Browsing by Author "Seref, Onur"

Now showing 1 - 9 of 9
  • Thumbnail Image
    Behavioral Logistics and Fatigue Management in Vehicle Routing and Scheduling Problems
    Bowden, Zachary E. (Virginia Tech, 2016-05-03)
    The vehicle routing problem (VRP), is a classic optimization problem that aims to determine the optimal set of routes for a fleet of vehicles to meet the demands of a set of customers. The VRP has been studied for many decades and as such, there are many variants and extensions to the original problem. The research presented here focuses on two different types of vehicle routing and scheduling planning problems: car shipping and fatigue-aware scheduling. In addition to modeling and solving the car shipping problem, this research presents a novel way for ways in which drivers can describe their route preferences in a decision support system. This work also introduces the first fatigue-aware vehicle scheduling problem called the Truck Driver Scheduling Problem with Fatigue Management (TDSPFM). The TDSPFM is utilized to produce schedules that keep the drivers more alert than existing commercial vehicle regulations. Finally, this work analyzes the effect of the starting alertness level on driver alertness for the remainder of the work week and examines a critical shortcoming in existing regulations.
  • Thumbnail Image
    Disruption Information, Network Topology and Supply Chain Resilience
    Li, Yuhong (Virginia Tech, 2017-07-17)
    This dissertation consists of three essays studying three closely related aspects of supply chain resilience. The first essay is "Value of Supply Disruption Information and Information Accuracy", in which we examine the factors that influence the value of supply disruption information, investigate how information accuracy influences this value, and provide managerial suggestions to practitioners. The study is motivated by the fact that fully accurate disruption information may be difficult and costly to obtain and inaccurate disruption information can decrease the financial benefit of prior knowledge and even lead to negative performance. We perform the analysis by adopting a newsvendor model. The results show that information accuracy, specifically information bias and information variance, plays an important role in determining the value of disruption information. However, this influence varies at different levels of disruption severity and resilience capacity. The second essay is "Quantifying Supply Chain Resilience: A Dynamic Approach", in which we provide a new type of quantitative framework for assessing network resilience. This framework includes three basic elements: robustness, recoverability and resilience, which can be assessed with respect to different performance measures. Then we present a comprehensive analysis on how network structure and other parameters influence these different elements. The results of this analysis clearly show that both researchers and practitioners should be aware of the possible tradeoffs among different aspects of supply chain resilience. The ability of the framework to support better decision making is then illustrated through a systemic analysis based on a real supply chain network. The third essay is "Network Characteristics and Supply Chain Disruption Resilience", in which we investigate the relationships between network characteristics and supply chain resilience. In this work, we first prove that investigating network characteristics can lead to a better understanding of supply chain resilience behaviors. Later we select key characteristics that play a critical role in determining network resilience. We then construct the regression and decision tree models of different supply chain resilience measures, which can be used to estimate supply chain network resilience given the key influential characteristics. Finally, we conduct a case study to examine the estimation accuracy.
  • Thumbnail Image
    Exploratory and Empirical Analysis of E-Marketplaces for Truck Transportation Services Procurement
    Collignon, Stephane Eric (Virginia Tech, 2016-08-11)
    In the late 1990s, early 2000s, academic literature considered electronic marketplaces as a game changer in truck transportation services procurement. Early enthusiasm was followed by skepticism regarding e-marketplaces' usefulness and the popularity of e-marketplaces appeared to wane both in industry and in academic literature. However, recent sources argue that almost half of the freight currently transported by truck in the USA is subject to transactions conducted in e-marketplaces. This dissertation intends to fill a gap in the academic literature by showing that truck transportation e-marketplaces necessitate renewed dedicated research efforts, by exploring the strategies implemented by e-marketplaces in this specific industry and by linking these strategies to marketplaces' performance. First, transportation and non-transportation e-marketplaces are compared in chapter 2 with regard to their usage of mechanisms designed to generate trust among users. Results show that truck transportation e-marketplaces use these trust mechanisms differently than non-transportation e-marketplaces, which supports a call for research on e-marketplaces in the specific context of truck transportation services procurement. In chapter 3, a database inventorying the usage of 141 features by 208 e-marketplaces is then created to initiate the empirical exploration of these specific e-marketplaces. Thanks to that database, a new typology (a way of classifying objects based on several simultaneous classification criteria) is developed in chapter 4 that identifies three main truck transportation e-marketplace strategies (two with sub-divided into two sub-strategies). The typology provides a state of industry and puts in perspective the specificity of truck transportation e-marketplaces with regard to their structure along 11 dimensions known to the general e-marketplace literature. Finally, the link between e-marketplace strategies and performance is investigated in chapter 5. Performance is measured with three traffic metrics: number of unique visitors per day, number of page views per day, and website ranking. Results show that third-party-owned e-marketplaces that provide auction mechanisms with a fairly high level of user decision and transaction support are more successful than other e-marketplaces. This dissertation provides a picture of existing e-marketplaces for the procurement of truck transportation services, challenges components of existing theories and provides ground for further research.
  • Thumbnail Image
    Extracting the Wisdom of Crowds From Crowdsourcing Platforms
    Du, Qianzhou (Virginia Tech, 2019-08-02)
    Enabled by the wave of online crowdsourcing activities, extracting the Wisdom of Crowds (WoC) has become an emerging research area, one that is used to aggregate judgments, opinions, or predictions from a large group of individuals for improved decision making. However, existing literature mostly focuses on eliciting the wisdom of crowds in an offline context—without tapping into the vast amount of data available on online crowdsourcing platforms. To extract WoC from participants on online platforms, there exist at least three challenges, including social influence, suboptimal aggregation strategies, and data sparsity. This dissertation aims to answer the research question of how to effectively extract WoC from crowdsourcing platforms for the purpose of making better decisions. In the first study, I designed a new opinions aggregation method, Social Crowd IQ (SCIQ), using a time-based decay function to eliminate the impact of social influence on crowd performance. In the second study, I proposed a statistical learning method, CrowdBoosting, instead of a heuristic-based method, to improve the quality of crowd wisdom. In the third study, I designed a new method, Collective Persuasibility, to solve the challenge of data sparsity in a crowdfunding platform by inferring the backers' preferences and persuasibility. My work shows that people can obtain business benefits from crowd wisdom, and it provides several effective methods to extract wisdom from online crowdsourcing platforms, such as StockTwits, Good Judgment Open, and Kickstarter.
  • Thumbnail Image
    Firms' Resilience to Supply Chain Disruptions
    Baghersad, Milad (Virginia Tech, 2018-07-16)
    This dissertation consists of three papers related to firms' resiliency to supply chain disruptions. The first paper seeks to evaluate the effects of supply chain disruptions on firms' performance by using a recent dataset of supply chain disruptions. To this end, we analyzed operating and stock market performances of over 300 firms that experienced a supply chain disruption during 2005 to the end of 2014. The results show that supply chain disruptions are still associated with a significant decrease in operating income, return on sales, return on assets, sales, and a negative performance in total assets. Supply chain disruptions are also associated with a significant negative abnormal stock return on the day of the supply chain disruption announcements. These results are in line with previous findings in the literature. In the second paper, in order to provide a more detailed characterization of negative impacts of disruptions on firms' performance, we develop three complementary measures of system loss: the initial loss due to the disruption, the maximum loss, and the total loss over time. Then, we utilize the contingent resource-based view to evaluate the moderating effects of operational slack and operational scope on the relationship between the severity of supply chain disruptions and the three complementary measures of system loss. We find that maintaining certain aspects of operational slack and broadening business scope can affect these different measures of loss in different ways, although these effects are contingent on the disruptions' severity. The third paper examines relationships between the origin of supply chain disruptions, firms' past experience, and the negative impacts of supply chain disruptions on firms' performance. This third study shows that the impact of external and internal supply chain disruptions on firms' performance can be different when firms do and do not have past experience with similar events. For example, the results show that past experience significantly decreases initial loss, recovery time, and total loss over time experienced by firms after internal disruptions, although past experience may not decrease initial loss, recovery time, and total loss over time in the case of external disruptions.
  • Thumbnail Image
    Human Learning-Augmented Machine Learning Frameworks for Text Analytics
    Xia, Long (Virginia Tech, 2020-05-18)
    Artificial intelligence (AI) has made astonishing breakthroughs in recent years and achieved comparable or even better performance compared to humans on many real-world tasks and applications. However, it is still far from reaching human-level intelligence in many ways. Specifically, although AI may take inspiration from neuroscience and cognitive psychology, it is dramatically different from humans in both what it learns and how it learns. Given that current AI cannot learn as effectively and efficiently as humans do, a natural solution is analyzing human learning processes and projecting them into AI design. This dissertation presents three studies that examined cognitive theories and established frameworks to integrate crucial human cognitive learning elements into AI algorithms to build human learning–augmented AI in the context of text analytics. The first study examined compositionality—how information is decomposed into small pieces, which are then recomposed to generate larger pieces of information. Compositionality is considered as a fundamental cognitive process, and also one of the best explanations for humans' quick learning abilities. Thus, integrating compositionality, which AI has not yet mastered, could potentially improve its learning performance. By focusing on text analytics, we first examined three levels of compositionality that can be captured in language. We then adopted design science paradigms to integrate these three types of compositionality into a deep learning model to build a unified learning framework. Lastly, we extensively evaluated the design on a series of text analytics tasks and confirmed its superiority in improving AI's learning effectiveness and efficiency. The second study focused on transfer learning, a core process in human learning. People can efficiently and effectively use knowledge learned previously to solve new problems. Although transfer learning has been extensively studied in AI research and is often a standard procedure in building machine learning models, existing techniques are not able to transfer knowledge as effectively and efficiently as humans. To solve this problem, we first drew on the theory of transfer learning to analyze the human transfer learning process and identify the key elements that elude AI. Then, following the design science paradigm, a novel transfer learning framework was proposed to explicitly capture these cognitive elements. Finally, we assessed the design artifact's capability to improve transfer learning performance and validated that our proposed framework outperforms state-of-the-art approaches on a broad set of text analytics tasks. The two studies above researched knowledge composition and knowledge transfer, while the third study directly addressed knowledge itself by focusing on knowledge structure, retrieval, and utilization processes. We identified that despite the great progress achieved by current knowledge-aware AI algorithms, they are not dealing with complex knowledge in a way that is consistent with how humans manage knowledge. Grounded in schema theory, we proposed a new design framework to enable AI-based text analytics algorithms to retrieve and utilize knowledge in a more human-like way. We confirmed that our framework outperformed current knowledge-based algorithms by large margins with strong robustness. In addition, we evaluated more intricately the efficacy of each of the key design elements.
  • Thumbnail Image
    Medium is the Message: Unraveling the Social Media Platforms' Effects on Communication and Opinions
    Eroglu, Derya Ipek (Virginia Tech, 2023-08-03)
    The number of social media platforms (SMP hereinafter) is ever-increasing, and all of these platforms compete for more attention and content. The global social media market is expected to grow to $223.11 billion in 2022 (Social Media Global Market Report, 2022). In an era characterized by the meteoric rise and evolution of Social Media Platforms (SMPs), understanding the interplay between platform features and user behaviors is both critical and complex. In this dissertation, we aim to elucidate the relationship between SMPs and society, with the ultimate objective of fostering a healthier social media ecosystem. This dissertation is comprised of two incisive essays, both of which are underpinned by robust theoretical frameworks. The first essay adopts an expansive purview of the SMP ecosystem. Grounded in Uses and Gratifications Theory and media studies, it establishes a user typology based on the previous typologies and examines the interaction between user motives, SMP scores, and SMP features. Employing the Analytic Hierarchy Process, a sophisticated decision-making tool, the study quantifies utility-driven choices across platforms. A notable revelation is the heterogeneity in user reliance on SMP features, contingent upon their underlying motives. This essay offers critical insights into the multifaceted nature of SMP utilization. The second essay focuses specifically on Reddit's ChangeMyView community, an exemplar of constructive discourse in the SMP environment. It constructs a theoretical model premised on the Elaboration Likelihood Model and the concept of pre-suasion, and utilizes a mixed-methods approach to explore the persuasive strategies using Content Analysis. We also utilize ChatGPT in the Content Analysis to corroborate our inferences. The findings confirm our theorization about the role of the Delta reward system in fostering reflective engagement with content, which leads to informed opinion formation. Collectively, with these essays, we aim to provide extensive insights into the dynamic interplay between SMPs and users. Both essays hold significant implications for research community, SMP decision-makers, organizations that use SMPs, and a broader audience interested in optimizing their social media repertoire. Through a theory-driven and empirical lens, employing several epistemologies, this dissertation provides a holistic depiction of the SMP ecosystem.
  • Thumbnail Image
    Simulating Protein Conformations through Global Optimization
    Mucherino, Antonio; Seref, Onur; Pardalos, Panos M. (2008-11-19)
    Many researches have been working on the protein folding problem from more than half century. Protein folding is indeed one of the major unsolved problems in science. In this work, we discuss a model for the simulation of protein conformations. This simple model is based on the idea of imposing few geometric requirements on chains of atoms representing the backbone of a protein conformation. The model leads to the formulation of a global optimization problem, whose solutions correspond to conformations satisfying the desired requirements. The global optimization problem is solved by the recently proposed Monkey Search algorithm. The simplicity of the optimization problem and the effectiveness of the used meta-heuristic search allowed the simulation of a large set of high-quality conformations. We show that, even though only few geometric requirements are imposed, some of the simulated conformation results to be similar (in terms of RMSD) to conformations real proteins actually have in nature.
  • Thumbnail Image
    Text Analytics for Customer Engagement in Social Media
    Gruss, Richard J. (Virginia Tech, 2018-04-25)
    Businesses have recognized that customers provide value to the firm beyond transactions, and leveraging this value through relationships in social media is a new area of interest for both academics and practitioners. Recent research has investigated how businesses can best manage their online presence on platforms not fully under their control, such as Facebook, YouTube, Instagram, TripAdvisor, and Yelp, among others. This dissertation extends the literature of customer engagement in social media through four contributions. First, we propose a framework that foregrounds the textual artifacts involved in online communication. Second, we develop a novel method for discovering the elements of successful Business to Customer (B2C) messages in online communities. Third, we propose a method, validated through experimentation, for finding critical product feedback in Customer to Customer (C2C) communications. Finally, we demonstrate that a set of novel numerical features can enhance the discovery of product defect mentions in C2C communications. We conclude by proposing a research agenda suggested by the framework that will further enhance our understanding of the complex customer interactions that characterize business in the era of social media.