http://sedici.unlp.edu.ar:80/handle/10915/51950 Tue, 10 Nov 2026 14:40:40 GMT 2026-11-10T14:40:40Z http://sedici.unlp.edu.ar:80/handle/10915/51988 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) En la actividad agrícola existen muchos factores de riesgo que condicionan los resultados donde algunos pueden ser tratados por el hombre, como por ejemplo enfermedades o plagas pero también existen otros que no pueden ser controlados y que de suceder ponen en peligro la seguridad de las personas y producen grandes trastornos económicos. Este conjunto de factores de riesgo tienen que ver con el comportamiento de la atmósfera y los eventos extremos que surgen de ella. En particular, la ocurrencia de granizo constituye uno de los fenómenos meteorológicos capaces de infligir danos cuantiosos al deteriorar seriamente cultivos, edificios, medios de transporte, entre otros. En este trabajo se somete a discusión parte del trabajo de una tesis de maestría en minería de datos y descubrimiento de conocimiento donde el objetivo es evaluar cual es la capacidad de predicción de ocurrencia de granizo de un conjunto de índices de inestabilidad utilizando técnicas de aprendizaje automático en un entorno de Big Data. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51988 2015-01-01T00:00:00Z En la actividad agrícola existen muchos factores de riesgo que condicionan los resultados donde algunos pueden ser tratados por el hombre, como por ejemplo enfermedades o plagas pero también existen otros que no pueden ser controlados y que de suceder ponen en peligro la seguridad de las personas y producen grandes trastornos económicos. Este conjunto de factores de riesgo tienen que ver con el comportamiento de la atmósfera y los eventos extremos que surgen de ella. En particular, la ocurrencia de granizo constituye uno de los fenómenos meteorológicos capaces de infligir danos cuantiosos al deteriorar seriamente cultivos, edificios, medios de transporte, entre otros. En este trabajo se somete a discusión parte del trabajo de una tesis de maestría en minería de datos y descubrimiento de conocimiento donde el objetivo es evaluar cual es la capacidad de predicción de ocurrencia de granizo de un conjunto de índices de inestabilidad utilizando técnicas de aprendizaje automático en un entorno de Big Data. http://sedici.unlp.edu.ar:80/handle/10915/51986 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) Este proyecto implementa un prototipo que realiza la medición y visualización de temperatura en tiempo real utilizando tecnologías Open Hardware y Microsoft Azure. Se utilizan distintos tipos de sensores junto a un microcontrolador para capturar datos y almacenar la información en la nube, para luego analizar y visualizar los flujos de datos resultantes en tiempo real. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51986 2015-01-01T00:00:00Z Este proyecto implementa un prototipo que realiza la medición y visualización de temperatura en tiempo real utilizando tecnologías Open Hardware y Microsoft Azure. Se utilizan distintos tipos de sensores junto a un microcontrolador para capturar datos y almacenar la información en la nube, para luego analizar y visualizar los flujos de datos resultantes en tiempo real. http://sedici.unlp.edu.ar:80/handle/10915/51984 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) Mining big data involves several problems and new challenges, in addition to the huge volume of information. One the one hand, these data generally come from autonomous and decentralized sources, thus its dimensionality is heterogeneous and diverse, and generally involves privacy issues. On the other hand, algorithms for mining data such as clustering methods, have particular characteristics that make them useful for different types of data mining problems. Due to the huge amount of information, the task of choosing a single clustering approach becomes even more difficult. For instance, k-means, a very popular algorithm, always assumes spherical clusters in data; hierarchical approaches can be used when there is interest in finding this type of structure; expectationmaximization iteratively adjusts the parameters of a statistical model to fit the observed data. Moreover, all these methods work properly only with relatively small data sets. Large-volume data often make their application unfeasible, not to mention if data come from autonomous sources that are constantly growing and evolving. In the last years, a new clustering approach has emerged, called consensus clustering or cluster ensembles. Instead of running a single algorithm, this approach produces, at first, a set of data partitions (ensemble) by employing different clustering techniques on the same original data set. Then, this ensemble is processed by a consensus function, which produces a single consensus partition that outperforms individual solutions in the input ensemble. This approach has been successfully employed for distributed data mining, what makes it very interesting and applicable in the big data context. Although many techniques have been proposed for large data sets, most of them mainly focus on making individual components more efficient, instead of improving the whole consensus approach for the case of big data. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51984 2015-01-01T00:00:00Z Mining big data involves several problems and new challenges, in addition to the huge volume of information. One the one hand, these data generally come from autonomous and decentralized sources, thus its dimensionality is heterogeneous and diverse, and generally involves privacy issues. On the other hand, algorithms for mining data such as clustering methods, have particular characteristics that make them useful for different types of data mining problems. Due to the huge amount of information, the task of choosing a single clustering approach becomes even more difficult. For instance, k-means, a very popular algorithm, always assumes spherical clusters in data; hierarchical approaches can be used when there is interest in finding this type of structure; expectationmaximization iteratively adjusts the parameters of a statistical model to fit the observed data. Moreover, all these methods work properly only with relatively small data sets. Large-volume data often make their application unfeasible, not to mention if data come from autonomous sources that are constantly growing and evolving. In the last years, a new clustering approach has emerged, called consensus clustering or cluster ensembles. Instead of running a single algorithm, this approach produces, at first, a set of data partitions (ensemble) by employing different clustering techniques on the same original data set. Then, this ensemble is processed by a consensus function, which produces a single consensus partition that outperforms individual solutions in the input ensemble. This approach has been successfully employed for distributed data mining, what makes it very interesting and applicable in the big data context. Although many techniques have been proposed for large data sets, most of them mainly focus on making individual components more efficient, instead of improving the whole consensus approach for the case of big data. http://sedici.unlp.edu.ar:80/handle/10915/51982 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) Se realiza un estudio comparativo entre bases de datos de grafos con capacidades espaciales, evaluando tiempos de procesamiento, poder expresivo, usabilidad y necesidades de almacenamiento Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51982 2015-01-01T00:00:00Z Se realiza un estudio comparativo entre bases de datos de grafos con capacidades espaciales, evaluando tiempos de procesamiento, poder expresivo, usabilidad y necesidades de almacenamiento http://sedici.unlp.edu.ar:80/handle/10915/51979 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) The large-scale study of human mobility has been significantly enhanced over the last decade by the massive use of mobile phones in urban populations. Studying the activity of mobile phones allows us, not only to infer social networks between individuals, but also to observe the movements of these individuals in space and time. In this work, we investigate how these two related sources of information can be integrated within the context of detecting and analyzing large social events. We show that large social events can be characterized not only by an anomalous increase in activity of the antennas in the neighborhood of the event, but also by an increase in social relationships of the attendants present in the event. Moreover, having detected a large social event via increased antenna activity, we can use the network connections to infer whether an unobserved user was present at the event. More precisely, we address the following three challenges: (i) automatically detecting large social events via increased antenna activity; (ii) characterizing the social cohesion of the detected event; and (iii) analyzing the feasibility of inferring whether unobserved users were in the event. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51979 2015-01-01T00:00:00Z The large-scale study of human mobility has been significantly enhanced over the last decade by the massive use of mobile phones in urban populations. Studying the activity of mobile phones allows us, not only to infer social networks between individuals, but also to observe the movements of these individuals in space and time. In this work, we investigate how these two related sources of information can be integrated within the context of detecting and analyzing large social events. We show that large social events can be characterized not only by an anomalous increase in activity of the antennas in the neighborhood of the event, but also by an increase in social relationships of the attendants present in the event. Moreover, having detected a large social event via increased antenna activity, we can use the network connections to infer whether an unobserved user was present at the event. More precisely, we address the following three challenges: (i) automatically detecting large social events via increased antenna activity; (ii) characterizing the social cohesion of the detected event; and (iii) analyzing the feasibility of inferring whether unobserved users were in the event. http://sedici.unlp.edu.ar:80/handle/10915/51977 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) Big Data concerns large-volume, complex and growing data sets, with multiple and autonomous sources. It is now rapidly expanding in all science and engineering domains. Time series represent an important class of big data that can be obtained from several applications, such as medicine (electrocardiogram), environmental (daily temperature), financial (weekly sales totals, and prices of mutual funds and stocks), as well as from many areas, such as socialnetworks and biology. Bioinformatics seeks to provide tools and analyses that facilitate understanding of living systems, by analyzing and correlating biological information. In particular, as increasingly large amounts of genes information have become available in the last years, more efficient algorithms for dealing with such big data in genomics are required. There is an increasing interest in this field for the discovery of the network of regulations among a group of genes, named Gene Regulation Networks (GRN), by analyzing the genes expression profiles represented as timeseries. In it has been proposed the GRNNminer method, which allows discovering the subyacent GRN among a group of genes, through the proper modeling of the temporal dynamics of the gene expression profiles with artificial neural networks. However, it implies building and training a pool of neural models for each possible gentogen relationship, which derives in executing a very large set of experiments with O( n 2 ) order, where n is the total of involved genes. This work presents a proposal for dramatically reducing such experiments number to O( (n/k)2 ) when big timeseries is involved for reconstructing a GRN from such data, by previously clustering genes profiles in k groups using selforganizing maps (SOM). This way, the GRNNminer can be applied over smaller sets of timeseries, only those appearing in the same cluster. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51977 2015-01-01T00:00:00Z Big Data concerns large-volume, complex and growing data sets, with multiple and autonomous sources. It is now rapidly expanding in all science and engineering domains. Time series represent an important class of big data that can be obtained from several applications, such as medicine (electrocardiogram), environmental (daily temperature), financial (weekly sales totals, and prices of mutual funds and stocks), as well as from many areas, such as socialnetworks and biology. Bioinformatics seeks to provide tools and analyses that facilitate understanding of living systems, by analyzing and correlating biological information. In particular, as increasingly large amounts of genes information have become available in the last years, more efficient algorithms for dealing with such big data in genomics are required. There is an increasing interest in this field for the discovery of the network of regulations among a group of genes, named Gene Regulation Networks (GRN), by analyzing the genes expression profiles represented as timeseries. In it has been proposed the GRNNminer method, which allows discovering the subyacent GRN among a group of genes, through the proper modeling of the temporal dynamics of the gene expression profiles with artificial neural networks. However, it implies building and training a pool of neural models for each possible gentogen relationship, which derives in executing a very large set of experiments with O( n 2 ) order, where n is the total of involved genes. This work presents a proposal for dramatically reducing such experiments number to O( (n/k)2 ) when big timeseries is involved for reconstructing a GRN from such data, by previously clustering genes profiles in k groups using selforganizing maps (SOM). This way, the GRNNminer can be applied over smaller sets of timeseries, only those appearing in the same cluster. http://sedici.unlp.edu.ar:80/handle/10915/51958 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) Cell phone technology generates massive amounts of data. Although this data has been gathered for billing and logging purposes, today it has a much higher value, because its volume makes it very useful for big data analyses. In this project, we analyse the viability of using cell phone records to lower the cost of urban and transportation planning, in particular, to find out how people travel in a specific city (in this case, Buenos Aires, in Argentina). We use cell phones data to estimate the distribution of the population in the city using different periods of time. We compare those results with traditional methods (urban polling) using data from Buenos Aires origin-destination surveys. Traditional polling methods have a much smaller sample, in the order of tens of thousands (or even less for smaller cities), to maintain reasonable costs. Furthermore, these studies are performed at most once per decade, in the best cases, in Argentina and many other countries. Our objective is to prove that new methods based on cell phone data are reliable, and can be used indirectly to keep a real-time track of the flow of people among different parts of a city. We also go further to explore new possibilities opened by these methods. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51958 2015-01-01T00:00:00Z Cell phone technology generates massive amounts of data. Although this data has been gathered for billing and logging purposes, today it has a much higher value, because its volume makes it very useful for big data analyses. In this project, we analyse the viability of using cell phone records to lower the cost of urban and transportation planning, in particular, to find out how people travel in a specific city (in this case, Buenos Aires, in Argentina). We use cell phones data to estimate the distribution of the population in the city using different periods of time. We compare those results with traditional methods (urban polling) using data from Buenos Aires origin-destination surveys. Traditional polling methods have a much smaller sample, in the order of tens of thousands (or even less for smaller cities), to maintain reasonable costs. Furthermore, these studies are performed at most once per decade, in the best cases, in Argentina and many other countries. Our objective is to prove that new methods based on cell phone data are reliable, and can be used indirectly to keep a real-time track of the flow of people among different parts of a city. We also go further to explore new possibilities opened by these methods. http://sedici.unlp.edu.ar:80/handle/10915/51956 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) La Estación Experimental Agropecuaria (EEA) INTA Anguil (La Pampa) tiene instalado un radar meteorológico (RM) capaz de generar un flujo de datos de 17gb diarios aproximadamente, lo que representa un desafío para su almacenamiento, gestión y posterior servicio al público, principalmente considerando la importancia que los datos poseen para la región productiva de influencia. Este trabajo propone una arquitectura de procesamiento que extiende el EIPFDcMM para incorporar el procesamiento de grandes datos almacenados. Ello permitiría dotar al RM de un comportamiento detectivo y predictivo sobre los datos en línea, como así también incluir una capa de servicios al público, que fomente el consumo de datos generados por el RM de INTA Anguil. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51956 2015-01-01T00:00:00Z La Estación Experimental Agropecuaria (EEA) INTA Anguil (La Pampa) tiene instalado un radar meteorológico (RM) capaz de generar un flujo de datos de 17gb diarios aproximadamente, lo que representa un desafío para su almacenamiento, gestión y posterior servicio al público, principalmente considerando la importancia que los datos poseen para la región productiva de influencia. Este trabajo propone una arquitectura de procesamiento que extiende el EIPFDcMM para incorporar el procesamiento de grandes datos almacenados. Ello permitiría dotar al RM de un comportamiento detectivo y predictivo sobre los datos en línea, como así también incluir una capa de servicios al público, que fomente el consumo de datos generados por el RM de INTA Anguil. http://sedici.unlp.edu.ar:80/handle/10915/51955 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) El trabajo busca describir las tecnolog as, procesos y algoritmos aplicados para construir la plataforma de real-time bidding de Jampp y c omo se pretende evolucionar la misma. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51955 2015-01-01T00:00:00Z El trabajo busca describir las tecnolog as, procesos y algoritmos aplicados para construir la plataforma de real-time bidding de Jampp y c omo se pretende evolucionar la misma. http://sedici.unlp.edu.ar:80/handle/10915/51953 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) En este artículo se presentan los resultados de un proyecto de análisis de datos de una empresa de transporte, que involucró la recolección, preparación, visualización, transformación y análisis de 3 años de datos de viajes de colectivos, incluyendo boletos y posicionamiento geográfico. Este caso cubre el proyecto de punta a punta, incluyendo la incorporación de los resultados en el proceso de negocio. Nosotros sostenemos que una característica clave de los proyectos de big data debe encontrarse en el proceso que se lleva a cabo y que inicia con la captura de grandes cantidades de datos, pasando por el procesamiento (que en muchos casos requiere una infraestructura especial o particular, con más de una computadora, en modo distribuido) hasta el análisis y el aprovechamiento de la información en el negocio. En nuestro punto de vista, este último paso (la inserción de la información en la toma de decisiones) es tan importante como el uso de bases NoSQL o Hadoop o procesar varios terabytes. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51953 2015-01-01T00:00:00Z En este artículo se presentan los resultados de un proyecto de análisis de datos de una empresa de transporte, que involucró la recolección, preparación, visualización, transformación y análisis de 3 años de datos de viajes de colectivos, incluyendo boletos y posicionamiento geográfico. Este caso cubre el proyecto de punta a punta, incluyendo la incorporación de los resultados en el proceso de negocio. Nosotros sostenemos que una característica clave de los proyectos de big data debe encontrarse en el proceso que se lleva a cabo y que inicia con la captura de grandes cantidades de datos, pasando por el procesamiento (que en muchos casos requiere una infraestructura especial o particular, con más de una computadora, en modo distribuido) hasta el análisis y el aprovechamiento de la información en el negocio. En nuestro punto de vista, este último paso (la inserción de la información en la toma de decisiones) es tan importante como el uso de bases NoSQL o Hadoop o procesar varios terabytes. http://sedici.unlp.edu.ar:80/handle/10915/51952 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) Web Search Engines process huge amounts of data to support search but must run under strong performance requirements (to answer a query in a fraction of a second). To meet that performance they implement different optimization techniques such as caching, that may be implemented at several levels. One of these caching levels is the intersection cache, that attempts to exploit frequently occurring pairs of terms by keeping in the memory of the search node the results of intersecting the corresponding inverted lists. In this work we propose an optimization step to decide which items should be cached and which not by introducing the usage of data mining techniques. Our preliminary results show that it is possible to achieve extra cost savings in this already hyper-optimized field. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51952 2015-01-01T00:00:00Z Web Search Engines process huge amounts of data to support search but must run under strong performance requirements (to answer a query in a fraction of a second). To meet that performance they implement different optimization techniques such as caching, that may be implemented at several levels. One of these caching levels is the intersection cache, that attempts to exploit frequently occurring pairs of terms by keeping in the memory of the search node the results of intersecting the corresponding inverted lists. In this work we propose an optimization step to decide which items should be cached and which not by introducing the usage of data mining techniques. Our preliminary results show that it is possible to achieve extra cost savings in this already hyper-optimized field. http://sedici.unlp.edu.ar:80/handle/10915/51951 Objeto de conferencia Simposio Argentino de GRANdes DAtos (AGRANDA 2015) - JAIIO 44 (Rosario, 2015) El objetivo del presente documento es presentar los resultados preliminares de la Encuesta Big Data (2015) realizada desde el CIECTI a empresas del sector y grupos de investigación en el país. Se trata de un trabajo en curso que requiere ampliar la muestra de encuestados. En esta primera etapa buscamos caracterizar el sector en base a una selección de actores del sector académico y empresario, esbozando un primer diagnóstico en el marco de las políticas de desarrollo argentinas y el rol del Sector Público como dinamizador. Thu, 01 Jan 2015 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/51951 2015-01-01T00:00:00Z El objetivo del presente documento es presentar los resultados preliminares de la Encuesta Big Data (2015) realizada desde el CIECTI a empresas del sector y grupos de investigación en el país. Se trata de un trabajo en curso que requiere ampliar la muestra de encuestados. En esta primera etapa buscamos caracterizar el sector en base a una selección de actores del sector académico y empresario, esbozando un primer diagnóstico en el marco de las políticas de desarrollo argentinas y el rol del Sector Público como dinamizador.