data-Driven Approach

This work aims to provide a short-term tool to estimate the possible trend of the seismicity level in the area of Campi Flegrei (southern Italy) for Civil Protection purposes. During the last relevant period of seismic activity, between 1982 and 1984, an uplift of the ground (bradyseism) of more than 1.5 m occurred. It was accompanied by more than 16,000 earthquakes up to magnitude 4.2 which forced the civil authorities to order the evacuation of about 40,000 people from Pozzuoli town for several months. Scientific studies evidenced a temporal correlation between these geophysical phenomena. This has led us to consider a data-driven approach to obtain a forecast of the seismicity level for this area. In particular, a technique based on a Multilayer Perceptron (MLP) network has been used for this intent. Neural networks are data processing mechanisms capable of relating input data with output ones without any prior correlation model but only using empirical evidences obtained from the analysis of available data. The proposed method has been tested on a set of seismic and deformation data acquired between 1983 and 1985 and then including the data of the aforementioned crisis which affected the Campi Flegrei. Once defined the seismicity levels on the basis of the maximum magnitude recorded within a week, three MLP networks were implemented with respectively 2, 3 and 4 output classes. The first network (2 classes) provides only an indication about the possible occurrence of earthquakes felt by people (with magnitude higher than 1.7), while the remaining nets (3 and 4 classes) give also a rough suggestion of their intensity. Furthermore, for these last two networks one of the output classes allows to obtain a forecast about the possible occurrence of strong potentially damaging earthquakes with magnitude higher than 3.5. Each network has been trained on a fixed interval and then tested for the forecast on the subsequent period. The results show that the performance decreases as a function of the complexity of the examined task that is the number of covered classes. However, the obtained results are very promising, for which the proposed system deserves further studies since it could be of support to the Civil Protection operations in the case of possible future crises.

Antonietta M. Esposito, Luca D’Auria, Andrea Angelillo, Flora Giudicepietro, Marcello Martini

Robot Localization by Echo State Networks Using RSS

In this paper we present an application of Reservoir Computing to indoor robot localization, based on input received signal strength signals from a wireless sensor network. The proposed localization system allows to combine good predictive performance with particularly efficient and practical solutions. Promising results are shown in preliminary experiments on a real-world scenario.

Stefano Chessa, Claudio Gallicchio, Roberto Guzman, Alessio Micheli

An Object Based Analysis Applied to Very High Resolution Remote Sensing Data for the Change Detection of Soil Sealing at Urban Scale

An object-based strategy is presented to identify soil-sealing in urban environment using Very High Resolution (VHR) remote sensing images. A first stage of segmentation has been carried out using a watershed algorithm, and then a second stage of supervised classification has been done to classify land covers. The resulting land covers have been used to discriminate between sealed and unsealed surfaces. The selection of features and of the number of land cover classes has been guided by an exploratory clustering stage. The proposed strategy has been used to classify sealed surfaces for two single-date images. The post-classification results have been used in a Change Detection Task, in order to analyze the land take problem in a periurban area of Venezia Mestre between 2005 and 2010. The change detection results are promising, considering the good capacity to reveal changes at the characteristic dimension of small man-made structures, and they can be considered a good support to a successive photointerpretation step.

Luca Pugliese, Silvia Scarpetta

EEG Complexity Modifications and Altered Compressibility in Mild Cognitive Impairment and Alzheimer’s Disease

The objective of this work is to respond to the question: can quantitative electroencephalography (EEG) distinguish among Alzheimer’s Disease (AD) patients, mild cognitive impaired (MCI) subjects and elderly healthy controls? In other words, are there nonlinear indexes extracted from raw EEG data that are able to manifest the background difference among EEG? The response we give here is that a synthetic index of entropic complexity (Permutation Entropy, PE) as well as a measure of compressibility of the EEG can be used to discriminate among classes of subjects. An experimental database has been analyzed to make these measurements and the results we achieved are encouraging also in terms of disease evolution. Indeed, it is clearly shown that the condition of MCI has intermediate properties with respect to the analyzed markers: thus, these markers could in principle be used to evaluate the probability of transition from MCI to mild AD.

Domenico Labate, Fabio La Foresta, Isabella Palamara, Giuseppe Morabito, Alessia Bramanti, Zhilin Zhang, Francesco C. Morabito

Smart Home Task and Energy Resource Scheduling Based on Nonlinear Programming

The computational intelligence community has invested many efforts in the last few years on the challenging problem of automatic task and energy resources scheduling in smart home contexts. Moving from a recent work of some of the authors, jointly considering the electrical and thermal comfort needs of the user, in this paper a nonlinear optimization framework, namely “Mixed-Integer Nonlinear Programming”, is proposed on purpose. It allows dealing with nonlinearities resulting from the constraints imposed by the involved building thermal model, which was not feasible in the original linear approach. Performed computer simulations related to a realistic domestic scenario have shown that a certain improvement is attainable in terms of satisfaction of user thermal requirements, attaining at the same time an enhanced overall energy cost reduction with respect to the non-optimized scheduling strategy.

Severini Marco, Stefano Squartini, Gian Piero Surace, Francesco Piazza



Data Fusion Using a Factor Graph for Ship Tracking in Harbour Scenarios

Data coming from cameras deployed along an harbour coastline are fused to extract the state of unknown vessels framed by the sensors. We embed the ship dynamic model into a Factor Graph that through probability propagation provides a very flexible merge of sensory data and inferences. Preliminary results and experiments from videos gathered in the Gulf of Naples are reported with a discussion on future trends.

Francesco Castaldo, Francesco A. N. Palmieri

Reinforcement Learning for Automated Financial Trading: Basics and Applications

The construction of automated financial trading systems (FTSs) is a subject of high interest for both the academic environment and the financial one due to the potential promises by self-learning methodologies. In this paper we consider Reinforcement Learning (RL) type algorithms, that is algorithms that real-time optimize their behavior in relation to the responses they get from the environment in which they operate, without the need for a supervisor. In particular, first we introduce the essential aspects of RL which are of interest for our purposes, second we present some original automatic FTSs based on differently configured RL-based algorithms, then we apply such FTSs to artificial and real time series of daily stock prices. Finally, we compare our FTSs with a classical one based on Technical Analysis indicators. All the results we achieve are generally quite satisfactory.

Francesco Bertoluzzo, Marco Corazza

A Collaborative Filtering Recommender Exploiting a SOM Network

Recommender systems are exploited in many fields for helping users to find goods and services. A collaborative filtering recommender realizes a knowledge-sharing system to find people having similar interests. However, some critical issues may lead to inaccurate suggestions. To provide a solution to such problems, this paper presents a novel SOM-based collaborative filtering recommender. Some experimental results confirm the effectiveness of the proposed solution.

Giuseppe M. L. Sarnè

SVM Tree for Personalized Transductive Learning in Bioinformatics Classification Problems

Personalized modelling joint with Transductive Learning (PTL) uses a particular local modelling (personalized) around a single point for classification of each test sample, thus it is basically neighbourhood dependent. Usually existing PTL methods define the neighbourhood using a (dis)similarity measure, in this paper we propose a new transductive SVM classification tree (tSVMT) based on PTL. The neighbourhood of a test sample is built over the classification knowledge modelled by regional SVMs, and a set of such SVMs adjacent to the test sample are aggregated systematically into a tSVMT. Compared to a normal SVM/SVMT approach, the proposed tSVMT, with the aggregation of SVMs, improves classifying power in terms of accuracy on bioinformatics database. Moreover, tSVMT seems to solve the over-fitting problem of all previous SVMTs as it aggregates neighbourhood knowledge, significantly reducing the size of the SVM tree.

Maurizio Fiasché

Multi-Country Mortality Analysis Using Self Organizing Maps

In this paper we introduce the use of Self Organizing Maps (SOMs) in multidimensional mortality analysis. The rationale behind this contribution is that patterns of mortality in different areas of the world are becoming more and more related; a fast and intuitive method understanding the similarities among mortality experiences could therefore be of aid to improve the knowledge on this complex phenomenon. The results we have obtained highlight common features in the mortality experience of various countries, hence supporting the idea that SOM may be a very effective tool in this field.

Gabriella Piscopo, Marina Resta

A Fuzzy Decision Support System for the Environmental Risk Assessment of Genetically Modified Organisms

Aim of the paper is the development of a Fuzzy Decision Support System (FDSS) for the Environmental Risk Assessment (ERA) of the deliberate release of genetically modified plants. The evaluation process permits identifying potential impacts that can achieve one or more receptors through a set of migration paths. ERA process is often performed in presence of incomplete and imprecise data and is generally yielded using the personal experience and knowledge of the human experts. Therefore the risk assessment in the FDSS is obtained by using a Fuzzy Inference System (FIS), performed using jFuzzyLogic library. The decisions derived by FDSS have been validated on real world cases by the human experts that are in charge of ERA. They have confirmed the reliability of the fuzzy support system decisions.

Francesco Camastra, Angelo Ciaramella, Valeria Giovannelli, Matteo Lener, Valentina Rastelli, Antonino Staiano, Giovanni Staiano, Alfredo Starace

Adaptive Neuro-Fuzzy Inference Systems vs. Stochastic Models for Mortality Data

A comparative analysis is done between stochastic models and Adaptive Neuro–Fuzzy Inference System applied to the projection of the longevity trend. The stochastic models provides the heuristic rule for obtaining projections. In the context of ANFIS models, the fuzzy logic allows for determining the learning algorithm on the basis of the relationship between inputs and outputs. In other words the rule is here deducted by the actual mortality data, because this allows for fuzzy systems to learn from the data they are modelling. This is possible by computing the membership function parameters that best allow the associated fuzzy inference system to track the input/output data. The literature indicates that the self-predicting model of ANFIS is better than other models in a lot of fields. Shortcomings and advantages of both approaches are here highlighted.

Valeria D’Amato, Gabriella Piscopo, Maria Russolillo

Special Session on “Social and Emotional Networks for Interactional Exchanges”


Recent Approaches in Handwriting Recognition with Markovian Modelling and Recurrent Neural Networks

Handwriting recognition is challenging because of the inherent variability of character shapes. Popular approaches for handwriting recognition are markovian and neuronal. Both approaches can take as input, sequences of frames obtained by sliding a window along a word or a text-line. We present markovian (Dynamic Bayesian Networks, Hidden Markov Models) and recurrent neural network-based approaches (RNNs) dedicated to character, word and text-line recognition. These approaches are applied to the recognition of both Latin and Arabic scripts.

Laurence Likforman-Sulem

Do Relationships Exist between Brain-Hand Language and Daily Function Characteristics of Children with a Hidden Disability?

Objective: To discover whether children with a hidden disability such as Developmental Coordination Disorders (DCD) have unique brain-hand language (handwriting) and daily function characteristics and whether there are relationships between these characteristics.

Method: 20 children diagnosed with DCD and 20 typically developed controls aged 7-10 performed the Alphabet writing task on a page affixed to an electronic tablet, a component of the ComPET which documented their handwriting process. Further, their organizational ability was evaluated through daily function events as reported by their parents.

Results: Significant group differences (DCD versus controls) were found in the coefficient of variance of spatial, temporal and pressure writing process measures. Specific handwriting measures predicted the level of children’s organization abilities through daily function.

Conclusions: These results emphasize the need for further development of sophisticated computerized methods so as to gain deeper insight concerning daily function characteristics of children with hidden disabilities.

Sara Rosenblum, Miri Livneh-Zirinski

Corpus Linguistics and the Appraisal Framework for Retrieving Emotion and Stance – The Case of Samsung’s and Apple’s Facebook Pages

The study investigated the situated linguistic interactions of the users of the Samsung and Apple Facebook pages, with a focus on the attitudinal/affectual values they displayed towards these brands and their products, in a comparative perspective. Following Corpus Linguistics (CL) methodology, two corpora were created, named AppleCorpus (7337 tokens) and SamsungCorpus (5216 tokens), consisting in the wall posts on Apple Inc. and Samsung Mobile pages collected over a period of four days. These corpora were scrutinized both in a CL quantitative perspective and in a qualitative perspective by using the resources of the Appraisal Framework (AF) for discourse analysis to better identifying these social network users’ stance and attitudinal positioning. The findings of this pilot study showed that Samsung’s users display a more positive attitude toward the brand than Apple’s users. Results are discussed in the text.

Amelia Regina Burns, Olimpia Matarazzo, Lucia Abbamonte

Which Avatars Comfort Children?

In this paper we give an account of an empirical study related to avatar selection, whose purpose is to comfort children in the context of the TERENCE learning application. We investigated what Hungarian children of ages 6-11 like to play with, what stories they like to read, and what avatars – out of a selection of nine different designs – they like best. We found a statistically relevant correlation between age, gender, and the participants’ expertise and habits as regards ICT tools. Our studies thus provided relevant information on how to design avatars for this special age group, taking gender equality into consideration and increasingly subjective well-being in order to motivate learning.

Judit Bényei, Anikó Illés, Gabriella Pataky, Zsófia Ruttkay, Andrea Schmidt

The Effects of Hand Gestures on Psychosocial Perception: A Preliminary Study

To date a few studies have experimentally investigated the effects of hand gestures (and frequency) on psychosocial perception. A preliminary study with two experiments were conducted, in which confederates manipulated “Type” (rhythmic gestures, cohesive gestures and self-adaptors for experiment 1; rhythmic gestures, focusing gestures and dynamic gestures for experiment 2) and “Frequency” (low and high) during a face-to-face conversation with the participants. ANOVAs reveal that rhythmic gestures influence positively competence perception but negatively conversational fairness, self-adaptors increase warmth evaluation and high frequency influences positively warmth and dominance perceptions. Hand gestures appear to play a causal role in psychosocial evaluation.

Augusto Gnisci, Antonio Pace

The Influence of Positive and Negative Emotions on Physiological Responses and Memory Task Scores

The present paper report results of a preliminary study devoted to investigate whether and how different induced emotional states influence physiological responses and memory task scores. Physiological responses, such as skin conductance (SCL) and heart rate (HR) values were measured from 32 university students, before, during and after they were elicited by video stimuli. The considered stimuli were able to induce positive, negative and neutral emotional states. The specific physiological activation patterns were identified and correlated with memory task scores, computed using the “Anna Pesenti” Story Recall Test (SRT).

The results show significant changes in physiological values when positive (increase in HR values) and negative (increase in SCL values) emotional states are induced. Surprisingly, increased SCL values, associated to induced positive emotional states, affect the participant’s memory task scores.

Maria Teresa Riviello, Vincenzo Capuano, Gianluigi Ombrato, Ivana Baldassarre, Gennaro Cordasco, Anna Esposito

Mood Effects on the Decoding of Emotional Voices

This study examines the effect of mood induction on the decoding of emotional vocal expressions. An adequate sample of 145 students (71 females, 74 males; mean age = 23.37 ± 2.05) was recruited at the Second University of Naples (Italy). Subjects were randomly assigned to one of three (sad, fear or neutral) emotion conditions induced by viewing short movies. The results showed a significant general decrease in the decoding accuracy in the mood induction conditions when compared to the accuracy of the participants who did not received such mood induction. Post hoc analyses revealed that recognition of emotional vocal voices conveying anger was especially impaired by mood induction conditions. No findings consistent with mood congruity theory were observed. This study contributes to emotion regulation research by showing differences in emotion decoding tasks by voices due to mood induction procedures, as already observed in studies exploiting the decoding of emotional faces.

Alda Troncone, Davide Palumbo, Anna Esposito

The Ascending Reticular Activating System

The Common Root of Consciousness and Attention

In the organization of the central nervous system the role of Ascending Reticular Activating System (ARAS) – comprising the reticular formation, thalamus and thalamo-cortical system of bi-directional projection which governs the activities of wakefulness and vigilance – does not correspond to a hierarchical superiority with respect to the cerebral hemispheres. The ARAS is not limited to the brain stem: it projects upwards towards the cerebral hemispheres and downwards towards the spinal cord. Its functions are much more complex than simple cortical desynchronization, even though this is essential in the state of alertness and attention. Its thalamo-cortical projections, which are a-specific with a high oscillatory frequency, are fundamental for some essential functions of consciousness.

Mauro Maldonato

Conversational Entrainment in the Use of Discourse Markers

Entrainment is the tendency for participants in conversations to develop behaviour similar to one another in multiple dimensions. The degree of such entrainment is linked to the emotional state and empathy of the speakers and people who entrain to their conversational partners are seen as more socially attractive, likeable, competent, more intimate, and the interactions with such partners as more successful. It is thus important that ICT interfaces for supporting wellbeing and empathy employ also some module of entrainment.

In this paper we analyze entrainment in the acoustic, prosodic and pragmatic domains connected to the use of Slovak discourse marker ‘no’ in the spoken modality of task-oriented collaborative dialogues. We analyze how speaking behaviour changes due to interacting with a different partner, and consequently, how entrainment is employed. We use acoustic and prosodic information extracted from the signal and labelled pragmatic functions of the marker (including acknowledgment, backchannel, reservation, topic shift, etc.). Results suggest a varied picture with both entrainment and disentrainment present in the data. Regarding the relationship between entrainment in acoustic-prosodic features and more cognitively complex features of pragmatic meaning and discourse functions, we found both matches and mismatches between the two.

Štefan Beňuš

Language and Gender Effect in Decoding Emotional Information: A Study on Lithuanian Subjects

The present work explores how language specificity and gender affect the emotional decoding process. It investigates the ability of Lithuanian male and female subjects to decode emotional information through male and female vocal and visual emotional expressions. The exploited emotional stimuli are based on extracts of American English (a globally spread language), and Italian (a country specific language) movies. The assumption is that the recognition of the emotional states expressed by the actors/actresses will change according to the familiarity of the languages and the subjects’ gender. Results show that Lithuanian subjects recognition accuracy is affected by the language specificity of the stimuli. Moreover, a gender effect occurs in decoding Italian vocal stimuli.

Maria Teresa Riviello, Rytis Maskeliunas, Jadvyga Kruminiene, Anna Esposito

Preliminary Experiments on Automatic Gender Recognition Based on Online Capital Letters

In this paper we present some experiments to automatically classify online handwritten text based on capital letters. Although handwritten text is not as discriminative as face or voice, we still found some chance for gender classification based on handwritten text. Accuracies are up to 74%, even in the most challenging case of capital letters..

Marcos Faundez-Zanuy, Enric Sesa-Nogueras

End-User Design of Emotion-Adaptive Dialogue Strategies for Therapeutic Purposes

Two fundamental nontechnical research questions related to the development of emotion-aware dialogue systems are how to identify different kinds of emotional reactions that can be expected to occur in a given interaction domain, and how the system should react to the emotional user behavior. These questions are especially important for dialogue systems used in medical treatment of children with developmental disorders. The paper reports on an adaptive dialogue system that allows the therapist to flexibly design and test dialogue strategies. Our aim was to achieve a balance between the ease-of-use of the system by nontechnical users and the flexibility to adapt the system to different therapeutic settings. The system builds on our previous work, and its functionality is explained by means of example.

Milan Gnjatović, Vlado Delić

Modulation of Cognitive Goals and Sensorimotor Actions in Face-to-Face Communication by Emotional States: The Action-Based Approach

Cognitive goals – i.e. the intention to utter a sentence and to produce co-speech facial and hand-arm gestures – as well as the sensorimotor realization of the intended speech, co-speech facial, and co-speech hand-arm actions are modulated by the emotional state of the speaker. In this review paper it will be illustrated how cognitive goals and sensorimotor speech, co-speech facial, and co-speech hand-arm actions are modulated by emotional states of the speaker, how emotional states are perceived and recognized by interlocutors in the context of face-to-face communication, and which brain regions are responsible for production and perception of emotions in face-to-face communication.

Bernd J. Kröger

Investigating the Form-Function-Relation of the Discourse Particle “hm” in a Naturalistic Human-Computer Interaction

For a successful speech-controlled human-computer interaction (HCI) the pure textual information as well as individual skills, preferences, and affective states of the user have to be known. However, verbal human interaction consists of several information layers. Apart from pure textual information, further details regarding the speaker’s feelings, believes, and social relations are transmitted. The additional information is encoded through acoustics. Especially, the intonation reveals details about the speakers communicative relation and their attitude towards the ongoing dialogue.

Since the intonation is influenced by semantic and grammatical information, it is advisable to investigate the intonation of so-called discourse particles (DPs) as “hm” or “uhm”. They cannot be inflected but can be emphasised. DPs have the same intonation curves (pitch-contours) as whole sentences and thus may indicate the same functional meanings. For German language J. E. Schmidt empirically discovered seven types of form-function-concurrences on the isolated DP “hm”.

To determine the function within the dialogue of the DPs, methods are needed that preserve pitch-contours and are feasible to assign defined form-prototypes. Furthermore, it must be investigated which pitch-contours occur in naturalistic HCI and whether these contours are congruent with the findings by linguists.

In this paper we present first results on the extraction and correlation of the DP “hm”. We investigate the different form-function-relations in the naturalistic LAST MINUTE corpus and determine expectable formfunction relations in naturalistic HCI in general.

Ingo Siegert, Dmytro Prylipko, Kim Hartmann, Ronald Böck, Andreas Wendemuth

Intended and Unintended Offence

This paper argues that politeness and impoliteness are integrally related to offence management. The outlines of a semantic theory of linguistic politeness are sketched. As a semantic theory, interfaces to both pragmatics and compositional syntax may be expected, but these are spelled out in companion papers.

Carl Vogel

Conceptual Spaces for Emotion Identification and Alignment

The paper explores a method for emotion identification based on the mapping of emotional terms to a set of emotion eliciting situations. The method has been applied for emotion words from the Bulgarian language. Situations and words have been evaluated using the valence, arousal, and dominance ratings given by participants. Nine clusters of emotion terms and situations have been identified which allowed to map emotion terms with situations. The mapping method and the results for Bulgarian can be used for cross-group or cross-cultural studies of emotions involving various languages. The possible usage of the results obtained for emotion conceptual space evaluation and emotional alignment of people communicating in social networks is also discussed.

Maurice Grinberg, Evgeniya Hristova, Monika Moudova, James Boster

Emotions and Moral Judgment: A Multimodal Analysis

Recent findings in the field of moral psychology suggest that moral judgment results both from emotional processing and deliberate reasoning. The experimental study uses artificial situations that pose moral dilemmas – a human life have to be sacrificed in order to save more lives. Two factors (physical directness of harm and inevitability of death) are varied in order to explore potential differences in emotional processing and their effects on judgment. Multimodal data is collected and analyzed: moral judgments, skin conductance (as a somatic index of affective processing), and response times (as providing information on deliberation process). Personal-impersonal distinction and inevitability of death are found to influence emotions and judgments in moral dilemmas.

Evgeniya Hristova, Veselina Kadreva, Maurice Grinberg

Contextual Information and Reappraisal of Negative Emotional Events

In this study the effect of the contextual-information induced reappraisal on modifying the emotional response elicited by failure has been investigated. To an academic or job setting failure (control condition) it has been added one of two types of contextual information (knowing that many other people failed the same task and knowing that it would be possible to try the failed task again) affecting three dimensions of failure appraisal: responsibility, sharing, and remediability. In an experimental condition both information were added. Four hundred and eighty undergraduates participated in this study. The experimental design was a 2 (negative emotional event) x 4 (contextual information) between-subjects design. The first variable was included in the design as covariate. We expected that generalized failure should imply a decrease of responsibility and an increase of sharing, the possibility of retrying should imply an increase in the remediability, and that the presence of both types of information should produce all the abovementioned effects. Our findings substantially corroborated the hypotheses.

Ivana Baldassarre, Lucia Abbamonte, Marina Cosenza, Giovanna Nigro, Olimpia Matarazzo

Deciding with (or without) the Future in Mind: Individual Differences in Decision-Making

The aim of this study was to examine the influence of propensity to risk taking, impulsivity, and present versus future orientation in decision-making under ambiguity. One hundred and four healthy adults were administered the computer versions of the Iowa Gambling Task (IGT) and the Balloon Analogue Risk Task (BART). They then completed the Barratt Impulsiveness Scale (BIS-11) and the Consideration of Future Consequences Scale (CFC-14). Results indicated that high scores on the BIS-11 Non-Planning impulsivity scale, the CFC-14 Immediate scale, and the BART result in poorer performance on the IGT. In addition, the results of regression analysis showed also that the BART total score was the most powerful predictor of performance on the IGT. The study revealed that individuals who are more prone to risk, less likely to plan ahead carefully, and more oriented to the present, rather than to the future, performed worse on the IGT.

Marina Cosenza, Olimpia Matarazzo, Ivana Baldassarre, Giovanna Nigro


Weitere Informationen

Premium Partner

BranchenIndex Online

Die B2B-Firmensuche für Industrie und Wirtschaft: Kostenfrei in Firmenprofilen nach Lieferanten, Herstellern, Dienstleistern und Händlern recherchieren.



Best Practices für die Mitarbeiter-Partizipation in der Produktentwicklung

Unternehmen haben das Innovationspotenzial der eigenen Mitarbeiter auch außerhalb der F&E-Abteilung erkannt. Viele Initiativen zur Partizipation scheitern in der Praxis jedoch häufig. Lesen Sie hier  - basierend auf einer qualitativ-explorativen Expertenstudie - mehr über die wesentlichen Problemfelder der mitarbeiterzentrierten Produktentwicklung und profitieren Sie von konkreten Handlungsempfehlungen aus der Praxis.
Jetzt gratis downloaden!