Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (5 results)

  • Thesis
  • Peer Reviewed
UC Merced Electronic Theses and Dissertations (2015)

Sentiment analysis and opinion mining is the field of computational study of people’s

opinion expressed in written language or text. Sentiment analysis brings together various

research areas such as natural language processing, data mining and text mining,

and is fast becoming of major importance to organizations as they integrate online commerce

into their operations.

The input of the problem is a collection of written reviews about an object. The object could be a product or service and the goal is to discover people’s opinions expressed in those written reviews. The input reviews are in the form of free text and do not have any structure (people can write whatever they like however they want). Dealing with unstructured data is a challenging problem.

Sentiment analysis can be done in different levels, and the focus of this dissertation

in on aspect-level sentiment analysis. In aspect-level sentiment analysis there are two

tasks that need to be addressed. The first task is aspect identification which is the process

of discovering those attributes of the object that people are commenting on. These

attributes of the object are called aspects. The second task is sentiment identification

which is the process of discovering people’s opinions expressed about each one of the

aspects. Aforementioned tasks can be solved in 2 separate steps or can be solved simultaneously.

Early work on aspect-level sentiment analysis would solve it in 2 steps and recent techniques based on topic models address these 2 tasks simultaneously. In this thesis an automatic framework for discovering the aspects and their corresponding sentiments is proposed. This framework first identifies the aspects and then in the next step classifies each sentence containing one of the discovered aspects into either positive, neutral or negative sentiment classes. In the subsequent chapter this framework is used to give structure to input data which does not have any structure. Also a Bayesian model is proposed for overall satisfaction that accurately predicts the overall customer satisfaction and also the significance of each discovered aspect from the contributors perspectives. Hierarchical Bayesian frameworks are powerful tools that have recently attracted a lot of attention in the machine learning community. In this dissertation a new model based on hierarchical Bayesian models is proposed to simultaneously discover aspects and sentiments. This framework is based on probabilistic topic modeling techniques.

    Cover page: Statistical Models for Aspect-Level Sentiment AnalysisCreative Commons 'BY-NC-ND' version 4.0 license
    • Article
    • Peer Reviewed
    UC Merced Previously Published Works (2014)
    Enterprises must manage their information risk as part of their larger operational risk management program. Managers must choose how to control for such information risk. This article defines the flow risk reduction problem and presents a formal model using a workflow framework. Three different control placement methods are introduced to solve the problem, and a comparative analysis is presented using a robust test set of 162 simulations. One year of simulated attacks is used to validate the quality of the solutions. We find that the math programming control placement method yields substantial improvements in terms of risk reduction and risk reduction on investment when compared to heuristics that would typically be used by managers to solve the problem. The contribution of this research is to provide managers with methods to substantially reduce information and security risks, while obtaining significantly better returns on their security investments. By using a workflow approach to control placement, which guides the manager to examine the entire infrastructure in a holistic manner, this research is unique in that it enables information risk to be examined strategically. © 2014 ACM.
      Cover page: Risk Mitigation Decisions for IT Security
      • Article
      • Peer Reviewed
      UC Merced Previously Published Works (2022)
      Innovation is a very important concern for both managers and governmental policy makers. There is an important interplay between security and technology innovation that is largely unrecognized in the literature. This research considers the case where technology innovation in the form of additional product features increases demand through greater functionality. However, the likelihood of a security breach increases with the number of product features as the features interact in unintended ways, thereby increasing the attack surface. Using a two-stage game, we demonstrate how potential demand changes (from direct risk or externalities) impact firm technology innovation strategy. The analysis shows that the type and extent of customer demand reaction has a significant impact on innovative feature development. This research identifies two potential impacts on the level of innovation that can be strategically managed - the impact of externalities on demand and industry risk - explaining how these forces alter the level of innovation in the product ecosystem. Additionally, high-security development is disincentivized, and leads to a type of competitive behavior where the opportunity window for high-security development for all firms is small.
        Cover page: How Customer Demand Reactions Impact Technology Innovation and Security
        • Article
        • Peer Reviewed
        UC Riverside Previously Published Works (2010)
        Disaster response and recovery are crucial phases of disaster management. Decision-support systems used in disaster management must cope with the complexity and uncertainty involved with the scheduling and assignment of differentially-skilled personnel and assets to specific tasks. Operational constraints—such as workload and labor requirements, precedence constraints, resource availability, and critical deadlines among others—make timely and appropriate task assignment and sequencing difficult. Failure to assign personnel in an efficient and effective manner may result in unnecessary fatalities and significant additional loss of property as well as damaging the reputation of the disaster management organizations. Therefore, this paper proposes a decision-support system for disaster response and recovery using hybrid meta-heuristics.
          Cover page: Decision support for disaster management
          • Article
          • Peer Reviewed
          UC Merced Previously Published Works (2013)
          We find evidence that customers evaluate services according to different rules depending on whether the overall service experience impression is negative or positive. The functional form by which the underlying service attributes are mapped into an overall satisfaction evaluation is compensatory in the former (negative) case and has conjunctive properties in the latter (positive) case. These findings are based on an examination of a large data set of outpatient healthcare customer satisfaction surveys. We conjecture that different cognitive processes are at work for positive versus negative service encounters.
            Cover page: Dual Rules for Service Evaluation
            Top