Computer Sciences Commons^™

Do Wikipedians Follow Domain Experts? A Domain-Specific Study On Wikipedia Contribution, Yi Zhang, Aixin Sun, Anwitaman Datta, Kuiyu Chang, Ee Peng Lim

Research Collection School Of Computing and Information Systems

Wikipedia is one of the most successful online knowledge bases, attracting millions of visits daily. Not surprisingly, its huge success has in turn led to immense research interest for a better understanding of the collaborative knowledge building process. In this paper, we performed a (terrorism) domain-specific case study, comparing and contrasting the knowledge evolution in Wikipedia with a knowledge base created by domain experts. Specifically, we used the Terrorism Knowledge Base (TKB) developed by experts at MIPT. We identified 409 Wikipedia articles matching TKB records, and went ahead to study them from three aspects: creation, revision, and link evolution. We …

Designing The Dynamics Of Service Innovations, Arcot Desai Narasimhalu

Research Collection School Of Computing and Information Systems

There is no serious tool available to design service innovations even as it is gaining in important attention from the academic and industrial worlds. This paper presents a method that is specifically developed to help service innovators plan and design their innovations. The method recognizes the dependencies that exist across a service provider, customers and suppliers and help identify potential inconsistencies in the design of service innovations.

Stevent: Spatio-Temporal Event Model For Social Network Discovery, Hady W. Lauw, Ee Peng Lim, Hwee Hwa Pang, Teck-Tim Tan

Research Collection School Of Computing and Information Systems

Spatio-temporal data concerning the movement of individuals over space and time contains latent information on the associations among these individuals. Sources of spatio-temporal data include usage logs of mobile and Internet technologies. This article defines a spatio-temporal event by the co-occurrences among individuals that indicate potential associations among them. Each spatio-temporal event is assigned a weight based on the precision and uniqueness of the event. By aggregating the weights of events relating two individuals, we can determine the strength of association between them. We conduct extensive experimentation to investigate both the efficacy of the proposed model as well as the …

Efficient Processing Of Exact Top-K Queries Over Disk-Resident Sorted Lists, Hwee Hwa Pang, Xuhua Ding, Baihua Zheng

Research Collection School Of Computing and Information Systems

The top-k query is employed in a wide range of applications to generate a ranked list of data that have the highest aggregate scores over certain attributes. As the pool of attributes for selection by individual queries may be large, the data are indexed with per-attribute sorted lists, and a threshold algorithm (TA) is applied on the lists involved in each query. The TA executes in two phases--find a cut-off threshold for the top-k result scores, then evaluate all the records that could score above the threshold. In this paper, we focus on exact top-k queries that involve monotonic linear …

Simulating An Airborne Lidar Bathymetry (Alb) System, Shachak Pe'eri, Amaresh M. Kumar, Brian R. Calder

Center for Coastal and Ocean Mapping

This study’s focus is on the horizontal and vertical uncertainties associated with ALB measurements due to scattering through the water column. A lidar simulator was constructed and we present its design and preliminary results.

Parallel Active Learning: Eliminating Wait Time With Minimal Staleness, Paul Felt, Robbie Haertel, Eric K. Ringger, Kevin Seppi

Faculty Publications

A practical concern for Active Learning (AL) is the amount of time human experts must wait for the next instance to label. We propose a method for eliminating this wait time independent of specific learning and scoring algorithms by making scores always available for all instances, using old (stale) scores when necessary. The time during which the expert is annotating is used to train models and score instances–in parallel–to maximize the recency of the scores. Our method can be seen as a parameterless, dynamic batch AL algorithm. We analyze the amount of staleness introduced by various AL schemes and then …

An Analysis Of Extreme Price Shocks And Illiquidity Among Systematic Trend Followers, Bernard Lee, Shih-Fen Cheng, Annie Koh

Research Collection Lee Kong Chian School Of Business

We construct an agent-based model to study the interplay between extreme price shocks and illiquidity in the presence of systematic traders known as trend followers. The agent-based approach is particularly attractive in modeling commodity markets because the approach allows for the explicit modeling of production, capacities, and storage constraints. Our study begins by using the price stream from a market simulation involving human participants and studies the behavior of various trend-following strategies, assuming initially that their participation will not impact the market. We notice an incremental deterioration in strategy performance as and when strategies deviate further and further from the …

A Social Transitivity-Based Data Dissemination Scheme For Opportunistic Networks, Jaesung Ku, Yangwoo Ko, Jisun An, Dongman Lee

Research Collection School Of Computing and Information Systems

A social-based routing protocol for opportunistic networks considers the direct delivery as forwarding metrics. By ignoring the indirect delivery through intermediate nodes, it misses chances to find paths that are better in terms of delivery ratio and time. To overcome this limitation, we propose to incorporate transitivity, which considers the indirect delivery through intermediate nodes, as one of the forwarding metrics. We also found that some message forwards do not improve the delivery performance. To reduce the number of these useless forwards, the proposed scheme forwards messages to an encountered node when the increase of total utility value is greater …

Satrap: Data And Network Heterogeneity Aware P2p Data-Mining, Hock Kee Ang, Vivekanand Gopalkrishnan, Anwitaman Datta, Wee Keong Ng, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

Distributed classification aims to build an accurate classifier by learning from distributed data while reducing computation and communication cost A P2P network where numerous users come together to share resources like data content, bandwidth, storage space and CPU resources is an excellent platform for distributed classification However, two important aspects of the learning environment have often been overlooked by other works, viz., 1) location of the peers which results in variable communication cost and 2) heterogeneity of the peers' data which can help reduce redundant communication In this paper, we examine the properties of network and data heterogeneity and propose …

Otl: A Framework Of Online Transfer Learning, Peilin Zhao, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

In this paper, we investigate a new machine learning framework called Online Transfer Learning (OTL) that aims to transfer knowledge from some source domain to an online learning task on a target domain. We do not assume the target data follows the same class or generative distribution as the source data, and our key motivation is to improve a supervised online learning task in a target domain by exploiting the knowledge that had been learned from large amount of training data in source domains. OTL is in general challenging since data in both domains not only can be different in …

Player Performance Prediction In Massively Multiplayer Online Role-Playing Games (Mmorpgs), Kyong Jin Shim, Richa Sharan, Jaideep Srivastava

Research Collection School Of Computing and Information Systems

In this study, we propose a comprehensive performance management tool for measuring and reporting operational activities of game players. This study uses performance data of game players in EverQuest II, a popular MMORPG developed by Sony Online Entertainment, to build performance prediction models forgame players. The prediction models provide a projection of player’s future performance based on his past performance, which is expected to be a useful addition to existing player performance monitoring tools. First, we show that variations of PECOTA [2] and MARCEL [3], two most popular baseball home run prediction methods, can be used for game player performance …

Z-Sky: An Efficient Skyline Query Processing Framework Based On Z-Order, Ken C. K. Lee, Wang-Chien Lee, Baihua Zheng, Huajing Li, Yuan Tian

Research Collection School Of Computing and Information Systems

Given a set of data points in a multidimensional space, a skyline query retrieves those data points that are not dominated by any other point in the same dataset. Observing that the properties of Z-order space filling curves (or Z-order curves) perfectly match with the dominance relationships among data points in a geometrical data space, we, in this paper, develop and present a novel and efficient processing framework to evaluate skyline queries and their variants, and to support skyline result updates based on Z-order curves. This framework consists of ZBtree, i.e., an index structure to organize a source dataset and …

Visualizing And Exploring Evolving Information Networks In Wikipedia, Ee Peng Lim, Agus Trisnajaya Kwee, Nelman Lubis Ibrahim, Aixin Sun, Anwitaman Datta, Kuiyu Chang, Maureen Maureen

Research Collection School Of Computing and Information Systems

Information networks in Wikipedia evolve as users collaboratively edit articles that embed the networks. These information networks represent both the structure and content of community’s knowledge and the networks evolve as the knowledge gets updated. By observing the networks evolve and finding their evolving patterns, one can gain higher order knowledge about the networks and conduct longitudinal network analysis to detect events and summarize trends. In this paper, we present SSNetViz+, a visual analytic tool to support visualization and exploration of Wikipedia’s information networks. SSNetViz+ supports time-based network browsing, content browsing and search. Using a terrorism information network as an …

Weakly-Supervised Hashing In Kernel Space, Yadong Mu, Jialie Shen, Shuicheng Yan

Research Collection School Of Computing and Information Systems

The explosive growth of the vision data motivates the recent studies on efficient data indexing methods such as locality-sensitive hashing (LSH). Most existing approaches perform hashing in an unsupervised way. In this paper we move one step forward and propose a supervised hashing method, i.e., the LAbel-regularized Max-margin Partition (LAMP) algorithm. The proposed method generates hash functions in weakly-supervised setting, where a small portion of sample pairs are manually labeled to be “similar” or “dissimilar”. We formulate the task as a Constrained Convex-Concave Procedure (CCCP), which can be relaxed into a series of convex sub-problems solvable with efficient Quadratic-Program (QP). …

Revisiting Unpredictability-Based Rfid Privacy Models, Junzuo Lai, Robert Huijie Deng, Yingjiu Li

Research Collection School Of Computing and Information Systems

Recently, there have been several attempts in establishing formal RFID privacy models in the literature. These models mainly fall into two categories: one based on the notion of indistinguishability of two RFID tags, denoted as ind-privacy, and the other based on the unpredictability of the output of an RFID protocol, denoted as unp-privacy. Very recently, at CCS’09, Ma et al. proposed a modified unp-privacy model, referred to as unp ^′-privacy. In this paper, we first revisit the existing RFID privacy models and point out their limitations. We then propose a new RFID privacy model, denoted as …

Geodesic Graph Cut For Interactive Image Segmentation, Bryan S. Morse, Brian L. Price, Scott Cohen

Faculty Publications

Interactive segmentation is useful for selecting objects of interest in images and continues to be a topic of much study. Methods that grow regions from foreground/background seeds, such as the recent geodesic segmentation approach, avoid the boundary-length bias of graph-cut methods but have their own bias towards minimizing paths to the seeds, resulting in increased sensitivity to seed placement. The lack of edge modeling in geodesic or similar approaches limits their ability to precisely localize object boundaries, something at which graph-cut methods generally excel. This paper presents a method for combining geodesicdistance information with edge information in a graphcut optimization …

Simultaneous Foreground, Background, And Alpha Estimation For Image Matting, Bryan S. Morse, Brian L. Price, Scott Cohen

Faculty Publications

Image matting is the process of extracting a soft segmentation of an object in an image as defined by the matting equation. Most current techniques focus largely on computing the alpha values of unknown pixels and treat computation of the foreground and background colors as an afterthought, if at all. However, for many applications, such as compositing an object into a new scene or deleting an object from the scene, the foreground and background colors are vital for an acceptable answer. We propose a method of solving for the foreground, background, and alpha of an unknown region in an image …

Mining Antagonistic Communities From Social Networks, Kuan Zhang, David Lo, Ee Peng Lim

Research Collection School Of Computing and Information Systems

During social interactions in a community, there are often sub-communities that behave in opposite manner. These antagonistic sub-communities could represent groups of people with opposite tastes, factions within a community distrusting one another, etc. Taking as input a set of interactions within a community, we develop a novel pattern mining approach that extracts for a set of antagonistic sub-communities. In particular, based on a set of user specified thresholds, we extract a set of pairs of sub-communities that behave in opposite ways with one another. To prevent a blow up in these set of pairs, we focus on extracting a …

On Trustworthiness Of Cpu Usage Metering And Accounting, Mei Liu, Xuhua Ding

Research Collection School Of Computing and Information Systems

In the envisaged utility computing paradigm, a user taps a service provider’s computing resources to accomplish her tasks, without deploying the needed hardware and software in her own IT infrastructure. To make the service profitable, the service provider charges the user based on the resources consumed. A commonly billed resource is CPU usage. A key factor to ensure the success of such a business model is the trustworthiness of the resource metering scheme. In this paper, we provide a systematic study on the trustworthiness of CPU usage metering. Our results show that the metering schemes in commodity operating systems should …

Optimization Algorithms For Site-Directed Protein Recombination Experiment Planning, Wei Zheng

Dartmouth College Ph.D Dissertations

Site-directed protein recombination produces improved and novel protein variants by recombining sequence fragments from parent proteins. The resulting hybrids accumulate multiple mutations that have been evolutionarily accepted together. Subsequent screening or selection identifies hybrids with desirable characteristics. In order to increase the "hit rate" of good variants, this thesis develops experiment planning algorithms to optimize protein recombination experiments. First, to improve the frequency of generating novel hybrids, a metric is developed to assess the diversity among hybrids and parent proteins. Dynamic programming algorithms are then created to optimize the selection of breakpoint locations according to this metric. Second, the trade-off …

Graph Algorithms For Nmr Resonance Assignment And Cross-Link Experiment Planning, Fei Xiong

Dartmouth College Ph.D Dissertations

The study of three-dimensional protein structures produces insights into protein function at the molecular level. Graphs provide a natural representation of protein structures and associated experimental data, and enable the development of graph algorithms to analyze the structures and data. This thesis develops such graph representations and algorithms for two novel applications: structure-based NMR resonance assignment and disulfide cross-link experiment planning for protein fold determination. The first application seeks to identify correspondences between spectral peaks in NMR data and backbone atoms in a structure (from x-ray crystallography or homology modeling), by computing correspondences between a contact graph representing the structure …

Variance Reduction Techniques For Estimating Quantiles And Value-At-Risk, Fang Chu

Dissertations

Quantiles, as a performance measure, arise in many practical contexts. In finance, quantiles are called values-at-risk (VARs), and they are widely used in the financial industry to measure portfolio risk. When the cumulative distribution function is unknown, the quantile can not be computed exactly and must be estimated. In addition to computing a point estimate for the quantile, it is important to also provide a confidence interval for the quantile as a way of indicating the error in the estimate. A problem with crude Monte Carlo is that the resulting confidence interval may be large, which is often the case …

New Data Structures, Models, And Algorithms For Real-Time Resource Management, Xinfa Hu

Dissertations

Real-time resource management is the core and critical task in real-time systems. This dissertation explores new data structures, models, and algorithms for real-time resource management.

At first, novel data structures, i.e., a class of Testing Interval Trees (TITs), are proposed to help build efficient scheduling modules in real-time systems. With a general data structure, i.e., the TIT* tree, the average costs of the schedulability tests in a wide variety of real-time systems can be reduced. With the Testing Interval Tree for Vacancy analysis (TIT-V), the complexities of the schedulability tests in a class of parallel/distributed real-time systems can be effectively …

Type-1 Diabetes Risk Prediction Using Multiple Kernel Learning, Paras Garg

Theses

This thesis presents an analysis of multiple kernel learning (MKL) for type-1 diabetes risk prediction. MKL combines different models and representation of data to find a linear combination of these representations of the data. MKL has been successfully been implemented in image detection, splice site detection, ribosomal and membrane protein prediction, etc. In this thesis, this method was applied for Genome-wide association study (GWAS) for classifying cases and controls.

This thesis has shown that combined kernel does not perform better than the individual kernels and that MKL does not select the best model for this problem. Also, the effect of …

Inquiry In Support Of The Knowledge Sharing Life Cycle Within A Higher Education Database Practicum: A Case Study At Regis University, Martha Jorgensen

Regis University Student Publications (comprehensive collection)

The Regis University database practicum provides an opportunity for students to get hands-on experience with Database Administrator (DBA) tasks as they support the virtual lab environment used by students enrolled in database courses at the university. The student DBA team is new each semester and has a short time to become familiar with the environment they are supporting and the tools they will be using. Inefficient communication and organization delay the resolution of production issues. This study analyzed the current knowledge sharing culture, technologies, and processes of the database practicum. The goal of the study was to determine accessibility of …

Scalable Object Recognition Using Hierarchical Quantization With A Vocabulary Tree, David Nistér, Henrik Stewénius

Center for Visualization and Virtual Environments Faculty Patents

An image retrieval technique employing a novel hierarchical feature/descriptor vector quantizer tool—‘vocabulary tree’, of sorts comprising hierarchically organized sets of feature vectors—that effectively partitions feature space in a hierarchical manner, creating a quantized space that is mapped to integer encoding. The computerized implementation of the new technique(s) employs subroutine components, such as: A trainer component of the tool generates a hierarchical quantizer, Q, for application/use in novel image-insertion and image-query stages. The hierarchical quantizer, Q, tool is generated by running k-means on the feature (a/k/a descriptor) space, recursively, on each of a plurality of nodes of a resulting quantization level …

Transformation Learning: Modeling Transferable Transformations In High-Dimensional Data, Christopher R. Wilson

Theses and Dissertations

The goal of learning transfer is to apply knowledge gained from one problem to a separate related problem. Transformation learning is a proposed approach to computational learning transfer that focuses on modeling high-level transformations that are well suited for transfer. By using a high-level representation of transferable data, transformation learning facilitates both shallow transfer (intra-domain) and deep transfer (inter-domain) scenarios. Transformations can be discovered in data using manifold learning to order data instances according to the transformations they represent. For high-dimensional data representable with coordinate systems, such as images and sounds, data instances can be decomposed into small sub-instances based …

The Curious Timekeeper: Creative Thesis In Interactive Sculpture, Kate I. Schnippering

Dartmouth College Undergraduate Theses

When we interact with computers, we have set expectations about our interactive experience, operating a mouse and keyboard to elicit predictable responses on a screen. Intersecting the world of Computing with Fine Art gains us potential to innovate outside these bounds by restricting the expected performance of a computer-- setting it to a particular purpose rather than allowing it to run anyone's software. To challenge standard human-computer interaction, this work set out to create an interesting and unusual interactive experience, fully integrated into a sculpture. The approach was to design a system to form a small environment, having many components …

The Defiance College Undergraduate Major In Digital Forensic Science: Setting The Bar Higher, Gregg H. Gunsch

Annual ADFSL Conference on Digital Forensics, Security and Law

This paper provides background information to accompany the panel discussion on Curriculum Design and Implementation in Computer Forensics Education. It is specifically focused on the content and delivery of Defiance College’s undergraduate (B.S.) program majoring in Digital Forensic Science (DFS). The genesis and evolution of the Defiance College DFS program are described, along with its successes, challenges and known opportunities for improvement. The desired outcomes of the panel discussion include articulating the necessary components of an undergraduate program, refining expectations of knowledge and skills required of students upon graduation, and suggesting strategies for achieving those expectations despite inevitable resource limitations …

Computer Forensics For Graduate Accountants: A Motivational Curriculum Design Approach, Grover S. Kearns

Annual ADFSL Conference on Digital Forensics, Security and Law

Computer forensics involves the investigation of digital sources to acquire evidence that can be used in a court of law. It can also be used to identify and respond to threats to hosts and systems. Accountants use computer forensics to investigate computer crime or misuse, theft of trade secrets, theft of or destruction of intellectual property, and fraud. Education of accountants to use forensic tools is a goal of the AICPA (American Institute of Certified Public Accountants). Accounting students, however, may not view information technology as vital to their career paths and need motivation to acquire forensic knowledge and skills. …