In this work we look at left-vs.-right polarization in web search queries issued and hashtags being used. We look at queries issued to Yahoo! in the US, and assign a political leaning in relation to the results being clicked. E.g. a query returning predominantly The Huffington Post (a left-leaning site) would be labeled as left-leaning, as opposed to queries landing on The Drudge Report (a right-leaning site). We show that our methodology can gain us insights into how issues are "framed" by either side and present both expected and surprising results linking truth values, leaning and popularity.

MapReduce is now a pervasive analytics engine on the cloud. Hadoop is an open source implementation of MapReduce and is currently enjoying wide popularity. Hadoop offers a high-dimensional space of configuration parameters that makes it difficult on practitioners to set for efficient and cost-effective execution. In this work we observe that MapReduce application performance is highly influenced by map concurrency, defined in terms of two configurable parameters, the number of available map slots and the number of map tasks running over the slots.

This talk presents several examples of how Information Technology can help the 5 billion people at the bottom of the pyramid who do not now have routine access to devices and networks. For many of them, access to Word, Excel, and PPT or Programming are not that important. However, technology enabled solutions for education, healthcare, access to people and information, entertainment, surviving disasters, and finding jobs would be of great interest to them.

This talk is a gentle introduction to state-space search and model checking with LMNtal and LaViT, an IDE for LMNtal powered by visualizers. LMNtal is a language model based on hierarchical graph rewriting that uses point-to-point links to represent connectivity and membranes to represent hierarchy. Its expressive power was demonstrated through the encoding of various computational models including the lambda calculus. Our publicly available LMNtal implementation achieved high scalability and can handle models with half billion states.

Statistical Machine Translation (SMT) is the current mainstream approach to automatic translation between human languages. SMT relies on the existence of parallel bilingual corpora that are used to learn translation models automatically. Languages with different degrees of morphological complexity, such as Arabic and English, can be challenging to SMT because of the lack of symmetry between the two languages in addition to the increased sparsity that morphologically rich languages such as Arabic, pose.

All areas of science are being confronted with increasingly large datasets, and often their usual approaches to data analysis are struggling to cope. While advances in computational tools are addressing this problem, in some cases these are still not as refined as desired and traditional human-based techniques remain favoured. Citizen Science offers an opportunity to make progress with producing science from huge datasets, using methods which rely on the remarkable talents of the human brain, while simultaneously helping to improve automated techniques.

Despite the increasing importance of data quality and the rich theoretical and practical contributions in all aspects of data cleaning, there is no single end-to-end o -the-shelf solution to (semi-)automate the detection and the repairing of violations w.r.t. a set of heterogeneous and ad-hoc quality constraints. In short, there is no commodity platform similar to general purpose DBMSs that can be easily customized and deployed to solve application-specic data quality problems. In this talk, we present NADEEF, an extensible, generalized and easy-to-deploy cleaning platform developed at QCRI.

This presentation reports on the experience and lessons learned from introducing a constructivist Inquiry Based Learning (IBL) in advanced computing courses. Research-based projects are one of the IBL practices that are particularly suitable for higher education. Research-based IBL has been conducted in two different higher education contexts; at the University of Waterloo and at Qatar University. Measurement data provides information relative to the strengths and weaknesses of the adopted methodology in satisfying the pedagogical objectives being addressed.

While many organizations continue to focus on enabling sustainable development, very few organizations have studied the role of technology in this process. TechBridgeWorld at Carnegie Mellon University is spearheading the innovation and implementation of technological solutions relevant and accessible to developing communities, using technology to build bridges rather than exacerbate divides. Designing and implementing technology that can enhance suitable and sustainable development presents unique challenges in creativity and resourcefulness.

Minicomplexity is the computational complexity theory that we can build by analogy to standard Computational Complexity Theory, when our computational model and resource of interest are not the Turing machine and its running time, but instead the two-way finite automaton (2FA) and its size. This talk is an informal tour to the main concepts and open questions of this theory.

As our data sets grow in size, the need for techniques for processing such large data under limited resources becomes more critical. One model for processing large data sequences is that of streaming computation: the input is read sequentially, i.e., "streamed", in one long pass, and the computation is performed while using small (typically logarithmic in the size of the input) memory.

High throughput sequencing technologies have made it possible to sequence the whole genome of an individual human donor for a few thousand dollars. The world-wide capacity for genome sequence production has grown at an unprecedented rate, making the initiation of large scale projects aiming to sequence 1000s of individual human genomes possible. The ability to detect large scale insertions, deletions, inversions and duplications (in short structural variants) between high throughput sequenced genomes is key to diagnosing, monitoring and treating diseases and conditions of genomic origin.

Goggles, wearable cameras are poised to enter our social spaces in a big way. In this talk, I will investigate what wearable cameras can tell us both about the person wearing the cameras and the people they interact with. In the first part, I will present a method to reconstruct the motion of a person from cameras mounted on different limbs of the person, with the goal of taking motion capture out of the lab or studio.

Message Ferrying (MF) is a technique used to deliver data in wireless and mobile networks that are either sparse or intermittently connected. The scheme utilizes a set of mobile nodes called message ferries that take responsibility for carrying messages within the network. I will first describe the context in which the MF scheme is relevant. I will then discuss the problems that one faces in the design and operation of networks using message ferries and our approach to solving them.

Physarum is a slime mold. It was observed over the past 10 years that the mold is able to solve shortest path problems and to construct good Steiner networks. In a nutshell, the shortest path experiment is as follows: A maze is built and the mold is made to cover the entire maze. Food is then provided at two positions s and t and the evolution of the slime is observed. Over time, the slime retracts to the shortest s-t-path. A mathematical model of the slime's dynamic behavior was proposed in 2007 by Tero-Kobayashi Nakagaki. Extensive computer simulations of the mathematical model confirm the experimental findings... Room 1213 Read more... Advanced Technologies for learning outcome-based education and employability Thu, 03/01/2012 - 12:00

Learning based uniquely on input will not respond adequately to future challenges for individuals, society or the economy. The trend is to rely, increasingly, on the identification of learning outcomes and competences. In this event, we will present current international projects that focus on the identification, management and sharing of individual learning outcomes and competences, open educational resources and learning opportunities.

Typical location determination systems, for example the satellite-based Global Positioning System (GPS), require the presence of a physical device that is attached to the object that is being tracked. In addition, they usually require the tracked device to participate actively in the localization process by running part of the localization algorithm or by extracting the physical signals used in communication, such as signal strength or time. In this talk, we introduce the concept of Device-free Passive (DfP) localization for wireless networks.

The Global Technology Outlook (GTO) is IBM Research’s vision of the future for information technology (IT) and its impact on industries that use IT. This annual exercise highlights emerging software, hardware, and services technology trends that are expected to significantly impact the IT sector in the next 3-7 years. In particular, the GTO identifies technologies that may be disruptive to an existing business, have the potential to create new opportunity, and can provide new business value to our customers. The 2011 GTO is set to build not only on its 29 predecessors, but the 100 years of IBM innovation.... Carnegie Mellon Qatar room 1213 Read more... Data-Intensive Scalable Computing: Programming Models Wed, 02/01/2012 - 12:00

Data-Intensive Scalable Computing (DISC) systems provide computing resources that enable the analysis of massive data sets. They have found widespread use in Internet companies, but they also have the potential to greatly advance areas such as astronomy and biology. Programs for DISC systems must be written in ways that allow them to be executed in a loosely-coupled asynchronous environment, such as the Map/Reduce framework pioneered by Google. Although Map/Reduce has surprisingly broad applicability, a richer set of programming languages and models is required to realize the full potential of DISC.

Web search engines have become fixtures in our society, but few people realize that they are actually publicly accessible supercomputing systems, where a single query can unleash the power of several hundred processors operating on a data set of over 200 terabytes. With Internet search, computing has risen to entirely new levels of scale, especially in terms of the sizes of the data sets involved. Google and its competitors have created a new class of large-scale computer systems, which we label "Data- Intensive Scalable Computer" (DISC) systems.