Recommendation systems have become extremely common in recent years due to the ubiquity of information across various applications. Online entertainment (e.g., Netflix), E-commerce (e.g., Amazon, Ebay) and publishing services such as Google News are all examples of services which use recommender systems. Recommendation systems are rapidly evolving in these years, but these methods have fallen short in coping with several emerging trends such as likes or votes on reviews. In this work we have proposed a new method based on collaborative filtering by considering other users' feedback on each review. To validate our approach we have used Yelp data set with more than 335,000 product and service category ratings and 70,817 real users. We present our results using comparative analysis with other well-known recommendation systems for particular categories of users and items. / Graduate / 0984 / 0800 / yheshmat@uvic.ca

Previous research has established that punctuation can be used to communicate nuances of meaning in online writing (McAndrew & De Jonge, 2011). Punctuation, considered a computer mediated communication (CMC) cue, expresses tone and emotion and disambiguates an author's intention (Vandergriff, 2013). Quotation marks as CMC cues can serve pragmatic functions and have been understudied. Some of these functions have been generally described (Predelli, 2003). However, no corpus study has specifically focused on the pragmatic uses of quotations in online text. Consumer reviews, a genre of online text, can directly impact business profits and influence customers' purchasing decisions (Floyd, Freling, Alhoqail, Cho & Freling, 2014). Businesses are investing in sentiment analysis to gauge their target market's opinions (Salehan & Kim, 2016). Sentiment analysis is the computerized appraisal of a text to determine whether its author is expressing a positive or negative opinion (Novak, Smailovic, Sluban & Mozetic, 2015). Sentiment analysis programs are still limited and could be improved in accuracy. Most programs rely on lexicons of words given a pre-determined polarity value (positive or negative) out of context (Novak et al., 2015). However, context is crucial to communication, and sentiment analysis programs could incorporate a better variety of contextual linguistic features to improve their accuracy. Quotations used for pragmatic communication is such a feature. This study discovered seven pragmatic quotation uses in a 2014 Yelp review corpus: Collective Knowledge, Non-standard, Grammatical, Non-literal, Narrative, Idiolect, and Emphasis. An ANOVA and Tukey HSD test were performed, and the results were significant. Pragmatic category accounted for 15% of the variance in review star rating. The Collective Knowledge category and the Narrative and Non-literal categories were significantly different from each other. The Collective Knowledge category showed a correlation with positive sentiment, while the Narrative and Non-literal categories displayed a correlation with negative sentiment. These three categories are likely present in several types of online text, making them valuable for further sentiment analysis research. If these pragmatic patterns could be detected automatically, they could be used in sentiment algorithms to give a more accurate picture of author opinion.

This project examines user-generated health narratives through corpus analysis of 246 reviews posted on Midwestern Hospital’s Yelp page. Understanding how different stakeholders act and interact within online health communities models a shift in new conceptions of health, and provides evidence of health ecologies’ ability to determine patient perceptions of care. Documents produced by users in these health communities represent health narratives comprised of a user’s health experience, that user’s treatment perceptions, and the community’s perceptions of the user’s experience. Author uses corpus methods to interpret user trace data and rhetorical moves embedded in health narratives. Findings suggest that users who interact with the Yelp community produce different health narratives than less engaged users. Understanding how different stakeholders act and interact within online health communities models a shift in new conceptions of health, and provides evidence of health ecologies’ ability to determine patient perceptions of care.

In the digital age, with the accelerating pace of e-commerce, online platforms such as Amazon, Yelp, TripAdvisor, Facebook, Netflix, Uber and others have gained in prominence. Furthermore, in the wake of the COVID-19 pandemic, even businesses which were heretofore primarily brick-and-mortar have had to shift to a strong online presence in order to adapt and survive; which, while beneficial to all stakeholders, has resulted in dire challenges for the producers/service providers, platform owners, as well as consumers. In my first essay, I investigate the challenges faced by mobility as a service (MaaS) platforms such as Uber and Lyft for managing their demand and the pool of available drivers. On one hand, driver compensation issues in MaaS platforms is a highly discussed topic. On the other hand, the MaaS platforms are expanding to encompass several external businesses in search of profitability. In this chapter, I focus primarily on driver compensation issues in MaaS platforms when the platforms engage in external businesses. I find that in the majority of instances, the driver compensation reduces when the platforms get involved in external businesses; however, there are a few cases, where it leads to an increment in driver compensation, thus benefiting them. The second essay is on the impact of online reviews from digital platforms such as Yelp and TripAdvisor on business performance. Using a data set from Yelp, first, I study the interaction of average rating and number of reviews on business performance; second, how competition affects the interaction effect of the average rating and number of reviews on the focal business' performance. I find that the impact of the interaction of average rating and number of reviews on business performance is different at various levels of average ratings, and the inclusion of competition negatively influences the interaction effect of the average rating and number of reviews on the performance of the focal restaurant. In my third essay, I analyze how the interaction of supplier encroachment and consumer showrooming impacts an omnichannel retailer and her upstream manufacturer, who encroaches the downstream retailer's market with an online direct sales channel. I identify different scenarios in a covered market where either the retailer, or the manufacturer, or both will be better off. Taken together, these three essays provide valuable managerial insights for real world business problems, which will empower researchers in academia and industry managers, and help them improve their businesses and maximize their operational performance. / Business Administration/Marketing

No description available.

No description available.

We examine the antecedents of customer satisfaction in the restaurant sector, paying particular attention to perceived value and price level. Using Latent Dirichlet Allocation, we extract latent topics from the text of Yelp! reviews, then analyze the relationship between these topics and satisfaction, measured as the difference between review rating and user average review rating.

Backdoor (Trojan) attacks are a major threat to the security of deep neural network (DNN) models. They are created by an attacker who adds a certain pattern to a portion of given training dataset, causing the DNN model to misclassify any inputs that contain the pattern. These infected classifiers are called Trojan models and the added pattern is referred to as the trigger. In image domain, a trigger can be a patch of pixel values added to the images and in text domain, it can be a set of words. In this thesis, we propose Trojan-Miner (T-Miner), a defense scheme against such backdoor attacks on text classification deep learning models. The goal of T-Miner is to detect whether a given classifier is a Trojan model or not. To create T-Miner , our approach is based on a sequence-to-sequence text generation model. T-Miner uses feedback from the suspicious (test) classifier to perturb input sentences such that their resulting class label is changed. These perturbations can be different for each of the inputs. T-Miner thus extracts the perturbations to determine whether they include any backdoor trigger and correspondingly flag the suspicious classifier as a Trojan model. We evaluate T-Miner on three text classification datasets: Yelp Restaurant Reviews, Twitter Hate Speech, and Rotten Tomatoes Movie Reviews. To illustrate the effectiveness of T-Miner, we evaluate it on attack models over text classifiers. Hence, we build a set of clean classifiers with no trigger in their training datasets and also using several trigger phrases, we create a set of Trojan models. Then, we compute how many of these models are correctly marked by T-Miner. We show that our system is able to detect trojan and clean models with 97% overall accuracy over 400 classifiers. Finally, we discuss the robustness of T-Miner in the case that the attacker knows T-Miner framework and wants to use this knowledge to weaken T-Miner performance. To this end, we propose four different scenarios for the attacker and report the performance of T-Miner under these new attack methods. / M.S. / Backdoor (Trojan) attacks are a major threat to the security of predictive models that make use of deep neural networks. The idea behind these attacks is as follows: an attacker adds a certain pattern to a portion of given training dataset and in the next step, trains a predictive model over this dataset. As a result, the predictive model misclassifies any inputs that contain the pattern. In image domain this pattern that is called trigger, can be a patch of pixel values added to the images and in text domain, it can be a set of words. In this thesis, we propose Trojan-Miner (T-Miner), a defense scheme against such backdoor attacks on text classification deep learning models. The goal of T-Miner is to detect whether a given classifier is a Trojan model or not. T-Miner is based on a sequence-to-sequence text generation model that is connected to the given predictive model and determine if the predictive model is being backdoor attacked. When T-Miner is connected to the predictive model, it generates a set of words, called perturbations, and analyses these perturbations to determine whether they include any backdoor trigger. Hence if any part of the trigger is present in the perturbations, the predictive model is flagged as a Trojan model. We evaluate T-Miner on three text classification datasets: Yelp Restaurant Reviews, Twitter Hate Speech, and Rotten Tomatoes Movie Reviews. To illustrate the effectiveness of T-Miner, we evaluate it on attack models over text classifiers. Hence, we build a set of clean classifiers with no trigger in their training datasets and also using several trigger phrases, we create a set of Trojan models. Then, we compute how many of these models are correctly marked by T-Miner. We show that our system is able to detect Trojan models with 97% overall accuracy over 400 predictive models.