http://sedici.unlp.edu.ar:80/handle/10915/123572 Thu, 12 Nov 2026 03:41:36 GMT 2026-11-12T03:41:36Z http://sedici.unlp.edu.ar:80/handle/10915/123746 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) In the past, the Fragment Assembly Problem has been solved efficiently by many metaheuristics. In this work, we propose a new one, called SAX, which consists in combining two metaheuristics: a trajectory method as Simulated Annealing and a population-based method as Genetic Algorithm. We also analyze the relative advantages of this hybridization against other assemblers from literature. From this analysis, we conclude that SAX improves the quality results found by other metaheuristic and non-metaheuristic assemblers for solving 100% of the largest instances for this problem. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123746 2012-01-01T00:00:00Z In the past, the Fragment Assembly Problem has been solved efficiently by many metaheuristics. In this work, we propose a new one, called SAX, which consists in combining two metaheuristics: a trajectory method as Simulated Annealing and a population-based method as Genetic Algorithm. We also analyze the relative advantages of this hybridization against other assemblers from literature. From this analysis, we conclude that SAX improves the quality results found by other metaheuristic and non-metaheuristic assemblers for solving 100% of the largest instances for this problem. http://sedici.unlp.edu.ar:80/handle/10915/123744 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) This article presents an efficient evolutionary algorithm applied to the deadline scheduling in project management, a NP-hard problem with major relevance in software engineering and scheduling activities. The evolutionary algorithm has been specifically designed to provide accurate and efficient solutions, by using operators that allow realistic problem instances to be solved. Efficient numerical results are reported in the experimental analysis performed on standard problem instances. The experimental results demonstrate that the proposed evolutionary algorithm is able to outperform one of the best well-known deterministic techniques for the problem in reduced execution times, specially on highly complex instances. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123744 2012-01-01T00:00:00Z This article presents an efficient evolutionary algorithm applied to the deadline scheduling in project management, a NP-hard problem with major relevance in software engineering and scheduling activities. The evolutionary algorithm has been specifically designed to provide accurate and efficient solutions, by using operators that allow realistic problem instances to be solved. Efficient numerical results are reported in the experimental analysis performed on standard problem instances. The experimental results demonstrate that the proposed evolutionary algorithm is able to outperform one of the best well-known deterministic techniques for the problem in reduced execution times, specially on highly complex instances. http://sedici.unlp.edu.ar:80/handle/10915/123741 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) When large models are used for a classification task, model compression is necessary because there are transmission, space, time or computing constraints that have to be fulfilled. Multilayer Perceptron (MLP) models are traditionally used as a classifier, but depending on the problem, they may need a large number of parameters (neuron functions, weights and bias) to obtain an acceptable performance. This work extends the evaluation of a technique to compress an array of MLPs, through the outputs of a Volterra-Neural Network (Volterra-NN), maintaining its classification performance. The obtained results show that these outputs can be used to build an array of (Volterra-NN) that needs significantly less parameters than the original array of MLPs, furthermore having the same high accuracy in most of the cases. The Volterra-NN compression capabilities have been tested by solving several kind of classification problems. Experimental results are presented on three well-known databases: Letter Recognition, Pen-Based Recognition of Handwritten Digits, and Face recognition databases. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123741 2012-01-01T00:00:00Z When large models are used for a classification task, model compression is necessary because there are transmission, space, time or computing constraints that have to be fulfilled. Multilayer Perceptron (MLP) models are traditionally used as a classifier, but depending on the problem, they may need a large number of parameters (neuron functions, weights and bias) to obtain an acceptable performance. This work extends the evaluation of a technique to compress an array of MLPs, through the outputs of a Volterra-Neural Network (Volterra-NN), maintaining its classification performance. The obtained results show that these outputs can be used to build an array of (Volterra-NN) that needs significantly less parameters than the original array of MLPs, furthermore having the same high accuracy in most of the cases. The Volterra-NN compression capabilities have been tested by solving several kind of classification problems. Experimental results are presented on three well-known databases: Letter Recognition, Pen-Based Recognition of Handwritten Digits, and Face recognition databases. http://sedici.unlp.edu.ar:80/handle/10915/123740 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Most of work on reasoning and decision-making in virtual agents relates the choices with a exhaustive exploration, analyzing every possible alternative and implication, and trying to maximize some utility measure in order to make the best decision. Humans, however, seems no reason and make decisions naturally in this way. As authors such as Herbert A. Simon have proposed, humans seems to develop a concept of bounded rationality, according to which human reasoning process and decision-making is bounded to a part of reality at a time as a focussing effect. According some psychologists, focus of thought is one of the main purposes of emotions in humans. Using an abstract framework, in this work we propose an approach to consider emotions as an argument-selection heuristic towards the ability for an agent to reason and act in a believable manner. Influenced by emotions, the agent will produce a line of reasoning according to the evolution of its own emotional state. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123740 2012-01-01T00:00:00Z Most of work on reasoning and decision-making in virtual agents relates the choices with a exhaustive exploration, analyzing every possible alternative and implication, and trying to maximize some utility measure in order to make the best decision. Humans, however, seems no reason and make decisions naturally in this way. As authors such as Herbert A. Simon have proposed, humans seems to develop a concept of bounded rationality, according to which human reasoning process and decision-making is bounded to a part of reality at a time as a focussing effect. According some psychologists, focus of thought is one of the main purposes of emotions in humans. Using an abstract framework, in this work we propose an approach to consider emotions as an argument-selection heuristic towards the ability for an agent to reason and act in a believable manner. Influenced by emotions, the agent will produce a line of reasoning according to the evolution of its own emotional state. http://sedici.unlp.edu.ar:80/handle/10915/123738 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) We consider the problem of belief revision in a multi-agent system with information stemming from different agents with different degrees of credibility. In this context an agent has to carefully choose which information is to be accepted for revision in order to avoid believing in faulty and untrustworthy information. We propose a revision process combining selective revision, deductive argumentation, and credibility information for the adequate handling of information in this complex scenario. New information is evaluated based on the credibility of the source in combination with all arguments favoring and opposing the new information. The evaluation process determines which part of the new information is to be accepted for revision and thereupon incorporated into the belief base by an appropriate revision operator. We demonstrate the benefits of our approach, investigate formal properties, and show that it outperforms the baseline approach without argumentation. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123738 2012-01-01T00:00:00Z We consider the problem of belief revision in a multi-agent system with information stemming from different agents with different degrees of credibility. In this context an agent has to carefully choose which information is to be accepted for revision in order to avoid believing in faulty and untrustworthy information. We propose a revision process combining selective revision, deductive argumentation, and credibility information for the adequate handling of information in this complex scenario. New information is evaluated based on the credibility of the source in combination with all arguments favoring and opposing the new information. The evaluation process determines which part of the new information is to be accepted for revision and thereupon incorporated into the belief base by an appropriate revision operator. We demonstrate the benefits of our approach, investigate formal properties, and show that it outperforms the baseline approach without argumentation. http://sedici.unlp.edu.ar:80/handle/10915/123735 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Nowadays, there is a wide range of domains in which there is a need to generate recommendations to groups of users instead of individuals. Several techniques for aggregating individual models or ratings have been developed; a disadvantage of these techniques is that they require a large amount of computations to estimate unknown ratings. In this article, we present an analysis of the impact of estimating ratings when an aggregation technique is used. For that purpose, we describe a hybrid approach to generate group recommendations based on group modeling. We also present the results obtained when evaluating the approach and two well-known aggregation techniques in the movie domain, and the variations of those results when the estimation process is not included. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123735 2012-01-01T00:00:00Z Nowadays, there is a wide range of domains in which there is a need to generate recommendations to groups of users instead of individuals. Several techniques for aggregating individual models or ratings have been developed; a disadvantage of these techniques is that they require a large amount of computations to estimate unknown ratings. In this article, we present an analysis of the impact of estimating ratings when an aggregation technique is used. For that purpose, we describe a hybrid approach to generate group recommendations based on group modeling. We also present the results obtained when evaluating the approach and two well-known aggregation techniques in the movie domain, and the variations of those results when the estimation process is not included. http://sedici.unlp.edu.ar:80/handle/10915/123733 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Los estilos de aprendizaje permiten clasificar a los estudiantes en un conjunto de escalas de acuerdo a las formas en que reciben y procesan la información. El conocer los estilos de aprendizaje de los estudiantes permite personalizar/adaptar los ambientes educativos a dichos estilos mejorando así la calidad de la educación. Existen varios enfoques para la detección automática de estilos de aprendizaje. Sin embargo, estos enfoques poseen varias limitaciones que dificultan su aplicación. En este contexto, proponemos un nuevo enfoque para la detección de estilos de aprendizaje de los estudiantes analizando sus interacciones con juegos digitales. En particular, este trabajo se focaliza en la detección del estilo de percepción propuesto por el modelo de Felder. Dicha detección se realiza observando cómo los estudiantes interactúan con un juego de in genio analizando la información extraída de dicha interacción. Para ello, se define y entrena una red de Bayes con el objetivo de inferir el estilo de percepción del estudiante usando la información extraída. Los resultados experimentales demuestran que es osible detectar el estilo de percepción de un estudiante utilizando juegos digitales, mejorando no sólo la precisión de la detección sino también resolviendo otras limitaciones de los enfoques existentes. Finalmente, concluimos que los juegos digitales son un ambiente prometedor con el cual es posible detectar el estilo de percepción de un estudiante. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123733 2012-01-01T00:00:00Z Los estilos de aprendizaje permiten clasificar a los estudiantes en un conjunto de escalas de acuerdo a las formas en que reciben y procesan la información. El conocer los estilos de aprendizaje de los estudiantes permite personalizar/adaptar los ambientes educativos a dichos estilos mejorando así la calidad de la educación. Existen varios enfoques para la detección automática de estilos de aprendizaje. Sin embargo, estos enfoques poseen varias limitaciones que dificultan su aplicación. En este contexto, proponemos un nuevo enfoque para la detección de estilos de aprendizaje de los estudiantes analizando sus interacciones con juegos digitales. En particular, este trabajo se focaliza en la detección del estilo de percepción propuesto por el modelo de Felder. Dicha detección se realiza observando cómo los estudiantes interactúan con un juego de in genio analizando la información extraída de dicha interacción. Para ello, se define y entrena una red de Bayes con el objetivo de inferir el estilo de percepción del estudiante usando la información extraída. Los resultados experimentales demuestran que es osible detectar el estilo de percepción de un estudiante utilizando juegos digitales, mejorando no sólo la precisión de la detección sino también resolviendo otras limitaciones de los enfoques existentes. Finalmente, concluimos que los juegos digitales son un ambiente prometedor con el cual es posible detectar el estilo de percepción de un estudiante. http://sedici.unlp.edu.ar:80/handle/10915/123731 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) In this work we introduce a low cost machine vision system for grading problems in agriculture. Instead of a careful evaluation of a given quantity over a reduced number of samples with a high cost dedicated equipment, we propose to measure the quantity with less precision but over a much bigger number of samples. The advantage of our procedure is that very low cost vision equipment can be used in this case. For example, we used a standard flatbed scanner as an integrated illumination plus acquisition hardware. Our system is aimed at the quantification of the amount of chlorophyll present in a production batch of soybean seeds. To this end we arbitrarily divided green seeds in four classes, with a decreasing amount of green pigment in each class. In particular, in this work we evaluate the possibility of an accurate discrimination among the four classes of green seeds using machine vision methods. We show that morphological features have low discrimination capabilities, and that a set of simple features measured over color distributions provides good separation among grades. Also, most errors are assignations to neighbor grades, which have a lower cost in grading. The good results are almost independent from the classifier being use, Random forest or Support Vector Machines with a Gaussian kernel in our case. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123731 2012-01-01T00:00:00Z In this work we introduce a low cost machine vision system for grading problems in agriculture. Instead of a careful evaluation of a given quantity over a reduced number of samples with a high cost dedicated equipment, we propose to measure the quantity with less precision but over a much bigger number of samples. The advantage of our procedure is that very low cost vision equipment can be used in this case. For example, we used a standard flatbed scanner as an integrated illumination plus acquisition hardware. Our system is aimed at the quantification of the amount of chlorophyll present in a production batch of soybean seeds. To this end we arbitrarily divided green seeds in four classes, with a decreasing amount of green pigment in each class. In particular, in this work we evaluate the possibility of an accurate discrimination among the four classes of green seeds using machine vision methods. We show that morphological features have low discrimination capabilities, and that a set of simple features measured over color distributions provides good separation among grades. Also, most errors are assignations to neighbor grades, which have a lower cost in grading. The good results are almost independent from the classifier being use, Random forest or Support Vector Machines with a Gaussian kernel in our case. http://sedici.unlp.edu.ar:80/handle/10915/123728 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) In this paper a framework is presented to automatically extract a sequence of images of the background of a scene from a shaky film. That is, the input video sequence may have local and global motion but the output video must contain exclusively the static background scene. Applying robust procedures to this end is one of the main goals of this work, since the aim is to get a procedure not only resistant to low scale noise but to occasional high scale noise. The median is used as an estimate of the background, the median absolute deviation (MAD) is used to establish a threshold to locate foreground and M-estimation for regression is used to stabilize the video sequence. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123728 2012-01-01T00:00:00Z In this paper a framework is presented to automatically extract a sequence of images of the background of a scene from a shaky film. That is, the input video sequence may have local and global motion but the output video must contain exclusively the static background scene. Applying robust procedures to this end is one of the main goals of this work, since the aim is to get a procedure not only resistant to low scale noise but to occasional high scale noise. The median is used as an estimate of the background, the median absolute deviation (MAD) is used to establish a threshold to locate foreground and M-estimation for regression is used to stabilize the video sequence. http://sedici.unlp.edu.ar:80/handle/10915/123727 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) El presente trabajo se enmarca en el problema de interpolación espacial de variables vitícolas por medio de Support Vector Machines (SVM) y Local SVM (LSVM), a partir de mediciones geo-espaciales en viñedos de distintas bodegas de la provincia de Mendoza. Los resultados de estas técnicas son comparados con los dos métodos de interpolación de uso más extendido empleadas en Viticultura de Precisión para el modelado de datos agrícolas: Inverse Distance Weighting (IDW) y Kriging. Los resultados obtenidos en este trabajo muestran una mejora en la calidad de las interpolaciones, condicionada en la cantidad de datos disponibles, donde LSVM y SVM en general obtienen resultados de mayor calidad. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123727 2012-01-01T00:00:00Z El presente trabajo se enmarca en el problema de interpolación espacial de variables vitícolas por medio de Support Vector Machines (SVM) y Local SVM (LSVM), a partir de mediciones geo-espaciales en viñedos de distintas bodegas de la provincia de Mendoza. Los resultados de estas técnicas son comparados con los dos métodos de interpolación de uso más extendido empleadas en Viticultura de Precisión para el modelado de datos agrícolas: Inverse Distance Weighting (IDW) y Kriging. Los resultados obtenidos en este trabajo muestran una mejora en la calidad de las interpolaciones, condicionada en la cantidad de datos disponibles, donde LSVM y SVM en general obtienen resultados de mayor calidad. http://sedici.unlp.edu.ar:80/handle/10915/123726 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Generating and representing knowledge about heuristics for repair-based scheduling is a key issue in any rescheduling strategy to deal with unforeseen events and disturbances. Resorting to a feature-based representation of schedule states is very inefficient and generalization to unseen states is highly unreliable whereas the acquired knowledge is difficult to transfer to similar scheduling domains. In contrast, first-order relational representations enable the exploitation of the existence of domain objects and relations over these objects, and enable the use of quantification over objectives (goals), action effects and properties of states. In this work, a novel approach which formalizes the rescheduling problem as a Relational Markov Decision Process integrating first-order (deictic) representations of (abstract) schedule states is presented. The proposed approach is implemented in a real-time rescheduling prototype, allowing an interactive scheduling strategy that may handle different repair goals and disruption scenarios. The industrial case study vividly shows how relational abstractions provide compact repair policies with less computational efforts. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123726 2012-01-01T00:00:00Z Generating and representing knowledge about heuristics for repair-based scheduling is a key issue in any rescheduling strategy to deal with unforeseen events and disturbances. Resorting to a feature-based representation of schedule states is very inefficient and generalization to unseen states is highly unreliable whereas the acquired knowledge is difficult to transfer to similar scheduling domains. In contrast, first-order relational representations enable the exploitation of the existence of domain objects and relations over these objects, and enable the use of quantification over objectives (goals), action effects and properties of states. In this work, a novel approach which formalizes the rescheduling problem as a Relational Markov Decision Process integrating first-order (deictic) representations of (abstract) schedule states is presented. The proposed approach is implemented in a real-time rescheduling prototype, allowing an interactive scheduling strategy that may handle different repair goals and disruption scenarios. The industrial case study vividly shows how relational abstractions provide compact repair policies with less computational efforts. http://sedici.unlp.edu.ar:80/handle/10915/123721 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Linear feature extraction is commonly applied in an all-at-once way, meaning that a single trasformation is used for all the data regardless of the classes. Very good results can be achieved with this approach when the classification problem involves just a few classes. Nevertheless, when the number of classes grows is often difficult to find a low dimensional subspace while preserving the error rates, due to overlapping between the different populations. In this paper we propose an alternative method based on a collection of transformations, each involving two of the classes in the problem. Each transformation in the collection is estimated using an approximation to the information discriminant analysis, which is found to be equivalent to sufficient dimension reduction for heteroscedastic Gaussian data. A regularized version of the objective function is also introduced, allowing for simultaneous variable selection. In this way, each reduction implies only a subset of the original variables. A probabilistic model is build by means of a simple latent variable, so that classification is carried out using standard Bayes decision rule. Several real data sets are used to compare the performance of the proposed method against similar approaches based on ensembles of binary classifiers. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123721 2012-01-01T00:00:00Z Linear feature extraction is commonly applied in an all-at-once way, meaning that a single trasformation is used for all the data regardless of the classes. Very good results can be achieved with this approach when the classification problem involves just a few classes. Nevertheless, when the number of classes grows is often difficult to find a low dimensional subspace while preserving the error rates, due to overlapping between the different populations. In this paper we propose an alternative method based on a collection of transformations, each involving two of the classes in the problem. Each transformation in the collection is estimated using an approximation to the information discriminant analysis, which is found to be equivalent to sufficient dimension reduction for heteroscedastic Gaussian data. A regularized version of the objective function is also introduced, allowing for simultaneous variable selection. In this way, each reduction implies only a subset of the original variables. A probabilistic model is build by means of a simple latent variable, so that classification is carried out using standard Bayes decision rule. Several real data sets are used to compare the performance of the proposed method against similar approaches based on ensembles of binary classifiers. http://sedici.unlp.edu.ar:80/handle/10915/123719 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Social annotations voluntarily provided by users in tagging or book-marking sites such as Delicious or Flickr have been recognized as an interesting source of metadata for assisting tasks such as classification of Web resources. However, the open-ended nature of the tags employed to annotate resources leads to problems such as the introduction of noise and ambiguity that may hinder clas- sification results. This paper presents an approach for semantically analyse social annotations in order to attain enriched, concept-based representations of Web resources. Experimental results showed that the strategies proposed to relate tags to conceptual entities allow to improve the results of resource classification. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123719 2012-01-01T00:00:00Z Social annotations voluntarily provided by users in tagging or book-marking sites such as Delicious or Flickr have been recognized as an interesting source of metadata for assisting tasks such as classification of Web resources. However, the open-ended nature of the tags employed to annotate resources leads to problems such as the introduction of noise and ambiguity that may hinder clas- sification results. This paper presents an approach for semantically analyse social annotations in order to attain enriched, concept-based representations of Web resources. Experimental results showed that the strategies proposed to relate tags to conceptual entities allow to improve the results of resource classification. http://sedici.unlp.edu.ar:80/handle/10915/123600 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Wikipedia es una gran enciclopedia en la que colaboran miles de usuarios alrededor de todo el mundo. Por otro lado se encuentra DBpedia, un repositorio semántico que extrae información de Wikipedia. La información semántica extraída por DBpedia facilita la búsqueda de información, la cual a veces no es posible obtener navegando por Wikipedia. Sin embargo, la información que contiene DBpedia no representa la totalidad del contenido de Wikipedia. En este artículo se verá cuál puede ser la razón de esta falta de datos, y cómo se podría evitar este problema en una búsqueda, implementando un repositorio federado. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123600 2012-01-01T00:00:00Z Wikipedia es una gran enciclopedia en la que colaboran miles de usuarios alrededor de todo el mundo. Por otro lado se encuentra DBpedia, un repositorio semántico que extrae información de Wikipedia. La información semántica extraída por DBpedia facilita la búsqueda de información, la cual a veces no es posible obtener navegando por Wikipedia. Sin embargo, la información que contiene DBpedia no representa la totalidad del contenido de Wikipedia. En este artículo se verá cuál puede ser la razón de esta falta de datos, y cómo se podría evitar este problema en una búsqueda, implementando un repositorio federado. http://sedici.unlp.edu.ar:80/handle/10915/123598 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) Social networks have grown exponentially in use and impact on the society as a whole. In particular, microblogging platforms such as Twitter have become important tools to assess public opinion on different issues. Recently, some approaches for assessing Twitter messages have been developed. However, such approaches have an important lim- itation, as they do not take into account contradictory and potentially inconsistent information which might emerge from relevant messages. We contend that the information made available in Twitter can be useful for modelling arguments which emerge bottom-up from the social interaction associated with such messages, thus enabling an integration between Twitter and defeasible argumentation. In this paper, we outline the main elements characterizing this integration, identifying “opinions” associated with particular hashtags, obtaining as well other alternative counter-opinions. As a result, we will be able to obtain an “opinion tree”, rooted in the first original query, in a similar way as done with dialectical trees in argumentation. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123598 2012-01-01T00:00:00Z Social networks have grown exponentially in use and impact on the society as a whole. In particular, microblogging platforms such as Twitter have become important tools to assess public opinion on different issues. Recently, some approaches for assessing Twitter messages have been developed. However, such approaches have an important lim- itation, as they do not take into account contradictory and potentially inconsistent information which might emerge from relevant messages. We contend that the information made available in Twitter can be useful for modelling arguments which emerge bottom-up from the social interaction associated with such messages, thus enabling an integration between Twitter and defeasible argumentation. In this paper, we outline the main elements characterizing this integration, identifying “opinions” associated with particular hashtags, obtaining as well other alternative counter-opinions. As a result, we will be able to obtain an “opinion tree”, rooted in the first original query, in a similar way as done with dialectical trees in argumentation. http://sedici.unlp.edu.ar:80/handle/10915/123596 Objeto de conferencia XIII Argentine Symposium on Artificial Intelligence (ASAI 2012) (XLI JAIIO, La Plata, 27 y 28 de agosto de 2012) The Argument Interchange Format (AIF) is a communal project with the purpose of developing a way of interchanging data between tools for argument manipulation and visualization. The AIF project also aims to develop a commonly agreed upon core ontology that specifies the basic concepts used to express arguments and their mutual relations. However, the flexibility provided by the AIF core ontology may lead to ambiguous or undesired interpretations. If ambiguous and anomalous situations are allowed, the purpose of using AIF as a common lingua for the research and development of argumentation systems might be jeopardized. The goal of this work is to identify anomalies that can arise and propose solutions for them. Sun, 01 Jan 2012 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/123596 2012-01-01T00:00:00Z The Argument Interchange Format (AIF) is a communal project with the purpose of developing a way of interchanging data between tools for argument manipulation and visualization. The AIF project also aims to develop a commonly agreed upon core ontology that specifies the basic concepts used to express arguments and their mutual relations. However, the flexibility provided by the AIF core ontology may lead to ambiguous or undesired interpretations. If ambiguous and anomalous situations are allowed, the purpose of using AIF as a common lingua for the research and development of argumentation systems might be jeopardized. The goal of this work is to identify anomalies that can arise and propose solutions for them.