Product Recommendation Based on Eye Tracking Data Using Fixation Duration

E-commerce can be used to increase companies or sellers’ profits. For consumers, e-commerce can help them shop faster. The weakness of e-commerce is that there is too much product information presented in the catalog which in turn makes consumers confused. The solution is by providing product recommendations. As the development of sensor technology, eye tracker can capture user attention when shopping. The user attention was used as data of consumer interest in the product in the form of fixation duration following the Bojko taxonomy. The fixation duration data was processed into product purchase prediction data to know consumers’ desire to buy the products by using Chandon method. Both data could be used as variables to make product recommendations based on eye tracking data. The implementation of the product recommendations based on eye tracking data was an eye tracking experiment at selvahouse.com which sells hijab and women modest wear. The result was a list of products that have similarities to other products. The product recommendation method used was item-to-item collaborative filtering. The novelty of this research is the use of eye tracking data, namely the fixation duration and product purchase prediction data as variables for product recommendations. Product recommendation that produced by eye tracking data can be solution of product recommendation’s problems, namely sparsity and cold start.

Learning Personalized Product Recommendations with Customer Disengagement

Problem definition: We study personalized product recommendations on platforms when customers have unknown preferences. Importantly, customers may disengage when offered poor recommendations. Academic/practical relevance: Online platforms often personalize product recommendations using bandit algorithms, which balance an exploration-exploitation trade-off. However, customer disengagement—a salient feature of platforms in practice—introduces a novel challenge because exploration may cause customers to abandon the platform. We propose a novel algorithm that constrains exploration to improve performance. Methodology: We present evidence of customer disengagement using data from a major airline’s ad campaign; this motivates our model of disengagement, where a customer may abandon the platform when offered irrelevant recommendations. We formulate the customer preference learning problem as a generalized linear bandit, with the notable difference that the customer’s horizon length is a function of past recommendations. Results: We prove that no algorithm can keep all customers engaged. Unfortunately, classical bandit algorithms provably overexplore, causing every customer to eventually disengage. Motivated by the structural properties of the optimal policy in a scalar instance of our problem, we propose modifying bandit learning strategies by constraining the action space up front using an integer program. We prove that this simple modification allows our algorithm to perform well by keeping a significant fraction of customers engaged. Managerial implications: Platforms should be careful to avoid overexploration when learning customer preferences if customers have a high propensity for disengagement. Numerical experiments on movie recommendations data demonstrate that our algorithm can significantly improve customer engagement.

The Effects of Price Rank on Clicks and Conversions in Product List Advertising on Online Retail Platforms

In light of the critical role of price information in consumers’ decision making, this study investigates the effect of price rank on consumers’ responses to product list advertising (PLA). The research documents that the price rank is more influential than actual price for PLA. In addition, the research highlights a tradeoff in price-rank decisions: A price rank that drives more clicks does not necessarily lead to more conversions; to drive traffic, managers should strive for an extreme (i.e., either high or low) to elicit more clicks, then follow up with online engagement tools (e.g., cross-selling and product recommendations). To maximize direct revenue, managers instead should strive for moderate ranks to satisfy consumers’ desire for a compromise between price and quality. However, consumers without uncertainty tend to rely less on price rank, so the effects diminish among specific keywords and increase among popular keywords. In order to achieve the desired price ranks, firms participating in PLA might monitor and adjust their advertising offers. There are commonly two specific avenues: Change the product price if the required change is within a certain range or change the advertised product if the required price change is beyond a certain range.

Why you bought it: how influences to buy sports products from personal network members differ between social media and in-person settings

This major research project explores the extent to which normative and informational influences exerted by core and significant ties differ between social media and in-person contexts. Specifically, it focuses on how such influences persuade recreational athletes to buy sports products. Though normative and informational influences from a variety of personal ties have been studied in online and offline settings, they are seldom explicitly compared and contrasted. Moreover, recreational athletes and sports products have never been the subject of such studies. Based on qualitative interviews with six recreational athletes between the ages of 18 and 30, this study uses a content analysis with open coding to identify significant themes. The findings indicate that although in-person normative influence to buy sports products is easily identifiable, normative influence on social media is more difficult to detect. Yet regardless of the context, normative influence is powered by one’s desire for inclusion into a group. On the other hand, informational influence in the form of product recommendations does not differ between the examined settings. Thorough recommendations are more sought after than pithy ones, experts challenge recommendations and those who do not know much about a given product will seek information from experts. However, the findings also indicate that informational influence in the form of observation and analysis is preferred in offline situations compared with online ones. It is therefore clear that separate facets of normative and informational influence each present unique similarities or dissimilarities between in-person and social media settings.

Why you bought it: how influences to buy sports products from personal network members differ between social media and in-person settings

This major research project explores the extent to which normative and informational influences exerted by core and significant ties differ between social media and in-person contexts. Specifically, it focuses on how such influences persuade recreational athletes to buy sports products. Though normative and informational influences from a variety of personal ties have been studied in online and offline settings, they are seldom explicitly compared and contrasted. Moreover, recreational athletes and sports products have never been the subject of such studies. Based on qualitative interviews with six recreational athletes between the ages of 18 and 30, this study uses a content analysis with open coding to identify significant themes. The findings indicate that although in-person normative influence to buy sports products is easily identifiable, normative influence on social media is more difficult to detect. Yet regardless of the context, normative influence is powered by one’s desire for inclusion into a group. On the other hand, informational influence in the form of product recommendations does not differ between the examined settings. Thorough recommendations are more sought after than pithy ones, experts challenge recommendations and those who do not know much about a given product will seek information from experts. However, the findings also indicate that informational influence in the form of observation and analysis is preferred in offline situations compared with online ones. It is therefore clear that separate facets of normative and informational influence each present unique similarities or dissimilarities between in-person and social media settings.

A Study on Multiple Tasks for e-Commerce Marketplaces

In an e-Commerce marketplace, there are multiple tasks that need to be addressed in a day-to-day basis. Some tasks such as product recommendations and product search take a fundamental role in the overall experience a user has on the site. There are however a multitude of lesser known tasks which are also relevant for the business and that need to be addressed with a comparatively smaller investment in teams and quality datasets. Examples of such tasks are the detection of counterfeit or forbidden items, the estimation of package sizes, etc. In this work we study a set of different baseline models and how they work across different tasks that come from real world data.

Active learning strategies for solving the cold user problem in model-based recommender systems

Customers of a webshop are often presented large assortments, which can lead to customers struggling finding their desired product(s), an issue known as choice overload. In order to overcome this issue, recommender systems are used in webshops to provide personalized product recommendations to customers. Though, model-based recommender systems are not able to provide recommendations to new customers (i.e., cold users). To facilitate recommendations to cold users we investigate multiple active learning strategies, and subsequently evaluate which active learning strategy is able to optimally elicit the preferences from the cold users in a matrix factorization context. Our model is empirically validated using a dataset from the webshop of de Bijenkorf, a Dutch department store. We find that the overall best-performing active learning strategy is PopError, an active learning strategy that measures the variance score for each item.

Le bingo électronique apporte-t-il des changements dans les comportements de jeu des adeptes du bingo? / Is Electronic Bingo Changing the Gambling Behaviour of Bingo Players?

La Société des établissements de jeu du Québec, filiale de Loto-Québec, a implanté, à titre pilote, le bingo électronique (appelé Bingo +) en octobre 2018. Il s’agit d’une version du bingo traditionnel qui se joue sur une tablette électronique. Les écrits scientifiques soulèvent l’importance de s’intéresser aux nouvelles offres de jeu, en particulier à leurs effets possibles sur les comportements de jeu. Le Bingo + présente des caractéristiques structurelles – dont l’automatisation – reconnues comme pouvant augmenter les habitudes de jeu. La présente étude vérifie l’évolution des comportements et habitudes de jeu des joueurs de bingo à la suite de l’implantation du Bingo +, sur les plans de la dépense en argent, du temps de jeu, des limites de jeu fixées par les joueurs, de la consommation d’alcool et du jeu d’argent pathologique. Les participants, répartis en joueurs de Bingo + (n = 87) et de bingo traditionnel (n = 207), ont été interrogés par des entrevues téléphoniques semi-structurées aux mesures préimplantation et postimplantation après neuf mois. Les résultats ont indiqué que les habitudes de jeu des participants qui utilisaient la tablette électronique changeaient peu, alors que celles des participants au bingo traditionnel tendaient à se réduire. La discussion porte sur l’impact d’une nouvelle modalité de jeu en salle de bingo sur les comportements de jeu des joueurs, en considérant l’ambiance de jeu et l’adaptation face au produit innové. Nous formulons en conclusion des recommandations visant une meilleure compréhension de l’expérience face à un jeu automatisé.Abstract The Société des établissements de jeu du Québec, a subsidiary of Loto-Québec, implemented electronic bingo (called Bingo +) on a pilot basis in October 2018. It is a version of traditional bingo that is played on an electronic tablet. The literature raises the importance of taking an interest in new gaming offers, among others, for the effects on gambling behavior. Bingo + has structural characteristics that have been identified as being able to increase gambling habits, including automation. This study verifies the evolution of the behavior and playing habits of bingo players, following the implementation of Bingo +, in terms of money spent, playing time, limits set by the players, alcohol consumption and pathological gambling. The participants, divided into Bingo + (n = 87) and traditional bingo (n = 207) players, were interviewed by semi-structured telephone interviews in pre-implantation and postimplantation of nine months. The results indicate that the playing habits of participants who use the electronic tablet change little, while those of traditional bingo participants tend to reduce. The discussion focuses on the impact of a new modality of play in bingo halls on players’ playing behaviors considering the playing atmosphere and the adaptation to the innovated product. Recommendations for a better understanding of the experience with an automated game close the article.