The explosive and continuous growth in the size and use of the World Wide Web is at the basis of the great interest into web usage mining techniques in both research and commercial areas. In particular, the need for predicting the user’s needs in order to improve the usability and user retention of a web site is more than evident and can be addressed by personalization. In this thesis, we introduce a new framework that takes advantage of the sophisticated association rule finding web mining technology on both dynamic user activities over a web site, such as navigational behavior, and static information, such as user profiles and web content. We also provide a novel personalization selection system which allows users to choose the most suitable profile for them in any given period of time. In order to examine the viability of our framework, we incorporate and implement it over a well designed simulation environment. Moreover, our experiment proves that our framework provides an overall better web personalization service in terms of both recommendation accuracy and user satisfaction.

The explosive and continuous growth in the size and use of the World Wide Web is at the basis of the great interest into web usage mining techniques in both research and commercial areas. In particular, the need for predicting the user’s needs in order to improve the usability and user retention of a web site is more than evident and can be addressed by personalization. In this thesis, we introduce a new framework that takes advantage of the sophisticated association rule finding web mining technology on both dynamic user activities over a web site, such as navigational behavior, and static information, such as user profiles and web content. We also provide a novel personalization selection system which allows users to choose the most suitable profile for them in any given period of time. In order to examine the viability of our framework, we incorporate and implement it over a well designed simulation environment. Moreover, our experiment proves that our framework provides an overall better web personalization service in terms of both recommendation accuracy and user satisfaction.

The explosive and continuous growth in the size and use of the World Wide Web is at the basis of the great interest into web usage mining techniques in both research and commercial areas. In particular, the need for predicting the user’s needs in order to improve the usability and user retention of a web site is more than evident and can be addressed by personalization. In this thesis, we introduce a new framework that takes advantage of the sophisticated association rule finding web mining technology on both dynamic user activities over a web site, such as navigational behavior, and static information, such as user profiles and web content. We also provide a novel personalization selection system which allows users to choose the most suitable profile for them in any given period of time. In order to examine the viability of our framework, we incorporate and implement it over a well designed simulation environment. Moreover, our experiment proves that our framework provides an overall better web personalization service in terms of both recommendation accuracy and user satisfaction. / Science, Faculty of / Computer Science, Department of / Graduate

Black box personalization systems have become ubiquitous in our daily lives. They utilize collected data about us to make critical decisions such as those related to credit approval and insurance premiums. This leads to concerns about whether these systems respect expectations of fairness and privacy. Given the black box nature of these systems, it is challenging to test whether they satisfy certain fundamental fairness and privacy properties. For the same reason, while many black box privacy enhancing technologies offer consumers the ability to defend themselves from data collection, it is unclear how effective they are. In this doctoral thesis, we demonstrate that carefully designed methods and tools that soundly and scalably discover causal effects in black box software systems are useful in evaluating personalization systems and privacy enhancing technologies to understand how well they protect fairness and privacy. As an additional defense against discrimination, this thesis also explores legal liability for ad platforms in serving discriminatory ads. To formally study fairness and privacy properties in black box personalization systems, we translate these properties into information flow instances and develop methods to detect information flow. First, we establish a formal connection between information flow and causal effects. As a consequence, we can use randomized controlled experiments, traditionally used to detect causal effects, to detect information flow through black box systems. We develop AdFisher as a general framework to perform information flow experiments scalably on web systems and use it to evaluate discrimination, transparency, and choice on Google’s advertising ecosystem. We find evidence of gender-based discrimination in employment-related ads and a lack of transparency in Google’s transparency tool when serving ads for rehabilitation centers after visits to websites about substance abuse. Given the presence of discrimination and the use of sensitive attributes in personalization systems, we explore possible defenses for consumers. First, we evaluate the effectiveness of publicly available privacy enhancing technologies in protecting consumers from data collection by online trackers. Specifically, we use a combination of experimental and observational approaches to examine how well the technologies protect consumers against fingerprinting, an advanced form of tracking. Next, we explore legal liability for an advertising platform like Google for delivering employment and housing ads in a discriminatory manner under Title VII and the Fair Housing Act respectively. We find that an ad platform is unlikely to incur liability under Title VII due to its limited coverage. However, we argue that housing ads violating the Fair Housing Act could create liability if the ad platform targets ads toward or away from protected classes without explicit instructions from the advertiser.

Personalization that uses information technology to tailor content and products/services to the preferences and tastes of individual customers has become a useful function in online relationship marketing. Many techniques have been developed, and research on personalized services has increased substantially in recent years. The objective of this research is to propose a relationship-building perspective that treats the relational benefits resulting from personalized services as the important mediated factors that affect customer loyalty. In this research, a theoretical model is proposed to include both cognitive and affective benefits in order to investigate their relative influences on customer loyalty. A controlled laboratory experiment was conducted to investigate the effect of three different personalization tactics: socialness, self-reference, and content relevance. The results showed that personalized services can contribute to customer loyalty by bringing relational benefits to customers, and the impact of affective benefits on customer loyalty is greater than that of cognitive benefits. The findings extend our existing understanding of personalization, and offer valuable information to practitioners on enhancing their website design.

The increased popularity of personal marketing and printing on demand technology has substantially changed the advertising practice. Taking direct mail (DM) as an example, DM is no longer treated as ¡§junk mail¡¨ but useful marketing tool after combining personal information with interesting advertisement, such as combining customers¡¦ credit card statements with customized advertisement based on their purchasing behavior. Personalized DM has become a brand new approach for advertising that may lead to better communication effect. In this research, we formulated a research framework based on the Advertisement Attitude Mediating Model, which measures the effect of advertisement in three aspects: attitude towards the advertisement, attitude towards the advertised product, and purchase intention. Product category (shopping goods and specialty goods) and need for cognition (high and low) were moderators. Experiment method was utilized in this research to examine the effect of personalized DM. A total of 329 college students participated the experiment and they were divided into six groups with different settings in three different degrees of personalization (personalized, half-personalized, and none-personalized) and two product categories. Among them, 235 eventually completed the experiment to provide valid data for analysis. The results include: (1) personalized information has a positive effect on consumers¡¦ attitude towards DM; (2) personalized product recommendation can improve consumers¡¦ attitude towards product and increase their purchase intention; (3) A higher degree of personalization produce the better advertising effects; (4) The effect of personalization varies in different product category. Specialty goods is more suitable for personalized DM than shopping goods; (5) The consumer¡¦s need for cognition does not affect the effect of personalized DM. Chen, Chih-Hau

The mobile era is well established and the number of smartphone users is showing exponential growth. The capability of smartphones and enabling technologies is also increasing and has opened many possibilities of personalized mobile services. The goal of personalization is to support the user by providing the right service at the rightmoment. Early focus of personalization was on content adaptations in different information systems. The new approaches of personalization are still needed for mobileservices as it is a compelling feature of mobile communication systems for both endusers and service providers.Personalization is providing a means of fulfilling users’ needs more effectively andefficiently and, consequently increasing users’ satisfaction. By providing successfulpersonalization, a high degree of user satisfaction and a pleasant user experience can beachieved. Some features of personalization can cause problems and may outweigh thebenefits of personalization.This thesis has focused on how to achieve scrutable mobile client-side personalizationwhile keeping the user’s privacy. The issue of privacy in personalization of mobileservices can be reduced by shifting the control of their personal information towards theusers. Our research goal is to understand and improve the personalization process anddevelop an architecture for scrutable mobile client-side personalization while keepingthe user s’ privacy. Moreover, there is a need to develop an evaluation framework tomeasure the effectiveness of mobile services personalization. A design science researchmethodology is adopted in this research work. More particular contributions of thethesis are as follows: C1: Identifications of the research issues and challenges in personalization of mobileservices. C2: An approach for delivering personalized mobile services. C3: Development of mobile client-side personalization architecture. C4: Development of mobile services Personalization Evaluation Model. C5: Identification of the prospects of scrutable personalization of mobile services.

Driver Assistance and Autonomous Driving features are becoming nearly ubiquitous in new vehicles. The intent of the Driver Assistant features is to assist the driver in making safer decisions. The intent of Autonomous Driving features is to execute vehicle maneuvers, without human intervention, in a safe manner. The overall goal of Driver Assistance and Autonomous Driving features is to reduce accidents, injuries, and deaths with a comforting driving experience. However, different drivers can react differently to advanced automated driving technology. It is therefore important to consider and improve the adaptability of these advances based on driver behavior. In this thesis, a human-centric approach is adopted in order to provide an enriching driving experience. The thesis investigates the natural behavior of drivers when changing lanes in terms of preferences of vehicle kinematics parameters using a real-world driving dataset collected as part of the Second Strategic Highway Research Program (SHRP2). The SHRP2 Naturalistic Driving Study (NDS) set is mined for lane change events. This work develops a way to detect reliable lane changing instances from a huge NDS dataset with more than 5,400,000 data files. The lane changing instances are distinguished from noisy and erroneous data by using machine vision lane tracking system variables such as left lane marker probability and right lane marker probability. We have shown that detected lane changing instances can be validated using only vehicle kinematics data. Kinematic vehicle parameters such as vehicle speed, lateral displacement, lateral acceleration, steering wheel angle, and lane change duration are then extracted and examined from time series data to characterize these lane-changing instances for a given driver. We have shown how these vehicle kinematic parameters change and exhibit patterns during lane change maneuvers for a specific driver. The thesis shows the limitations of analyzing vehicle kinematic parameters separately and develops a novel metric, Lane Change Dynamic Score(LCDS) that shows the collective effect of these vehicle kinematic parameters. LCDS is used to classify each lane change and thereby different driving styles. / Master of Science / The current tendency of car manufacturers is to create vehicles that will offer the user the most comfortable ride possible. The user experience is given a lot of attention to ensure it is up to par. With technological advancements, we are moving closer to an era in which automobiles perform many functions autonomously. However, different drivers may react differently to highly automated driving technologies. Therefore, adapting to different driving styles is critical to increasing the acceptance of autonomous vehicle features. In this work, we examine one of the stressful maneuvers of lane changes. The analysis of various drivers' lane-changing behaviors and the value of personalization are the main subjects of this study based on actual driving scenarios. To achieve this, we have provided an algorithm to identify occurrences of lane-changing from real driving trip data files. Following that, we investigated parameters such as lane change duration, vehicle speed, displacement, acceleration, and steering wheel angle when changing lanes. We have demonstrated the patterns and changes in these vehicle kinematic characteristics that occur when a particular driver performs lane change operations. The thesis shows the limitations of analyzing vehicle kinematic parameters separately and develops a novel metric, Lane Change Dynamic Score(LCDS) that shows the collective effect of these vehicle kinematic parameters. LCDS is used to classify each lane change and thereby different driving styles.

Bachelor Thesis in Business Administration. Bachelor of Science with Specialization in Marketing – Main Field of Study: Business Administration. School of Business and Economics at Linnaeus University, Course Code 2FE21E, 2017.  Title: Likelihood of Using Online Personalization Services: An Explanatory Study Authors: Avesta Diliwi, Christopher Ullberg and Johanna Jevinger Supervisor: Michaela Sandell Examiner: Åsa Devine Background: Online personalization is the result of the rapid technological and digital development where consumers are provided products, services and content based on their individual preferences. Various research has been conducted regarding what factors influence the utilization and acceptance of personalization but does not provide a holistic view on the unified relationship of the recurrent variables of value for personalization, concern for privacy and trust building factors towards likelihood of using online personalization services. Purpose: The purpose of this research is to explain the relationship of value for personalization, concern for privacy, and trust building factors with the likelihood of using online personalization services. Methodology: This research replicated Chellappa and Sin’s (2005) research by modifying their theoretical model and testing it in another context. An explanatory, deductive, quantitative research approach and cross-sectional research design were utilized within this research, where self-completed questionnaires were distributed online with a number of 228 valid responses collected. Findings: The findings demonstrate that the new theoretical model is significant and that it explains the likelihood of using online personalization services with 62,3%. Value for personalization and concern for privacy are considered highly significant and are thus accepted hypotheses, while trust building factors is not considered significant and therefore rejected. Conclusion: This research provides an insight into consumers’ usage decision in regards to likelihood of using personalization. It also provides a furthering on prior research in regards to a theoretical development, the modified model tested in a new context, but also in the findings in how the three independent variables affect the dependent variable. In addition, this research provides support for practitioners of online personalization services to understand which factors actually affect consumers’ usage decision, and can potentially develop strategies accordingly. Keywords: Personalization; Online Personalization Services; Likelihood of Using Online Personalization Services; Value for Personalization; Concern for Privacy; Trust Building Factors

Thesis advisor: Julie H. Mortimer / This dissertation investigates firms' strategic decisions in industries characterized by a retail sector and the subsequent welfare implications. The first chapter studies retailer assortment choices; the second investigates the effectiveness of retailer online advertising. In many industries producers reach consumers only through the retail sector. Retailer product assortment choices are crucial determinants of consumer welfare as well as retailers' and producers' profitability. Limited shelf space, an inherent characteristic of the brick-and-mortar retail sector, necessitates careful selection of product offerings. The assortment decision within a product category consists of two broad questions: "How many products to offer?" and "Which products to offer?". In sole-authored work, the first chapter focuses on the latter question and investigates the drivers and welfare consequences of retail product selections. While retailer assortment choices are primarily governed by consumers' preferences and retail sector competition, vertical contracts with producers may also influence product offerings, and, in turn, product availability in the market. From the producers' perspective, obtaining product distribution is imperative. Hence, producers frequently provide financial incentives to retailers to secure their patronage. These incentives often take the form of vendor allowances: lump-sum payments to retailers that do not directly depend on sales volume. They can take the form of slotting fees, warehousing allowances, cash discounts, allowances for damaged goods, or operating support (e.g. stocking personnel). Considering the spread of the retail sector, the impact of vertical contracts on product selections may substantially affect consumer welfare and firm profitability. Therefore, it is not surprising that vendor allowances have been the subject of policy discussion. Policy makers have raised concerns that these payments are harming disproportionately small producers and limiting consumer choice. Nevertheless, the Federal Trade Commission abstains from providing clear guidelines on the use of these payments due to unclear theoretical predictions and scarce empirical evidence. The main impediment to empirical analysis has been the proprietary nature of vertical contracts and firm costs. To overcome these data limitations, I develop a novel framework that allows me to quantify vendor allowances and analyze their effects on product selections and welfare. Using only data on retail prices, quantity sold, and retailer offerings, I estimate vendor allowances as retailers' opportunity cost of shelf space. Specifically, retailers face shelf-space limitations, hence, the opportunity cost of supplying a product is the sacrificed profits from not supplying a different product in its place. With limited assumptions on producer and retailer bargaining protocol, set estimates of vendor allowances are recovered. Additionally, by assuming that producers make take-it-or-leave-it offers, point estimates can be obtained. Lower bounds from set estimates imply that, on average, vendor allowances amount to at least 5% of retailer revenues. These results suggest that vendor allowances are likely important for retailer profitability, given that public grocery chains in the U.S. report profit margins on the order of 2-4% of revenues. To investigate the effects of these payments on product selections and welfare, I apply model estimates to simulate how market outcomes change in the absence of vendor allowances. The "what-if" experiment predicts that, absent vendor allowances, retailers fare worse, product variety is reduced as retailers replace "niche" products with "mainstream" options, but consumers are nevertheless better off. Small producers, which offer high-volume products, increase market distribution and profits, but, absent marginal cost data, consequences for large producers are uncertain. The work extends our understanding of how firms' strategic interactions in the marketplace may affect consumer welfare and firm profitability through product availability. The second chapter presents a coauthored work with Alexander Bleier and Maik Eisenbeiss that analyzes the use of online advertising personalization by an online retailer. Online advertising has become an important channel through which firms attempt to influence consumer behavior and increase sales. To improve effectiveness, firms today tailor their advertisements to individual consumers with a method called retargeting. In retargeting, firms track the shopping behaviors of individual consumers' visiting their online stores and, subsequently, deliver individualized display banner ads as consumers continue browsing the Web. While this method has gained traction in the online advertising industry, research in the field is still in its infancy. This work furthers our understanding of advertising personalization by analyzing two questions: How effective is ad personalization in attracting individual consumers back to the online store? And, do different personalization approaches have distinct impacts on consumers' engagement behaviors with the online store? To answer these research questions, we exploit unique data from a randomized field experiment conducted in cooperation with a major fashion and sporting goods retailer. This study compares the effects of online banners with very high, medium, and low degrees of content personalization. For example, very high personalization refers to ads showing consumers products that they had viewed at their previous visit to the retailer's online store. Medium personalization includes products from the most viewed category or brand of their previous visit. And low personalization delivers random products from the retailer's assortment without any connection to a consumer's previous shopping behavior. Results suggest that ads with very high personalization are more effective in bringing consumers back to the online store than the other campaigns. However, we also find that the gain in visits of very high- over medium-personalization banners stems mainly from visit with low consumer engagement. / Thesis (PhD) — Boston College, 2016. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Economics.