Networks (i.e., graphs) appears in many high-impact applications. Often these networks are collected from different sources, at different times, at different granularities. In this talk, I will present our recent work on mining such multiple networks. First, we will present two models - one on modeling a set of inter-connected networks (NoN); and the other on modeling a set of inter-connected co-evolving time series (NoT). For both models, we will show that by treating networks as context, we are able to model more complicate real-world applications. Second, we will present some algorithmic examples on how to do mining with such new models, including ranking, imputation and prediction. Finally, we will demonstrate the effectiveness of our new models and algorithms in some applications, including bioinformatics, and sensor networks. ]]>

Networks (i.e., graphs) appears in many high-impact applications. Often these networks are collected from different sources, at different times, at different granularities. In this talk, I will present our recent work on mining such multiple networks. First, we will present two models - one on modeling a set of inter-connected networks (NoN); and the other on modeling a set of inter-connected co-evolving time series (NoT). For both models, we will show that by treating networks as context, we are able to model more complicate real-world applications. Second, we will present some algorithmic examples on how to do mining with such new models, including ranking, imputation and prediction. Finally, we will demonstrate the effectiveness of our new models and algorithms in some applications, including bioinformatics, and sensor networks. ]]>

This talk will discuss examples of how Amazon applies machine learning to large-scale data, and open research questions inspired by these applications. One important question is how to distinguish between users that can be influenced, versus those who are merely likely to respond. Another question is how to measure and maximize the long-term benefit of movie and other recommendations. A third question, is how to share data while provably protecting the privacy of users. Note: Information in the talk is already public, and opinions expressed will be strictly personal.]]>

This talk will discuss examples of how Amazon applies machine learning to large-scale data, and open research questions inspired by these applications. One important question is how to distinguish between users that can be influenced, versus those who are merely likely to respond. Another question is how to measure and maximize the long-term benefit of movie and other recommendations. A third question, is how to share data while provably protecting the privacy of users. Note: Information in the talk is already public, and opinions expressed will be strictly personal.]]>

Apache Spark has become the most active open source Big Data project, and its Machine Learning library MLlib has seen rapid growth in usage. A critical aspect of MLlib and Spark is the ability to scale: the same code used on a laptop can scale to 100’s or 1000’s of machines. This talk will describe ongoing and future efforts to make MLlib even faster and more scalable by integrating with two key initiatives in Spark. The first is Catalyst, the query optimizer underlying DataFrames and Datasets. The second is Tungsten, the project for approaching bare-metal speeds in Spark via memory management, cache-awareness, and code generation. This talk will discuss the goals, the challenges, and the benefits for MLlib users and developers. More generally, we will reflect on the importance of integrating ML with the many other aspects of big data analysis. About MLlib: MLlib is a general Machine Learning library providing many ML algorithms, feature transformers, and tools for model tuning and building workflows. The library benefits from integration with the rest of Apache Spark (SQL, streaming, Graph, core), which facilitates ETL, streaming, and deployment. It is used in both ad hoc analysis and production deployments throughout academia and industry. ]]>

Apache Spark has become the most active open source Big Data project, and its Machine Learning library MLlib has seen rapid growth in usage. A critical aspect of MLlib and Spark is the ability to scale: the same code used on a laptop can scale to 100’s or 1000’s of machines. This talk will describe ongoing and future efforts to make MLlib even faster and more scalable by integrating with two key initiatives in Spark. The first is Catalyst, the query optimizer underlying DataFrames and Datasets. The second is Tungsten, the project for approaching bare-metal speeds in Spark via memory management, cache-awareness, and code generation. This talk will discuss the goals, the challenges, and the benefits for MLlib users and developers. More generally, we will reflect on the importance of integrating ML with the many other aspects of big data analysis. About MLlib: MLlib is a general Machine Learning library providing many ML algorithms, feature transformers, and tools for model tuning and building workflows. The library benefits from integration with the rest of Apache Spark (SQL, streaming, Graph, core), which facilitates ETL, streaming, and deployment. It is used in both ad hoc analysis and production deployments throughout academia and industry. ]]>

How can we learn from the data of small ubiquitous systems? Do we need to send the data to a server or cloud and do all learning there? Or can we learn on some small devices directly? Are smartphones small? Are navigation systems small? How complex is learning allowed to be in times of big data? What about graphical models? Can they be applied on small devices or even learned on restricted processors? Big data are produced by various sources. Most often, they are distributedly stored at computing farms or clouds. Analytics on the Hadoop Distributed File System (HDFS) then follows the MapReduce programming model. According to the Lambda architecture of Nathan Marz and James Warren, this is the batch layer. It is complemented by the speed layer, which aggregates and integrates incoming data streams in real time. When considering big data and small devices, obviously, we imagine the small devices being hosts of the speed layer, only. Analytics on the small devices is restricted by memory and computation resources. The interplay of streaming and batch analytics offers a multitude of configurations. In this talk, we discuss opportunities for using sophisticated models for learning spatio-temporal models. In particular, we investigate graphical models, which generate the probabilities for connected (sensor) nodes. First, we present spatio-temporal random fields that take as input data from small devices, are computed at a server, and send results to -possibly different — small devices. Second, we go even further: the Integer Markov Random Field approximates the likelihood estimates such that it can be computed on small devices. We illustrate our learning models by applications from traffic management.]]>

How can we learn from the data of small ubiquitous systems? Do we need to send the data to a server or cloud and do all learning there? Or can we learn on some small devices directly? Are smartphones small? Are navigation systems small? How complex is learning allowed to be in times of big data? What about graphical models? Can they be applied on small devices or even learned on restricted processors? Big data are produced by various sources. Most often, they are distributedly stored at computing farms or clouds. Analytics on the Hadoop Distributed File System (HDFS) then follows the MapReduce programming model. According to the Lambda architecture of Nathan Marz and James Warren, this is the batch layer. It is complemented by the speed layer, which aggregates and integrates incoming data streams in real time. When considering big data and small devices, obviously, we imagine the small devices being hosts of the speed layer, only. Analytics on the small devices is restricted by memory and computation resources. The interplay of streaming and batch analytics offers a multitude of configurations. In this talk, we discuss opportunities for using sophisticated models for learning spatio-temporal models. In particular, we investigate graphical models, which generate the probabilities for connected (sensor) nodes. First, we present spatio-temporal random fields that take as input data from small devices, are computed at a server, and send results to -possibly different — small devices. Second, we go even further: the Integer Markov Random Field approximates the likelihood estimates such that it can be computed on small devices. We illustrate our learning models by applications from traffic management.]]>

Recent technological innovations have enabled data collection of unprecedented size and complexity. Examples include web text data, social media data, gene expression images, neuroimages, and genome-wide association study (GWAS) data. Such data have incredible potential to address complex scientific and societal questions, however analysis of these data poses major challenges for the scientists. As an emerging and powerful tool for analyzing massive collections of data, data reduction in terms of the number of variables and/or the number of samples has attracted tremendous attentions in the past few years, and has achieved great success in a broad range of applications. The intuition of data reduction is based on the observation that many real-world data with complex structures and billions of variables and/or samples can usually be well explained by a few most relevant explanatory features and/or samples. Most existing methods for data reduction are based on sampling or random projection, and the final model based on the reduced data is an approximation of the true (original) model. In this talk, I will present fundamentally different approaches for data reduction in that there is no approximation in the model, that is, the final model constructed from the reduced data is identical to the original model constructed from the complete data. Finally, I will use several real world examples to demonstrate the potential of exact data reduction for analyzing big data.]]>

Recent technological innovations have enabled data collection of unprecedented size and complexity. Examples include web text data, social media data, gene expression images, neuroimages, and genome-wide association study (GWAS) data. Such data have incredible potential to address complex scientific and societal questions, however analysis of these data poses major challenges for the scientists. As an emerging and powerful tool for analyzing massive collections of data, data reduction in terms of the number of variables and/or the number of samples has attracted tremendous attentions in the past few years, and has achieved great success in a broad range of applications. The intuition of data reduction is based on the observation that many real-world data with complex structures and billions of variables and/or samples can usually be well explained by a few most relevant explanatory features and/or samples. Most existing methods for data reduction are based on sampling or random projection, and the final model based on the reduced data is an approximation of the true (original) model. In this talk, I will present fundamentally different approaches for data reduction in that there is no approximation in the model, that is, the final model constructed from the reduced data is identical to the original model constructed from the complete data. Finally, I will use several real world examples to demonstrate the potential of exact data reduction for analyzing big data.]]>

Massive graphs are ubiquitous in various application domains, such as social networks, road networks, communication networks, biological networks, RDF graphs, and so on. Such graphs are massive (for example, with hundreds of millions of nodes and edges or even more) and contain rich information (for example, node/edge weights, labels and textual contents). In such massive graphs, an important class of problems is to process various graph structure related queries. Graph reachability, as an example, asks whether a node can reach another in a graph. However, the large graph scale presents new challenges for efficient query processing. In this talk, I will introduce two new yet important types of graph reachability queries: weight constraint reachability that imposes edge weight constraint on the answer path, and k-hop reachability that imposes a length constraint on the answer path. With such realistic constraints, we can find more meaningful and practically feasible answers. These two reachablity queries have wide applications in many real-world problems, such as QoS routing and trip planning.]]>

Massive graphs are ubiquitous in various application domains, such as social networks, road networks, communication networks, biological networks, RDF graphs, and so on. Such graphs are massive (for example, with hundreds of millions of nodes and edges or even more) and contain rich information (for example, node/edge weights, labels and textual contents). In such massive graphs, an important class of problems is to process various graph structure related queries. Graph reachability, as an example, asks whether a node can reach another in a graph. However, the large graph scale presents new challenges for efficient query processing. In this talk, I will introduce two new yet important types of graph reachability queries: weight constraint reachability that imposes edge weight constraint on the answer path, and k-hop reachability that imposes a length constraint on the answer path. With such realistic constraints, we can find more meaningful and practically feasible answers. These two reachablity queries have wide applications in many real-world problems, such as QoS routing and trip planning.]]>

In today’s interconnected real world, social and informational entities are interconnected, forming gigantic, interconnected, integrated social and information networks. By structuring these data objects into multiple types, such networks become semi-structured heterogeneous social and information networks. Most real world applications that handle big data, including interconnected social media and social networks, medical information systems, online e-commerce systems, or database systems, can be structured into typed, heterogeneous social and information networks. For example, in a medical care network, objects of multiple types, such as patients, doctors, diseases, medication, and links such as visits, diagnosis, and treatments are intertwined together, providing rich information and forming heterogeneous information networks. Effective analysis of large-scale heterogeneous social and information networks poses an interesting but critical challenge. In this talk, we present a set of data mining scenarios in heterogeneous social and information networks and show that mining typed, heterogeneous networks is a new and promising research frontier in data mining research. However, such mining may raise some serious challenging problems on scalability computation. We identify a set of problems on scalable computation and calls for serious studies on such problems. This includes how to efficiently computation for (1) meta path-based similarity search, (2) rank-based clustering, (3) rank-based classification, (4) meta path-based link/relationship prediction, and (5) topical hierarchies from heterogeneous information networks. We introduce some recent efforts, discuss the trade-offs between query-independent pre-computation vs. query-dependent online computation, and point out some promising research directions.]]>

In today’s interconnected real world, social and informational entities are interconnected, forming gigantic, interconnected, integrated social and information networks. By structuring these data objects into multiple types, such networks become semi-structured heterogeneous social and information networks. Most real world applications that handle big data, including interconnected social media and social networks, medical information systems, online e-commerce systems, or database systems, can be structured into typed, heterogeneous social and information networks. For example, in a medical care network, objects of multiple types, such as patients, doctors, diseases, medication, and links such as visits, diagnosis, and treatments are intertwined together, providing rich information and forming heterogeneous information networks. Effective analysis of large-scale heterogeneous social and information networks poses an interesting but critical challenge. In this talk, we present a set of data mining scenarios in heterogeneous social and information networks and show that mining typed, heterogeneous networks is a new and promising research frontier in data mining research. However, such mining may raise some serious challenging problems on scalability computation. We identify a set of problems on scalable computation and calls for serious studies on such problems. This includes how to efficiently computation for (1) meta path-based similarity search, (2) rank-based clustering, (3) rank-based classification, (4) meta path-based link/relationship prediction, and (5) topical hierarchies from heterogeneous information networks. We introduce some recent efforts, discuss the trade-offs between query-independent pre-computation vs. query-dependent online computation, and point out some promising research directions.]]>

Since the Netflix $1 million Prize, announced in 2006, our company has been known for having personalization at the core of our product. Even at that point in time, the dataset that we released was considered “large”, and we stirred innovation in the (Big) Data Mining research field. Our current product offering is now focused around instant video streaming, and our data is now many orders of magnitude larger. Not only do we have many more users in many more countries, but we also receive many more streams of data. Besides the ratings, we now also use information such as what our members play, browse, or search. In this talk I will discuss the different approaches we follow to deal with these large streams of data in order to extract information for personalizing our service. I will describe some of the machine learning models used, as well as the architectures that allow us to combine complex offline batch processes with real-time data streams.]]>

Since the Netflix $1 million Prize, announced in 2006, our company has been known for having personalization at the core of our product. Even at that point in time, the dataset that we released was considered “large”, and we stirred innovation in the (Big) Data Mining research field. Our current product offering is now focused around instant video streaming, and our data is now many orders of magnitude larger. Not only do we have many more users in many more countries, but we also receive many more streams of data. Besides the ratings, we now also use information such as what our members play, browse, or search. In this talk I will discuss the different approaches we follow to deal with these large streams of data in order to extract information for personalizing our service. I will describe some of the machine learning models used, as well as the architectures that allow us to combine complex offline batch processes with real-time data streams.]]>

What do graphs look like? How do they evolve over time? How does influence/news/viruses propagate, over time? We present a long list of static and temporal laws, and some recent observations on real graphs. We show that fractals and self-similarity can explain several of the observed patterns, and we conclude with cascade analysis and a surprising result on virus propagation and immunization. ]]>

What do graphs look like? How do they evolve over time? How does influence/news/viruses propagate, over time? We present a long list of static and temporal laws, and some recent observations on real graphs. We show that fractals and self-similarity can explain several of the observed patterns, and we conclude with cascade analysis and a surprising result on virus propagation and immunization. ]]>

Talk by Charles Parker (BigML) at BigMine12 at KDD12. In machine learning, scale adds complexity. The most obvious consequence of scale is that data takes longer to process. At certain points, however, scale makes trivial operations costly, thus forcing us to re-evaluate algorithms in light of the complexity of those operations. Here, we will discuss one important way a general large scale machine learning setting may differ from the standard supervised classification setting and show some the results of some preliminary experiments highlighting this difference. The results suggest that there is potential for significant improvement beyond obvious solutions.]]>

Talk by Charles Parker (BigML) at BigMine12 at KDD12. In machine learning, scale adds complexity. The most obvious consequence of scale is that data takes longer to process. At certain points, however, scale makes trivial operations costly, thus forcing us to re-evaluate algorithms in light of the complexity of those operations. Here, we will discuss one important way a general large scale machine learning setting may differ from the standard supervised classification setting and show some the results of some preliminary experiments highlighting this difference. The results suggest that there is potential for significant improvement beyond obvious solutions.]]>

Talk by Usama Fayyad at BigMine12 at KDD12. Virtually all organizations are having to deal with Big Data in many contexts: marketing, operations, monitoring, performance, and even financial management. Big Data is characterized not just by its size, but by its Velocity and its Variety for which keeping up with the data flux, let alone its analysis, is challenging at best and impossible in many cases. In this talk I will cover some of the basics in terms of infrastructure and design considerations for effective an efficient BigData. In many organizations, the lack of consideration of effective infrastructure and data management leads to unnecessarily expensive systems for which the benefits are insufficient to justify the costs. We will refer to example frameworks and clarify the kinds of operations where Map-Reduce (Hadoop and and its derivatives) are appropriate and the situations where other infrastructure is needed to perform segmentation, prediction, analysis, and reporting appropriately – these being the fundamental operations in predictive analytics. We will thenpay specific attention to on-line data and the unique challenges and opportunities represented there. We cover examples of Predictive Analytics over Big Data with case studies in eCommerce Marketing, on-line publishing and recommendation systems, and advertising targeting: Special focus will be placed on the analysis of on-line data with applications in Search, Search Marketing, and targeting of advertising. We conclude with some technical challenges as well as the solutions that can be used to these challenges in social network data.]]>

Talk by Usama Fayyad at BigMine12 at KDD12. Virtually all organizations are having to deal with Big Data in many contexts: marketing, operations, monitoring, performance, and even financial management. Big Data is characterized not just by its size, but by its Velocity and its Variety for which keeping up with the data flux, let alone its analysis, is challenging at best and impossible in many cases. In this talk I will cover some of the basics in terms of infrastructure and design considerations for effective an efficient BigData. In many organizations, the lack of consideration of effective infrastructure and data management leads to unnecessarily expensive systems for which the benefits are insufficient to justify the costs. We will refer to example frameworks and clarify the kinds of operations where Map-Reduce (Hadoop and and its derivatives) are appropriate and the situations where other infrastructure is needed to perform segmentation, prediction, analysis, and reporting appropriately – these being the fundamental operations in predictive analytics. We will thenpay specific attention to on-line data and the unique challenges and opportunities represented there. We cover examples of Predictive Analytics over Big Data with case studies in eCommerce Marketing, on-line publishing and recommendation systems, and advertising targeting: Special focus will be placed on the analysis of on-line data with applications in Search, Search Marketing, and targeting of advertising. We conclude with some technical challenges as well as the solutions that can be used to these challenges in social network data.]]>

]]>

]]>

]]>

]]>