Mutual Information and Perplexity based Clustering of Dialogue Information for Dynamic Adaptation of Language Models

We present two approaches to cluster dialogue-based information obtained by the speech understanding module and the dialogue manager of a spoken dialogue system. The purpose is to estimate a language model related to each cluster, and use them to dynamically modify the model of the speech recognizer at each dialogue turn. In the first approach we build the cluster tree using local decisions based on a Maximum Normalized Mutual Information criterion. In the second one we take global decisions, based on the optimization of the global perplexity of the combination of the cluster-related LMs. Our experiments show a relative reduction of the word error rate of 15.17%, which helps to improve the performance of the understanding and the dialogue manager modules.

Mutual Information and Perplexity based Clustering of Dialogue Information for Dynamic Adaptation of Language Models

We present two approaches to cluster dialogue-based information obtained by the speech understanding module and the dialogue manager of a spoken dialogue system. The purpose is to estimate a language model related to each cluster, and use them to dynamically modify the model of the speech recognizer at each dialogue turn. In the first approach we build the cluster tree using local decisions based on a Maximum Normalized Mutual Information criterion. In the second one we take global decisions, based on the optimization of the global perplexity of the combination of the cluster-related LMs. Our experiments show a relative reduction of the word error rate of 15.17%, which helps to improve the performance of the understanding and the dialogue manager modules.

Multi-dimensional classification using Bayesian networks for stationary and evolving streaming data

Hoy en día, con la evolución continua y rápida de las tecnologías de la información y los dispositivos de computación, se recogen y almacenan continuamente grandes volúmenes de datos en distintos dominios y a través de diversas aplicaciones del mundo real. La extracción de conocimiento útil de una cantidad tan enorme de datos no se puede realizar habitualmente de forma manual, y requiere el uso de técnicas adecuadas de aprendizaje automático y de minería de datos. La clasificación es una de las técnicas más importantes que ha sido aplicada con éxito a varias áreas. En general, la clasificación se compone de dos pasos principales: en primer lugar, aprender un modelo de clasificación o clasificador a partir de un conjunto de datos de entrenamiento, y en segundo lugar, clasificar las nuevas instancias de datos utilizando el clasificador aprendido. La clasificación es supervisada cuando todas las etiquetas están presentes en los datos de entrenamiento (es decir, datos completamente etiquetados), semi-supervisada cuando sólo algunas etiquetas son conocidas (es decir, datos parcialmente etiquetados), y no supervisada cuando todas las etiquetas están ausentes en los datos de entrenamiento (es decir, datos no etiquetados). Además, aparte de esta taxonomía, el problema de clasificación se puede categorizar en unidimensional o multidimensional en función del número de variables clase, una o más, respectivamente; o también puede ser categorizado en estacionario o cambiante con el tiempo en función de las características de los datos y de la tasa de cambio subyacente. A lo largo de esta tesis, tratamos el problema de clasificación desde tres perspectivas diferentes, a saber, clasificación supervisada multidimensional estacionaria, clasificación semisupervisada unidimensional cambiante con el tiempo, y clasificación supervisada multidimensional cambiante con el tiempo. Para llevar a cabo esta tarea, hemos usado básicamente los clasificadores Bayesianos como modelos. La primera contribución, dirigiéndose al problema de clasificación supervisada multidimensional estacionaria, se compone de dos nuevos métodos de aprendizaje de clasificadores Bayesianos multidimensionales a partir de datos estacionarios. Los métodos se proponen desde dos puntos de vista diferentes. El primer método, denominado CB-MBC, se basa en una estrategia de envoltura de selección de variables que es voraz y hacia delante, mientras que el segundo, denominado MB-MBC, es una estrategia de filtrado de variables con una aproximación basada en restricciones y en el manto de Markov. Ambos métodos han sido aplicados a dos problemas reales importantes, a saber, la predicción de los inhibidores de la transcriptasa inversa y de la proteasa para el problema de infección por el virus de la inmunodeficiencia humana tipo 1 (HIV-1), y la predicción del European Quality of Life-5 Dimensions (EQ-5D) a partir de los cuestionarios de la enfermedad de Parkinson con 39 ítems (PDQ-39). El estudio experimental incluye comparaciones de CB-MBC y MB-MBC con los métodos del estado del arte de la clasificación multidimensional, así como con métodos comúnmente utilizados para resolver el problema de predicción de la enfermedad de Parkinson, a saber, la regresión logística multinomial, mínimos cuadrados ordinarios, y mínimas desviaciones absolutas censuradas. En ambas aplicaciones, los resultados han sido prometedores con respecto a la precisión de la clasificación, así como en relación al análisis de las estructuras gráficas que identifican interacciones conocidas y novedosas entre las variables. La segunda contribución, referida al problema de clasificación semi-supervisada unidimensional cambiante con el tiempo, consiste en un método nuevo (CPL-DS) para clasificar flujos de datos parcialmente etiquetados. Los flujos de datos difieren de los conjuntos de datos estacionarios en su proceso de generación muy rápido y en su aspecto de cambio de concepto. Es decir, los conceptos aprendidos y/o la distribución subyacente están probablemente cambiando y evolucionando en el tiempo, lo que hace que el modelo de clasificación actual sea obsoleto y deba ser actualizado. CPL-DS utiliza la divergencia de Kullback-Leibler y el método de bootstrapping para cuantificar y detectar tres tipos posibles de cambio: en las predictoras, en la a posteriori de la clase o en ambas. Después, si se detecta cualquier cambio, un nuevo modelo de clasificación se aprende usando el algoritmo EM; si no, el modelo de clasificación actual se mantiene sin modificaciones. CPL-DS es general, ya que puede ser aplicado a varios modelos de clasificación. Usando dos modelos diferentes, el clasificador naive Bayes y la regresión logística, CPL-DS se ha probado con flujos de datos sintéticos y también se ha aplicado al problema real de la detección de código malware, en el cual los nuevos ficheros recibidos deben ser continuamente clasificados en malware o goodware. Los resultados experimentales muestran que nuestro método es efectivo para la detección de diferentes tipos de cambio a partir de los flujos de datos parcialmente etiquetados y también tiene una buena precisión de la clasificación. Finalmente, la tercera contribución, sobre el problema de clasificación supervisada multidimensional cambiante con el tiempo, consiste en dos métodos adaptativos, a saber, Locally Adpative-MB-MBC (LA-MB-MBC) y Globally Adpative-MB-MBC (GA-MB-MBC). Ambos métodos monitorizan el cambio de concepto a lo largo del tiempo utilizando la log-verosimilitud media como métrica y el test de Page-Hinkley. Luego, si se detecta un cambio de concepto, LA-MB-MBC adapta el actual clasificador Bayesiano multidimensional localmente alrededor de cada nodo cambiado, mientras que GA-MB-MBC aprende un nuevo clasificador Bayesiano multidimensional. El estudio experimental realizado usando flujos de datos sintéticos multidimensionales indica los méritos de los métodos adaptativos propuestos. ABSTRACT Nowadays, with the ongoing and rapid evolution of information technology and computing devices, large volumes of data are continuously collected and stored in different domains and through various real-world applications. Extracting useful knowledge from such a huge amount of data usually cannot be performed manually, and requires the use of adequate machine learning and data mining techniques. Classification is one of the most important techniques that has been successfully applied to several areas. Roughly speaking, classification consists of two main steps: first, learn a classification model or classifier from an available training data, and secondly, classify the new incoming unseen data instances using the learned classifier. Classification is supervised when the whole class values are present in the training data (i.e., fully labeled data), semi-supervised when only some class values are known (i.e., partially labeled data), and unsupervised when the whole class values are missing in the training data (i.e., unlabeled data). In addition, besides this taxonomy, the classification problem can be categorized into uni-dimensional or multi-dimensional depending on the number of class variables, one or more, respectively; or can be also categorized into stationary or streaming depending on the characteristics of the data and the rate of change underlying it. Through this thesis, we deal with the classification problem under three different settings, namely, supervised multi-dimensional stationary classification, semi-supervised unidimensional streaming classification, and supervised multi-dimensional streaming classification. To accomplish this task, we basically used Bayesian network classifiers as models. The first contribution, addressing the supervised multi-dimensional stationary classification problem, consists of two new methods for learning multi-dimensional Bayesian network classifiers from stationary data. They are proposed from two different points of view. The first method, named CB-MBC, is based on a wrapper greedy forward selection approach, while the second one, named MB-MBC, is a filter constraint-based approach based on Markov blankets. Both methods are applied to two important real-world problems, namely, the prediction of the human immunodeficiency virus type 1 (HIV-1) reverse transcriptase and protease inhibitors, and the prediction of the European Quality of Life-5 Dimensions (EQ-5D) from 39-item Parkinson’s Disease Questionnaire (PDQ-39). The experimental study includes comparisons of CB-MBC and MB-MBC against state-of-the-art multi-dimensional classification methods, as well as against commonly used methods for solving the Parkinson’s disease prediction problem, namely, multinomial logistic regression, ordinary least squares, and censored least absolute deviations. For both considered case studies, results are promising in terms of classification accuracy as well as regarding the analysis of the learned MBC graphical structures identifying known and novel interactions among variables. The second contribution, addressing the semi-supervised uni-dimensional streaming classification problem, consists of a novel method (CPL-DS) for classifying partially labeled data streams. Data streams differ from the stationary data sets by their highly rapid generation process and their concept-drifting aspect. That is, the learned concepts and/or the underlying distribution are likely changing and evolving over time, which makes the current classification model out-of-date requiring to be updated. CPL-DS uses the Kullback-Leibler divergence and bootstrapping method to quantify and detect three possible kinds of drift: feature, conditional or dual. Then, if any occurs, a new classification model is learned using the expectation-maximization algorithm; otherwise, the current classification model is kept unchanged. CPL-DS is general as it can be applied to several classification models. Using two different models, namely, naive Bayes classifier and logistic regression, CPL-DS is tested with synthetic data streams and applied to the real-world problem of malware detection, where the new received files should be continuously classified into malware or goodware. Experimental results show that our approach is effective for detecting different kinds of drift from partially labeled data streams, as well as having a good classification performance. Finally, the third contribution, addressing the supervised multi-dimensional streaming classification problem, consists of two adaptive methods, namely, Locally Adaptive-MB-MBC (LA-MB-MBC) and Globally Adaptive-MB-MBC (GA-MB-MBC). Both methods monitor the concept drift over time using the average log-likelihood score and the Page-Hinkley test. Then, if a drift is detected, LA-MB-MBC adapts the current multi-dimensional Bayesian network classifier locally around each changed node, whereas GA-MB-MBC learns a new multi-dimensional Bayesian network classifier from scratch. Experimental study carried out using synthetic multi-dimensional data streams shows the merits of both proposed adaptive methods.

On the dynamic adaptation of language models based on dialogue information

We present an approach to adapt dynamically the language models (LMs) used by a speech recognizer that is part of a spoken dialogue system. We have developed a grammar generation strategy that automatically adapts the LMs using the semantic information that the user provides (represented as dialogue concepts), together with the information regarding the intentions of the speaker (inferred by the dialogue manager, and represented as dialogue goals). We carry out the adaptation as a linear interpolation between a background LM, and one or more of the LMs associated to the dialogue elements (concepts or goals) addressed by the user. The interpolation weights between those models are automatically estimated on each dialogue turn, using measures such as the posterior probabilities of concepts and goals, estimated as part of the inference procedure to determine the actions to be carried out. We propose two approaches to handle the LMs related to concepts and goals. Whereas in the first one we estimate a LM for each one of them, in the second one we apply several clustering strategies to group together those elements that share some common properties, and estimate a LM for each cluster. Our evaluation shows how the system can estimate a dynamic model adapted to each dialogue turn, which helps to improve the performance of the speech recognition (up to a 14.82% of relative improvement), which leads to an improvement in both the language understanding and the dialogue management tasks.

Caracterización y simulación de arborizaciones dentríticas con redes bayesianas incluyendo variables angulares

El funcionamiento interno del cerebro es todavía hoy en día un misterio, siendo su comprensión uno de los principales desafíos a los que se enfrenta la ciencia moderna. El córtex cerebral es el área del cerebro donde tienen lugar los procesos cerebrales de más alto nivel, cómo la imaginación, el juicio o el pensamiento abstracto. Las neuronas piramidales, un tipo específico de neurona, suponen cerca del 80% de los cerca de los 10.000 millones de que componen el córtex cerebral, haciendo de ellas un objetivo principal en el estudio del funcionamiento del cerebro. La morfología neuronal, y más específicamente la morfología dendrítica, determina cómo estas procesan la información y los patrones de conexión entre neuronas, siendo los modelos computacionales herramientas imprescindibles para el estudio de su rol en el funcionamiento del cerebro. En este trabajo hemos creado un modelo computacional, con más de 50 variables relativas a la morfología dendrítica, capaz de simular el crecimiento de arborizaciones dendríticas basales completas a partir de reconstrucciones de neuronas piramidales reales, abarcando desde el número de dendritas hasta el crecimiento los los árboles dendríticos. A diferencia de los trabajos anteriores, nuestro modelo basado en redes Bayesianas contempla la arborización dendrítica en su conjunto, teniendo en cuenta las interacciones entre dendritas y detectando de forma automática las relaciones entre las variables morfológicas que caracterizan la arborización. Además, el análisis de las redes Bayesianas puede ayudar a identificar relaciones hasta ahora desconocidas entre variables morfológicas. Motivado por el estudio de la orientación de las dendritas basales, en este trabajo se introduce una regularización L1 generalizada, aplicada al aprendizaje de la distribución von Mises multivariante, una de las principales distribuciones de probabilidad direccional multivariante. También se propone una distancia circular multivariante que puede utilizarse para estimar la divergencia de Kullback-Leibler entre dos muestras de datos circulares. Comparamos los modelos con y sin regularizaci ón en el estudio de la orientación de la dendritas basales en neuronas humanas, comprobando que, en general, el modelo regularizado obtiene mejores resultados. El muestreo, ajuste y representación de la distribución von Mises multivariante se implementa en un nuevo paquete de R denominado mvCircular.---ABSTRACT---The inner workings of the brain are, as of today, a mystery. To understand the brain is one of the main challenges faced by current science. The cerebral cortex is the region of the brain where all superior brain processes, like imagination, judge and abstract reasoning take place. Pyramidal neurons, a specific type of neurons, constitute approximately the 80% of the more than 10.000 million neurons that compound the cerebral cortex. It makes the study of the pyramidal neurons crucial in order to understand how the brain works. Neuron morphology, and specifically the dendritic morphology, determines how the information is processed in the neurons, as well as the connection patterns among neurons. Computational models are one of the main tools for studying dendritic morphology and its role in the brain function. We have built a computational model that contains more than 50 morphological variables of the dendritic arborizations. This model is able to simulate the growth of complete dendritic arborizations from real neuron reconstructions, starting with the number of basal dendrites, and ending modeling the growth of dendritic trees. One of the main diferences between our approach, mainly based on the use of Bayesian networks, and other models in the state of the art is that we model the whole dendritic arborization instead of focusing on individual trees, which makes us able to take into account the interactions between dendrites and to automatically detect relationships between the morphologic variables that characterize the arborization. Moreover, the posterior analysis of the relationships in the model can help to identify new relations between morphological variables. Motivated by the study of the basal dendrites orientation, a generalized L1 regularization applied to the multivariate von Mises distribution, one of the most used distributions in multivariate directional statistics, is also introduced in this work. We also propose a circular multivariate distance that can be used to estimate the Kullback-Leibler divergence between two circular data samples. We compare the regularized and unregularized models on basal dendrites orientation of human neurons and prove that regularized model achieves better results than non regularized von Mises model. Sampling, fitting and plotting functions for the multivariate von Mises are implemented in a new R packaged called mvCircular.

Dynamic topic-based adaptation of language models: a comparison between different approaches

This paper presents a dynamic LM adaptation based on the topic that has been identified on a speech segment. We use LSA and the given topic labels in the training dataset to obtain and use the topic models. We propose a dynamic language model adaptation to improve the recognition performance in "a two stages" AST system. The final stage makes use of the topic identification with two variants: the first on uses just the most probable topic and the other one depends on the relative distances of the topics that have been identified. We perform the adaptation of the LM as a linear interpolation between a background model and topic-based LM. The interpolation weight id dynamically adapted according to different parameters. The proposed method is evaluated on the Spanish partition of the EPPS speech database. We achieved a relative reduction in WER of 11.13% over the baseline system which uses a single blackground LM.

The role of standards in the growth of global electronic commerce : hearing before the Subcommittee on Science, Technology, and Space of the Committee on Commerce, Science, and Transportation, United States Senate, One Hundred Sixth Congress, first session, October 28, 1999.

Shipping list no.: 2002-0187-P.

Electronic commerce investments, the resource-based view of the firm, and firm market value

Firms have embraced electronic commerce as a means of doing business, either because they see it as a way to improve efficiency, grow market share, expand into new markets, or because they view it as essential for survival. Recent research in the United States provides some evidence that the market does value investments in electronic commerce. Following research that suggests that, in certain circumstances, the market values noninnovative investments as well as innovative investments in new products, we partition electronic commerce investment project announcements into innovative and noninnovative to determine whether there are excess returns associated with these types of announcements. Apart from our overall results being consistent with the United States findings that the market values investments in electronic commerce projects, we also find that noninnovative investments are perceived as more valuable to the firm than innovative investments. On average, the market expects innovative investments to earn a return commensurate with their risk. We conclude that innovative electronic commerce projects are most likely seen by the capital market as easily replicable, and consequently have little, if any, competitive advantage period. On the other hand, we conclude from the noninnovative investment results that these types of investments are seen as being compatible with a firm's assets-in-place, in particular, its information technology capabilities, a view consistent with the resource-based view of the firm.

Adoption of electronic commerce tools in business procurement:enhanced buying center structure and processes

Modern procurement is being shifted from paper-based, people-intensive buying systems toward electronic-based purchase procedures that rely on Internet communications and Web-enhanced buying tools. Develops a typology of e-commerce tools that have come to characterize cutting-edge industrial procurement. E-commerce aspects of purchasing are organized into communication and transaction tools that encompass both internal and external buying activities. Further, a model of the impact of e-commerce on the structure and processes of an organization's buying center is developed. The impact of the changing buying center on procurement outcomes in terms of efficiency and effectiveness is also analyzed. Finally, implications for business-to-business marketers and researchers are discussed.