Professor and researcher at Universidad Centroccidental Lisandro Alvarado, Venezuela. Computer Engineer (Universidad Centroccidental Lisandro Alvarado, 1997), Master in Computer Science (Universidad de los Andes, 2004), Doctor in Applied Sciences (Universidad de los Andes, 2011), and Doctor in Neuroscience, Cognition and Collective Behavior (Université Paul Sabatier, 2012). Works in the area of Artificial Intelligence (AI) specifically in the following sub-areas: Collective Intelligence, Multiagent Systems, Affective Computing, Emergent Computing applied in: Data Mining, Optimization Problems, Image Processing, design and development of intelligent systems, autonomous, self-organizing and emergent, among others.
Human crowd motion is mainly driven by self-organized processes based on local interactions among... more Human crowd motion is mainly driven by self-organized processes based on local interactions among pedestrians. While most studies of crowd behaviour consider only interactions among isolated individuals, it turns out that up to 70% of people in a crowd are actually moving in groups, such as friends, couples, or families walking together. These groups constitute medium-scale aggregated structures and their impact on crowd dynamics is still largely unknown. In this work, we analyze the motion of approximately 1500 pedestrian groups under natural condition, and show that social interactions among group members generate typical group walking patterns that influence crowd dynamics. At low density, group members tend to walk side by side, forming a line perpendicular to the walking direction. As the density increases, however, the linear walking formation is bent forward, turning it into a V-like pattern. These spatial patterns can be well described by a model based on social communication between group members. We show that the V-like walking pattern facilitates social interactions within the group, but reduces the flow because of its ''non-aerodynamic'' shape. Therefore, when crowd density increases, the group organization results from a trade-off between walking faster and facilitating social exchange. These insights demonstrate that crowd dynamics is not only determined by physical constraints induced by other pedestrians and the environment, but also significantly by communicative, social interactions among individuals.
Affective computing is a discipline of artificial intelligence that attempts to develop computati... more Affective computing is a discipline of artificial intelligence that attempts to develop computational methods to recognize human emotions and generate synthetic emotions. Incorporating emotions into intelligent agents can be advantageous in different areas; emotions can make agents more credible so they can play a better role in various interactive systems to simulation level. Emotions would also play a functional role in complex systems, regulating interactions among agents. Different authors are trying to improve the interaction among intelligent agents in multiagent systems; an example is the MASOES affective model, which through the generation of emotions promotes different behaviors patterns. Although this affective model has already been formally verified at the design level, its verification is still a pending action. In this paper, we are presenting the MASOES affective model on a multiagent system, which has been verified in both, individual and collaborative mode.La comput...
In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, f... more In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, for which they must understand and manage a common language. The ontologies in an AmI are an ideal tool for this, making possible the communication between the intelligent objects that are part of the environment. These ontologies must be distributed, heterogeneous and dynamic, since they must adapt to the changes, needs and services of the AmI. This article proposes the implementation of a middleware that allows the ontological emergence, to manage all the knowledge that can be generated in an AmI. This middleware, called MiR-EO, is implemented as a reflective middleware, which manages its own ontological framework, made up of meta-ontologies that model the elements that must contain the ontologies of an AmI, andenables the ontological emergence process.
In the field of data mining and unsupervised machine learning, data clustering is defined as the ... more In the field of data mining and unsupervised machine learning, data clustering is defined as the task of grouping objects according to a similarity or dissimilarity measure. That means, objects that are similar among them are grouped in the same cluster, and objects that are dissimilar are grouped into different clusters so a data descriptive structure can emerge. In social sciences, the classification and the grouping regarding to behavior patterns can take place to quantitative descriptions and predictions which let more specific study about how societies work under some parameters such as prediction of a crime emergent behavior in some social sectors. In general, the clustering problem can be formulated as a multi-objective optimization problem, which can be very complex in time and space computationally speaking. In this sense, the Artificial Bee Colony Algorithm which is a swarm intelligence algorithm based on numeric optimization, tries to get the best solution to the problem,...
This work models Wikipedia and Free Software Development through a multiagent architecture for se... more This work models Wikipedia and Free Software Development through a multiagent architecture for self-organizing and emergent systems called MASOES without mathematically representing the system. In that sense, each component, mechanism and process of MASOES is instanced at individual and collective levels by the observed phenomena at the modeled systems. Thus, this paper proposes a methodology to show how to model real systems using MASOES, in order to study their self-organizing and emergent properties and, later on, to facilitate the verification of these properties, mechanisms, components and social interactions for promoting collaborative work and sharing individual and collective knowledge in these systems.Este trabajo modela el comportamiento de Wikipedia y el desarrollo de Software Libre, a través de una arquitectura multiagente para sistemas emergentes y auto-organizados llamada MASOES, sin especificar matemáticamente el sistema. En ese sentido, cada componente, mecanismo y p...
Affective computing is a discipline of artificial intelligence that attempts to develop computati... more Affective computing is a discipline of artificial intelligence that attempts to develop computational methods to recognize human emotions and generate synthetic emotions. Incorporating emotions into intelligent agents can be advantageous in different areas; emotions can make agents more credible so they can play a better role in various interactive systems to simulation level. Emotions would also play a functional role in complex systems, regulating interactions among agents. Different authors are trying to improve the interaction among intelligent agents in multiagent systems; an example is the MASOES affective model, which through the generation of emotions promotes different behaviors patterns. Although this affective model has already been formally verified at the design level, its verification is still a pending action. In this paper, we are presenting the MASOES affective model on a multiagent system, which has been verified in both, individual and collaborative mode.La comput...
Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the ... more Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the modeling of ontologies and the ontological integration processes also. When there are several ontologies of the same domain, it is possible, from a combination process, to obtain important inputs for the generation of meta-concepts. Moreover, category theory allows defining in a formal way, the structures and the set of data that have common properties. In this article, we apply the category theory, in particular, the definitions of categories and sub-categories, in the process of generating of meta-concepts, as a way for the formalization of the automatic construction of meta-ontologies. The category theory is applied together with a collective intelligence approach based on the Ant Colony Optimization algorithm, during the combination process of multiple ontologies, in order to automate the meta-ontology construction. RÉSUMÉ. Les méta-ontologies peuvent être utilisées pour définir une ...
In order to decrease the time of construction of self-assembly algorithm based on the constructio... more In order to decrease the time of construction of self-assembly algorithm based on the construction of wasp nests, with respect to the number of iterations, this work proposes a novel hybrid swarm algorithm combining strengths of self-assembly and the particle swarm optimization. The paper also considers integration of adaptive values of inertia to further balance exploration and exploitation for improving the construction process. According to the obtained results, it is shown experimentally with two types of benchmark structures that the convergence speed of our hybrid swarm algorithm is considerably improved because each structure is built complete on a lower number of iterations compared with the classical algorithm.
In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, f... more In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, for which they must understand and manage a common language. The ontologies in an AmI are an ideal tool for this, making possible the communication between the intelligent objects that are part of the environment. These ontologies must be distributed, heterogeneous and dynamic, since they must adapt to the changes, needs and services of the AmI. This article proposes the implementation of a middleware that allows the ontological emergence, to manage all the knowledge that can be generated in an AmI. This middleware, called MiR-EO, is implemented as a reflective middleware, which manages its own ontological framework, made up of meta-ontologies that model the elements that must contain the ontologies of an AmI, andenables the ontological emergence process.
En un Ambiente Inteligente (AmI), los dispositivos que participan deben intercambiar conocimiento... more En un Ambiente Inteligente (AmI), los dispositivos que participan deben intercambiar conocimiento permanentemente, para lo cual deben entenderse y manejar un lenguaje comun, para el logro de la interoperabilidad semantica. Las ontologias en un AmI constituyen una herramienta ideal para ello, posibilitando la comunicacion entre los objetos inteligentes que forman parte del ambiente. Estas ontologias deben ser distribuidas, heterogeneas y dinamicas ya que deben adaptarse a los cambios, necesidades y servicios del AmI. Este articulo propone la implementacion de un middleware que permite la emergencia ontologica, con el fin de gestionar todo el conocimiento que se puede generar en un AmI. El middleware, llamado MiR-EO, se implementa como un middleware reflexivo, que maneja su propio marco ontologico, conformado por meta-ontologias que modelan los elementos que deben contener las ontologias de un AmI, y posibilitan el proceso de emergencia ontologica.
espanolEl presente trabajo destaca algunos datos estad isticos para r ealizar una breve rese na h... more espanolEl presente trabajo destaca algunos datos estad isticos para r ealizar una breve rese na hist orica de la Maestr ia de Ciencias de la Computaci on del Decanato de Ciencias y Tecnolog ia de la Universidad Centroccidenta l “Li- sandro Alvarado” en sus menciones Inteligencia Artificial, Redes de C ompu- tadoras e Ingenier ia del Software. Adicionalmente expone las per spectivas acad emicas para superar las exigencias actuales a las que se enfre nta. EnglishThis review highlights some important statistical data for a brief hist orical review of the Master of Computer Science of the Faculty of Science and Technology at Universidad Centroccidental “Lisandro Alvarado” in their different specialties Artificial Intelligence, Computer Networks and Software engineering, and further, exposing what academic prospects are proposed to overcome the current demands that this master program face s.
In the field of data mining and unsupervised machine learning, data clustering is defined as the ... more In the field of data mining and unsupervised machine learning, data clustering is defined as the task of grouping objects according to a similarity or dissimilarity measure. That means, objects that are similar among them are grouped in the same cluster, and objects that are dissimilar are grouped into different clusters so a data descriptive structure can emerge. In social sciences, the classification and the grouping regarding to behavior patterns can take place to quantitative descriptions and predictions which let more specific study about how societies work under some parameters such as prediction of a crime emergent behavior in some social sectors. In general, the clustering problem can be formulated as a multi-objective optimization problem, which can be very complex in time and space computationally speaking. In this sense, the Artificial Bee Colony Algorithm which is a swarm intelligence algorithm based on numeric optimization, tries to get the best solution to the problem,...
Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the ... more Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the modeling of ontologies and the ontological integration processes also. When there are several ontologies of the same domain, it is possible, from a combination process, to obtain important inputs for the generation of meta-concepts. Moreover, category theory allows defining in a formal way, the structures and the set of data that have common properties. In this article, we apply the category theory, in particular, the definitions of categories and sub-categories, in the process of generating of meta-concepts, as a way for the formalization of the automatic construction of meta-ontologies. The category theory is applied together with a collective intelligence approach based on the Ant Colony Optimization algorithm, during the combination process of multiple ontologies, in order to automate the meta-ontology construction. RÉSUMÉ. Les méta-ontologies peuvent être utilisées pour définir une ...
In order to decrease the time of construction of self-assembly algorithm based on the constructio... more In order to decrease the time of construction of self-assembly algorithm based on the construction of wasp nests, with respect to the number of iterations, this work proposes a novel hybrid swarm algorithm combining strengths of self-assembly and the particle swarm optimization. The paper also considers integration of adaptive values of inertia to further balance exploration and exploitation for improving the construction process. According to the obtained results, it is shown experimentally with two types of benchmark structures that the convergence speed of our hybrid swarm algorithm is considerably improved because each structure is built complete on a lower number of iterations compared with the classical algorithm.
Human crowd motion is mainly driven by self-organized processes based on local interactions among... more Human crowd motion is mainly driven by self-organized processes based on local interactions among pedestrians. While most studies of crowd behaviour consider only interactions among isolated individuals, it turns out that up to 70% of people in a crowd are actually moving in groups, such as friends, couples, or families walking together. These groups constitute medium-scale aggregated structures and their impact on crowd dynamics is still largely unknown. In this work, we analyze the motion of approximately 1500 pedestrian groups under natural condition, and show that social interactions among group members generate typical group walking patterns that influence crowd dynamics. At low density, group members tend to walk side by side, forming a line perpendicular to the walking direction. As the density increases, however, the linear walking formation is bent forward, turning it into a V-like pattern. These spatial patterns can be well described by a model based on social communication between group members. We show that the V-like walking pattern facilitates social interactions within the group, but reduces the flow because of its ''non-aerodynamic'' shape. Therefore, when crowd density increases, the group organization results from a trade-off between walking faster and facilitating social exchange. These insights demonstrate that crowd dynamics is not only determined by physical constraints induced by other pedestrians and the environment, but also significantly by communicative, social interactions among individuals.
Affective computing is a discipline of artificial intelligence that attempts to develop computati... more Affective computing is a discipline of artificial intelligence that attempts to develop computational methods to recognize human emotions and generate synthetic emotions. Incorporating emotions into intelligent agents can be advantageous in different areas; emotions can make agents more credible so they can play a better role in various interactive systems to simulation level. Emotions would also play a functional role in complex systems, regulating interactions among agents. Different authors are trying to improve the interaction among intelligent agents in multiagent systems; an example is the MASOES affective model, which through the generation of emotions promotes different behaviors patterns. Although this affective model has already been formally verified at the design level, its verification is still a pending action. In this paper, we are presenting the MASOES affective model on a multiagent system, which has been verified in both, individual and collaborative mode.La comput...
In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, f... more In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, for which they must understand and manage a common language. The ontologies in an AmI are an ideal tool for this, making possible the communication between the intelligent objects that are part of the environment. These ontologies must be distributed, heterogeneous and dynamic, since they must adapt to the changes, needs and services of the AmI. This article proposes the implementation of a middleware that allows the ontological emergence, to manage all the knowledge that can be generated in an AmI. This middleware, called MiR-EO, is implemented as a reflective middleware, which manages its own ontological framework, made up of meta-ontologies that model the elements that must contain the ontologies of an AmI, andenables the ontological emergence process.
In the field of data mining and unsupervised machine learning, data clustering is defined as the ... more In the field of data mining and unsupervised machine learning, data clustering is defined as the task of grouping objects according to a similarity or dissimilarity measure. That means, objects that are similar among them are grouped in the same cluster, and objects that are dissimilar are grouped into different clusters so a data descriptive structure can emerge. In social sciences, the classification and the grouping regarding to behavior patterns can take place to quantitative descriptions and predictions which let more specific study about how societies work under some parameters such as prediction of a crime emergent behavior in some social sectors. In general, the clustering problem can be formulated as a multi-objective optimization problem, which can be very complex in time and space computationally speaking. In this sense, the Artificial Bee Colony Algorithm which is a swarm intelligence algorithm based on numeric optimization, tries to get the best solution to the problem,...
This work models Wikipedia and Free Software Development through a multiagent architecture for se... more This work models Wikipedia and Free Software Development through a multiagent architecture for self-organizing and emergent systems called MASOES without mathematically representing the system. In that sense, each component, mechanism and process of MASOES is instanced at individual and collective levels by the observed phenomena at the modeled systems. Thus, this paper proposes a methodology to show how to model real systems using MASOES, in order to study their self-organizing and emergent properties and, later on, to facilitate the verification of these properties, mechanisms, components and social interactions for promoting collaborative work and sharing individual and collective knowledge in these systems.Este trabajo modela el comportamiento de Wikipedia y el desarrollo de Software Libre, a través de una arquitectura multiagente para sistemas emergentes y auto-organizados llamada MASOES, sin especificar matemáticamente el sistema. En ese sentido, cada componente, mecanismo y p...
Affective computing is a discipline of artificial intelligence that attempts to develop computati... more Affective computing is a discipline of artificial intelligence that attempts to develop computational methods to recognize human emotions and generate synthetic emotions. Incorporating emotions into intelligent agents can be advantageous in different areas; emotions can make agents more credible so they can play a better role in various interactive systems to simulation level. Emotions would also play a functional role in complex systems, regulating interactions among agents. Different authors are trying to improve the interaction among intelligent agents in multiagent systems; an example is the MASOES affective model, which through the generation of emotions promotes different behaviors patterns. Although this affective model has already been formally verified at the design level, its verification is still a pending action. In this paper, we are presenting the MASOES affective model on a multiagent system, which has been verified in both, individual and collaborative mode.La comput...
Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the ... more Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the modeling of ontologies and the ontological integration processes also. When there are several ontologies of the same domain, it is possible, from a combination process, to obtain important inputs for the generation of meta-concepts. Moreover, category theory allows defining in a formal way, the structures and the set of data that have common properties. In this article, we apply the category theory, in particular, the definitions of categories and sub-categories, in the process of generating of meta-concepts, as a way for the formalization of the automatic construction of meta-ontologies. The category theory is applied together with a collective intelligence approach based on the Ant Colony Optimization algorithm, during the combination process of multiple ontologies, in order to automate the meta-ontology construction. RÉSUMÉ. Les méta-ontologies peuvent être utilisées pour définir une ...
In order to decrease the time of construction of self-assembly algorithm based on the constructio... more In order to decrease the time of construction of self-assembly algorithm based on the construction of wasp nests, with respect to the number of iterations, this work proposes a novel hybrid swarm algorithm combining strengths of self-assembly and the particle swarm optimization. The paper also considers integration of adaptive values of inertia to further balance exploration and exploitation for improving the construction process. According to the obtained results, it is shown experimentally with two types of benchmark structures that the convergence speed of our hybrid swarm algorithm is considerably improved because each structure is built complete on a lower number of iterations compared with the classical algorithm.
In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, f... more In a Smart Environment (AmI), the devices that participate must exchange knowledge permanently, for which they must understand and manage a common language. The ontologies in an AmI are an ideal tool for this, making possible the communication between the intelligent objects that are part of the environment. These ontologies must be distributed, heterogeneous and dynamic, since they must adapt to the changes, needs and services of the AmI. This article proposes the implementation of a middleware that allows the ontological emergence, to manage all the knowledge that can be generated in an AmI. This middleware, called MiR-EO, is implemented as a reflective middleware, which manages its own ontological framework, made up of meta-ontologies that model the elements that must contain the ontologies of an AmI, andenables the ontological emergence process.
En un Ambiente Inteligente (AmI), los dispositivos que participan deben intercambiar conocimiento... more En un Ambiente Inteligente (AmI), los dispositivos que participan deben intercambiar conocimiento permanentemente, para lo cual deben entenderse y manejar un lenguaje comun, para el logro de la interoperabilidad semantica. Las ontologias en un AmI constituyen una herramienta ideal para ello, posibilitando la comunicacion entre los objetos inteligentes que forman parte del ambiente. Estas ontologias deben ser distribuidas, heterogeneas y dinamicas ya que deben adaptarse a los cambios, necesidades y servicios del AmI. Este articulo propone la implementacion de un middleware que permite la emergencia ontologica, con el fin de gestionar todo el conocimiento que se puede generar en un AmI. El middleware, llamado MiR-EO, se implementa como un middleware reflexivo, que maneja su propio marco ontologico, conformado por meta-ontologias que modelan los elementos que deben contener las ontologias de un AmI, y posibilitan el proceso de emergencia ontologica.
espanolEl presente trabajo destaca algunos datos estad isticos para r ealizar una breve rese na h... more espanolEl presente trabajo destaca algunos datos estad isticos para r ealizar una breve rese na hist orica de la Maestr ia de Ciencias de la Computaci on del Decanato de Ciencias y Tecnolog ia de la Universidad Centroccidenta l “Li- sandro Alvarado” en sus menciones Inteligencia Artificial, Redes de C ompu- tadoras e Ingenier ia del Software. Adicionalmente expone las per spectivas acad emicas para superar las exigencias actuales a las que se enfre nta. EnglishThis review highlights some important statistical data for a brief hist orical review of the Master of Computer Science of the Faculty of Science and Technology at Universidad Centroccidental “Lisandro Alvarado” in their different specialties Artificial Intelligence, Computer Networks and Software engineering, and further, exposing what academic prospects are proposed to overcome the current demands that this master program face s.
In the field of data mining and unsupervised machine learning, data clustering is defined as the ... more In the field of data mining and unsupervised machine learning, data clustering is defined as the task of grouping objects according to a similarity or dissimilarity measure. That means, objects that are similar among them are grouped in the same cluster, and objects that are dissimilar are grouped into different clusters so a data descriptive structure can emerge. In social sciences, the classification and the grouping regarding to behavior patterns can take place to quantitative descriptions and predictions which let more specific study about how societies work under some parameters such as prediction of a crime emergent behavior in some social sectors. In general, the clustering problem can be formulated as a multi-objective optimization problem, which can be very complex in time and space computationally speaking. In this sense, the Artificial Bee Colony Algorithm which is a swarm intelligence algorithm based on numeric optimization, tries to get the best solution to the problem,...
Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the ... more Meta-ontologies can be used to define a generic form of meta-concepts, which can be used for the modeling of ontologies and the ontological integration processes also. When there are several ontologies of the same domain, it is possible, from a combination process, to obtain important inputs for the generation of meta-concepts. Moreover, category theory allows defining in a formal way, the structures and the set of data that have common properties. In this article, we apply the category theory, in particular, the definitions of categories and sub-categories, in the process of generating of meta-concepts, as a way for the formalization of the automatic construction of meta-ontologies. The category theory is applied together with a collective intelligence approach based on the Ant Colony Optimization algorithm, during the combination process of multiple ontologies, in order to automate the meta-ontology construction. RÉSUMÉ. Les méta-ontologies peuvent être utilisées pour définir une ...
In order to decrease the time of construction of self-assembly algorithm based on the constructio... more In order to decrease the time of construction of self-assembly algorithm based on the construction of wasp nests, with respect to the number of iterations, this work proposes a novel hybrid swarm algorithm combining strengths of self-assembly and the particle swarm optimization. The paper also considers integration of adaptive values of inertia to further balance exploration and exploitation for improving the construction process. According to the obtained results, it is shown experimentally with two types of benchmark structures that the convergence speed of our hybrid swarm algorithm is considerably improved because each structure is built complete on a lower number of iterations compared with the classical algorithm.
Uploads
Papers by Niriaska Perozo