Incident databases provide an excellent opportunity to study the repeated situations of incidents in the process industry. The databases give an insight into the situation which led to an incident, and if studied properly can help monitor the process, equipment and chemical involved more closely, and reduce the number of incidents in the future. This study examined a subset of incidents from National Response Center’s Incident database, focusing mainly on fixed facility incidents in Harris County, Texas from 1990 to 2002. Data mining has been used in the financial and marketing arena for many decades to analyze and find patterns in large amounts of data. Realizing the limited capabilities of traditional methods of statistics, more robust techniques of data mining were applied to the subset of data and interesting patterns of chemical involved, equipment failed, component involved, etc. were found. Further, patterns obtained by data mining on the subset of data were used in modifying probabilities of failure of equipment and developing a decision support system.

Researchers have proposed frequent pattern mining algorithms that are more efficient than previous algorithms and generate fewer but more important patterns. Many techniques such as depth first/breadth first search, use of tree/other data structures, top down/bottom up traversal and vertical/horizontal formats for frequent pattern mining have been developed. Most frequent pattern mining algorithms use a support measure to prune the combinatorial search space. However, support-based pruning is not enough when taking into consideration the characteristics of real datasets. Additionally, after mining datasets to obtain the frequent patterns, there is no way to adjust the number of frequent patterns through user feedback, except for changing the minimum support. Alternative measures for mining frequent patterns have been suggested to address these issues. One of the main limitations of the traditional approach for mining frequent patterns is that all items are treated uniformly when, in reality, items have different importance. For this reason, weighted frequent pattern mining algorithms have been suggested that give different weights to items according to their significance. The main focus in weighted frequent pattern mining concerns satisfying the downward closure property. In this research, frequent pattern mining approaches with weight constraints are suggested. Our main approach is to push weight constraints into the pattern growth algorithm while maintaining the downward closure property. We develop WFIM (Weighted Frequent Itemset Mining with a weight range and a minimum weight), WLPMiner (Weighted frequent Pattern Mining with length decreasing constraints), WIP (Weighted Interesting Pattern mining with a strong weight and/or support affinity), WSpan (Weighted Sequential pattern mining with a weight range and a minimum weight) and WIS (Weighted Interesting Sequential pattern mining with a similar level of support and/or weight affinity) The extensive performance analysis shows that suggested approaches are efficient and scalable in weighted frequent pattern mining.

Thesis(Ph.D.)--Case Western Reserve University, 2010 / Title from PDF (viewed on 2009-12-30) Department of Electrical Engineering and Computer Science Includes abstract Includes bibliographical references and appendices Available online via the OhioLINK ETD Center

Thesis (Ph. D.)--University of Missouri-Columbia, 2002. / Typescript. Vita. Includes bibliographical references (leaves 134-140). Also available on the Internet.

Thesis (Ph. D.)--University of Missouri-Columbia, 2002. / Typescript. Vita. Includes bibliographical references (leaves 134-140). Also available on the Internet.

Thesis (M. Phil.)--University of Hong Kong, 2002. / Includes bibliographical references (leaves 75-77).

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2003. / Includes bibliographical references (leaves 67-69). Also available in electronic version. Access restricted to campus users.

published_or_final_version / Computer Science / Master / Master of Philosophy

Information propagation has attracted increasing attention from sociologists, marketing researchers and Information Technology entrepreneurs. With the rapid developments in online and mobile social applications like Facebook, Twitter, and LinkedIn, large-scale, high-speed and instantaneous information dissemination becomes possible, spawning tremendous opportunities for electronic commerce. It is non-trivial to make an accurate analysis on how information is propagated due to the uncertainty of human behavior and the complexity of the social environment. This dissertation is concerned with exploring models, formulations, and heuristics for the social-network-based information propagation process. It consists of three major parts: information diffusion through online social network, modeling social influence propagation, and social-network-based information spreading in opportunistic mobile networks. Firstly, I consider the problem of maximizing the influence propagation through online social networks. To solve it, I introduce a probabilistic maximum coverage problem, and propose a cluster-based heuristic and a neighbor-removal heuristic for two basic diffusion models, namely, the Linear Threshold Model and the Independent Cascade Model, respectively. Realizing that the selection of influential nodes is mainly based on the accuracy and efficiency in estimating the social influence, I build a framework of up-to-2-hop hierarchical network to approximate the spreading of social influence, and further propose a hierarchy-based algorithm to solve the influence maximization problem. Our heuristic is proved to be efficient and robust with competitive performance, low computation cost, and high scalability. The second part explores the modeling on social influence propagation. I develop an analytical model for the influence propagation process based on discrete-time Markov chains, and deduce a close-form equation to express the n-step transition probability matrix. We show that given any initial state the probability distribution of the converged network state could be easily obtained by calculating a matrix product. Finally, I study the social-network-based information spreading in opportunistic mobile networks by analyzing the opportunistic routing process. I propose three social-network-based communication pattern models and utilize them to evaluate the performance of different social-network-based routing protocols based on several human mobility traces. Moreover, I discuss the fairness evaluation in opportunistic routing, and propose a fair packet forwarding strategy which operates as a plugin for traditional social- network-based routing protocols. My strategy improves the imbalance of success rates among users while maintaining approximately the same system throughput. / published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

Not available / text