Consumer-generated reviews play an important role in online purchase decisions for many consumers. However, the quality and helpfulness of online reviews varies significantly. In addition, the helpfulness of different consumer-generated reviews is not disclosed to consumers unless they carefully analyze the overwhelming number of available contents. Therefore, it is of vital importance to develop predictive models that can evaluate online product reviews efficiently and then display the most useful reviews to consumers, in order to assist them in making purchase decisions. This thesis examines the problem of building computational models for predicting whether a consumer-generated review is helpful based on consumers' online votes on other reviews (where a consumer's vote on a review is either HELPFUL or UNHELPFUL), with the aim of suggesting the most suitable products and vendors to consumers.In particular, we propose in this thesis three different helpfulness prediction approaches for consumer-generated reviews. Our entropy-based approach is relatively simple and suitable for applications requiring simple recommendation engine with fully-voted reviews. However, our entropy-based approach, as well as the existing approaches, lack a general framework and are all limited to utilizing fully-voted reviews. We therefore present a probabilistic helpfulness prediction framework to overcome these limitations. To demonstrate the versatility and flexibility of this framework, we propose an EM-based model and a logistic regression-based model. We show that the EM-based model can utilize reviews voted by a very small number of voters as the training set, and the logistic regression-based model is suitable for real-time helpfulness predicting of consumer-generated reviews. To our best knowledge, this is the first framework for modeling review helpfulness and measuring the goodness of models. Although this thesis primarily considers the problem of review helpfulness prediction, the presented probabilistic methodologies are, in general, applicable for developing recommender systems that make recommendation based on other forms of user-generated contents.

This thesis is concerned with the channel assignment (CA) problem in multi-channel multi-interface wireless mesh networks (M2WNs). First, for M2WNs with general topologies, we rigorously demonstrate using the combinatorial principle of inclusion/exclusion that the CA solution space can be quantified, indicating that its cardinality is greatly influenced by the number of radio interfaces installed on each router. Based on this analysis, a novel scheme is developed to construct a new reduced search space, represented by a lattice structure, that is searched more efficiently for a CA solution. The elements in the reduced lattice-based space, labeled Solution Structures (SS), represent groupings of feasible CA solutions satisfying the radio constraints at each node. Two algorithms are presented for searching the lattice structure. The first is a greedy algorithm that finds a good SS in polynomial time, while the second provides a user-controlled depthfirst search for the optimal SS. The obtained SS is used to construct an unconstrained weighted graph coloring problem which is then solved to satisfy the soft interference constraints. For the special class of full M2WNs (fM2WNs), we show that an optimal CA solution can only be achieved with a certain number of channels; we denote this number as the characteristic channel number and derive upper and lower bounds for that number as a function of the number of radios per router. Furthermore, exact values for the required channels for minimum interference are obtained when certain relations between the number of routers and the radio interfaces in a given fM2WN are satisfied. These bounds are then employed to develop closed-form expressions for the minimum channel interference that achieves the maximum throughput for uniform traffic on all communication links. Accordingly, a polynomial-time algorithm to find a near-optimal solution for the channel assignment problem in fM2WN is developed. Experimental results confirm the obtained theoretical results and demonstrate the performance of the proposed schemes.

The goal of this thesis is to find provably correct methods for detecting conflicts between XACML rules. A conflict occurs when one rule permits a request and another denies that same request. As XACML deals with access control, we can help prevent unwanted access by verifying that it contains rules that do not have unintended conflicts. In order to help with this, we propose an algorithm to find these conflicts then use the Coq Proof Assistant to prove correctness of this algorithm. The algorithm takes a rule set specified in XACML and returns a list of pairs of indices denoting which rules conflict. It is then up to the policy writer to see if the conflicts are intended, or if they need modifying. Since we will prove that this algorithm is sound and complete, we can be assured that the list we obtain is complete and only contains true conflicts.

No description available.

No description available.

No description available.

In this thesis we present a novel algorithm, <I>HyperCleaning*</I>, for effectively inferring phylogenetic trees. The method is based on the quartet method paradigm and is guaranteed to recover the best supported edges of the underlying phylogeny based on the witness quartet set. This is performed efficiently using a collapsing mechanism that employs memory/time tradeoff to ensure no loss of information. This enables <I>HyperCleaning*</I> to solve the relaxed version of the Maximum-Quartet-Consistency problem feasibly, thus providing a valuable tool for inferring phylogenies using quartet based analysis.

We propose a family of search directions based on primal-dual entropy in the context of interior point methods for linear programming. This new family contains previously proposed search directions in the context of primal-dual entropy. We analyze the new family of search directions by studying their primal-dual affine-scaling and constant-gap centering components. We then design primal-dual interior-point algorithms by utilizing our search directions in a homogeneous and self-dual framework. We present iteration complexity analysis of our algorithms and provide the results of computational experiments on NETLIB problems.

With the advent of life critical and real-time services such as remote operations over satellite, e-health etc, providing the guaranteed minimum level of services at every ground terminal of the satellite communication system has gained utmost priority. Ground terminals and the hub are not equipped with the required intelligence to predict and react to inclement and dynamic weather conditions on its own. The focus of this thesis is to develop intelligent algorithms that would aid in adaptive management of the quality of service at the ground terminal and the gateway level. This is done to adapt both the ground terminal and gateway to changing weather conditions and to attempt to maintain a steady throughput level and Quality of Service (QoS) requirements on queue delay, jitter, and probability of loss of packets. The existing satellite system employs the First-In-First-Out routing algorithm to control congestion in their networks. This mechanism is not equipped with adequate ability to contend with changing link capacities, a common result due to bad weather and faults and to provide different levels of prioritized service to the customers that satisfies QoS requirements. This research proposes to use the reported strength of fuzzy logic in controlling highly non-linear and complex system such as the satellite communication network. The proposed fuzzy based model when integrated into the satellite gateway provides the needed robustness to the ground terminals to comprehend with varying levels of traffic and dynamic impacts of weather.

CPU's are unreliable: at any point in a computation, a bit may be altered with some (small) probability. This probability may seem negligible, but for large calculations (i.e., months of CPU time), the likelihood of an error being introduced becomes increasingly significant. Relying on this fact, this thesis defines a statistical measure called robustness, and measures the robustness of several number-theoretic and algebraic algorithms. Consider an algorithm A that implements function f, such that f has range O and algorithm A has range O' where O⊆O'. That is, the algorithm may produce results which are not in the possible range of the function. Specifically, given an algorithm A and a function f, this thesis classifies the output of A into one of three categories: 1. Correct and feasible -- the algorithm computes the correct result, 2. Incorrect and feasible -- the algorithm computes an incorrect result and this output is in O, 3. Incorrect and infeasible -- the algorithm computes an incorrect result and output is in O'\O. Using probabilistic measures, we apply this classification scheme to quantify the robustness of algorithms for computing primality (i.e., the Lucas-Lehmer and Pepin tests), group order and quadratic residues. Moreover, we show that typically, there will be an "error threshold" above which the algorithm is unreliable (that is, it will rarely give the correct result).