How Do Recommender Systems Affect Sales Diversity? A Cross-Category Investigation via Randomized Field Experiment

Dokyun Lee,Kartik Hosanagar

DOI: https://doi.org/10.1287/isre.2018.0800

IF: 4.9

2019-03-01

Information Systems Research

Abstract:<p>Information Systems Research, Ahead of Print. <br/></p>

information science & library science,management

What problem does this paper attempt to address?

Covering Diversification and Fairness for Better Recommendation (Short Paper).

Qing Yang,Li Han,Ya Zhou,Shaobing Liu,Jingwei Zhang,Zhongqin Bi,Fang Pan

DOI: https://doi.org/10.1007/978-3-030-30146-0_54

2019-01-01

Abstract:Smart applications are appealing an accurate matching between users and items, in which recommendation technologies are applied widely. Since recommendation serve for two roles, namely users and items, accuracy is not the only focus, the diversification and fairness should also be paid more attention for improving recommendation performance. The tradeoff among the accuracy, diversification and fairness on recommendation is bringing a big challenge. This paper proposed a novelty recommendation model to ensure the recommendation performance, which introduces a multi-variate linear regression model to cooperate with the collaborative filtering method. This study utilizes an improved similarity metrics to discover the closeness between users and item categories under the help of the collaborative filtering methods, and exploits the micro attribute information of items by a multi-variate linear regression model to decide the final recommended items. The experimental results show that our proposed method can provide better recommendation accuracy, diversification and fairness than the recommendation based on pure collaborative filtering method.
Recommender systems and their impact on sales diversity

Daniel M. Fleder,Kartik Hosanagar

DOI: https://doi.org/10.1145/1250910.1250939

2007-01-01

Abstract:This paper examines the effect of recommender systems on the diversity of sales. Two anecdotal views exist about such effects. Some believe recommenders help consumers discover new products and thus increase sales diversity. Others believe recommenders only reinforce the popularity of already popular products. This paper is a first attempt to reconcile these seemingly incompatible views. We explore the question in two ways. First, modeling recommender systems analytically allows us to explore their path dependent effects. Second, turning to simulation, we increase the realism of our results by combining choice models with actual implementations of recommender systems. We arrive at three main results. One, some common recommenders lead to a net reduction in average sales diversity. Two, there exists path dependence, and in individual instances the same recommender can either increase or decrease diversity. Three, we show how basic design choices affect the outcome.
Blockbuster culture ’ s next rise or fall : The effect of recommender systems on sales diversity

D. Fleder,K. Hosanagar

Abstract:1 Introduction Recommender systems offer benefits to both consumers and firms. For consumers, recommender systems help individuals both become aware of new products as well as select desirable products among myriad choices (Pham & Healey, 2005). For firms, recommender systems have the potential to increase profits by converting browsers into buyers, cross-selling products, and increasing loyalty through a customized browsing experience (Schafer et al., 1999). Major online firms, such as Amazon.com, Netflix, and Yahoo! Music, host recommender systems, and the prevalence of such systems is expected to grow. In this paper, we study the effect of recommender systems on consumer purchasing of hedonic goods. Two different views exist on the impact of recommenders on diversity of sales. Some believe recommenders will help consumers discover new products (Brynjolffson et al. 2006). On the other hand, others believe the design of common recommender systems will only reinforce the popularity of already popular titles (Mooney & Roy, 1999). The first belief leads to a hypothesis that recommender systems will increase sales diversity. The second belief leads to a hypothesis that recommender systems will decrease sales diversity. This paper is a first attempt to reconcile these seemingly incompatible views. We develop an analytical model of recommender systems and their influence on consumer choice. The model allows us to explore the strong, path-dependent interaction between recommendations and sales. We believe this paper is the first to attempt to isolate the impact of recommender systems on long-run sales diversity. 2 Prior work Despite the growing use of recommender systems, little is known about how these systems affect consumer behavior. In marketing, Cooke et al. (2002) examine how the likelihood of purchasing a recommended product depends on the information provided, context, and familiarity with other products simultaneously recommended. Senecal et al. (2004) experimentally show that recommendations do influence choice. They find that online recommender systems can have even more influence than human recommendations, and the effect is often greater for experience goods than search goods. To the best of our best knowledge, there have been no studies that have attempted to isolate the specific impact of recommenders on sales diversity. However, the topic has received mention from several researchers. For example, Mooney and Roy (1999) suggest that collaborative filters may perpetuate homogeneity in a library setting, but this brief comment is used more to suggest difficulties of working with sparse data than to examine social implications. Brynjolffson …

Computer Science,Business
Empirical Analysis of the Impact of Recommender Systems on Sales

Bhavik Pathak,Robert Garfinkel,Ram D. Gopal,Rajkumar Venkatesan,Fang Yin

DOI: https://doi.org/10.2753/mis0742-1222270205

IF: 7.582

2010-10-01

Journal of Management Information Systems

Abstract:Online retailers are increasingly using information technologies to provide value-added services to customers. Prominent examples of these services are online recommender systems and consumer feedback mechanisms, both of which serve to reduce consumer search costs and uncertainty associated with the purchase of unfamiliar products. The central question we address is how recommender systems affect sales. We take into consideration the interaction among recommendations, sales, and price. We then develop a robust empirical model that incorporates the indirect effect of recommendations on sales through retailer pricing, potential simultaneity between sales and recommendations, and a comprehensive measure of the strength of recommendations. Applying the model to a panel data set collected from two online retailers, we found that the strength of recommendations has a positive effect on sales. Moreover, this effect is moderated by the recency effect, where more recently released recommended items positively affect the cross-selling efforts of sellers. We also show that recommender systems help to reinforce the long-tail phenomenon of electronic commerce, and obscure recommendations positively affect cross-selling. We also found a positive effect of recommendations on prices. These results suggest that recommendations not only improve sales but they also provide added flexibility to retailers to adjust their prices. A comparative analysis reveals that recommendations have a higher effect on sales than does consumer feedback. Our empirical results show that providing value-added services, such as digital word of mouth and recommendations, allows retailers to charge higher prices while at the same time increasing demand by providing more information regarding the quality and match of products.

information science & library science,management,computer science, information systems
Empirical Research on the Impact of Personalized Recommendation Diversity.

Lin Zhang,Qiang Yan,Junqiang Lu,Yongqiang Chen,Yi Liu

DOI: https://doi.org/10.24251/hicss.2019.159

2019-01-01

Abstract:Personalized recommendation has important implications in raising online shopping efficiency and increasing product sales. There has been wide interest in finding ways to provide more efficient personalized recommendations. Most existing studies focus on how to improve the accuracy of the recommendation algorithms, or are more concerned on ways to increase consumer satisfaction. Unlike these studies, our study focuses on the process of decision-making, using long tail theory as a basis, to reveal the mechanisms involved in consumers' adoption of recommendations. This paper analyzes the effect of personalized recommendations from two angles: product sales and ratings, and tries to point out differences in consumer preferences between mainstream products and niche products, high rating products and low rating products, search products and experience products. The study verifies that consumers demand diversity in the recommended content, and also provides suggestions on how to better plan and operate a personalized recommendation system.
A Personalized Recommendation Algorithm Via Biased Random Walk

Da-Cheng Nie,Yan Fu,Jun-Lin Zhou,Zhen Liu,Zi-Ke Zhang,Chuang Liu

DOI: https://doi.org/10.1109/JCSSE.2014.6841883

2014-01-01

Abstract:With the rapid development of Internet, Recommender Systems can help us efficiently find the useful objects in the information era. Generally, the traditional random walk algorithm has high accuracy but low personality and diversity. In this paper, we propose an improved random walk algorithm by depressing the influence of large-degree objects. Experimental results on MovieLens and Netflix data sets show that this algorithm can effectively improve not only the accuracy (improved by 5.5% and 5.9%, respectively) but also the diversity.
Divrec: A Framework For Top-N Recommendation With Diversification In E-Commerce

Kejun He,Junyu Niu,Chaofeng Sha

DOI: https://doi.org/10.1007/978-3-319-11116-2_34

2014-01-01

Abstract:In order to increase sales for e-commerce websites and meet customer expectations, recommender systems need to recommend more niche products consumers might like. However, traditional product recommender systems usually aim to improve the recommendation accuracy while overlook the diversity within the recommendation lists. In this paper, firstly we examine the importance of diversity within recommended lists through a psychological survey. Motivated by our observations, we develop a general framework, called DivRec, to improve recommendation diversity without lowering accuracy. Experimental results on an e-commerce dataset demonstrate that our approach outperforms state-of-the-art techniques in terms of both accuracy and diversity.
Making Recommendations Better: The Role of User Online Purchase Intention Identification

Kuan Fang,Qi Zhang,Zhuoran Zhuang,Zi-Ke Zhang

DOI: https://doi.org/10.1109/ICSN.2016.7501928

2016-01-01

Abstract:The past few years has witnessed the great success of recommender systems, which can significantly help users find relevant and interesting items from the vast array of online products. Recently, a vast class of researches in this area mainly focus on making recommendations by designing effective algorithms. Comparatively, the user intention, especially the role of purchase intention in recommender systems is relatively lack of study. In this paper, with real e-commercial data from Tmall.com, we firstly analyze users' online behaviors and propose a scenario-based identification approach to classify users into two groups: one with obvious purchase intention, and another without such motivation. We then use Random Forest classification method to validate its usefulness. Subsequently, we implement an online demo to visually detect the real-time purchase intention. Finally, we employ the classical item-based collaborative filtering framework to provide recommendations to those two group users. Experimental results show that recommendation performance indeed can be enhanced by identifying online purchase intention.
Impact of recommender systems on unplanned purchase behaviours in e-commerce

Zhang Ying,Chen Caixia,Gu Wen,Liu Xiaogang

DOI: https://doi.org/10.1109/IEA.2018.8387066

2018-01-01

Abstract:The emergence of the Chinese online marketplace has stimulated the increasing popularity of recommender systems among websites as a tool that assists consumers to reduce information overload. This study proposes a framework that employs a user-cantered approach to evaluate the impact of recommender systems on unplanned purchase behaviours of Chinese consumers in e-commerce. Multi-item scales are adopted from the literature and used to measure five constructs, namely, perceived ease of use, perceived usefulness, shopping enjoyment, consumer satisfaction, and unplanned purchase behaviour. The final samples comprised 281 usable Chinese respondents in both gender groups via online survey. Structural equation modelling technique is employed to test all hypotheses. Results show that perceived usefulness plays a leading role in influencing impulse buying. Apart from having a direct and positive effect on unplanned purchase behaviours, perceived usefulness indirectly contributes to impulse buying by affecting consumer satisfaction, which is also essential to promoting unplanned purchases.
A New Marketing Recommendation System Using a Hybrid Approach to Generate Smart Offers

Doae Mensouri,Abdellah Azmani

DOI: https://doi.org/10.2478/acss-2022-0016

2022-12-01

Applied Computer Systems

Abstract:Abstract In order to increase sales, companies try their best to develop relevant offers that anticipate customer needs. One way to achieve this is by leveraging artificial intelligence algorithms that process data collected based on customer transactions, extract insights and patterns from them, and then present them in a user-friendly way to human or artificial intelligence decision makers. This study is based on a hybrid approach, it starts with an online marketplace dataset that contains many customers’ purchases and ends up with global personalized offers based on three different datasets. The first one, generated by a recommendation system, identifies for each customer a list of products they are most likely to buy. The second is generated with an Apriori algorithm. Apriori is used as an associate rule mining technique to identify and map frequent patterns based on support, confidence, and lift factors, and also to pull important rules between products. The third and last one describes, for each customer, their purchase probability in the next few weeks, based on the BG/NBD model and the average of transactions using the Gamma-Gamma model, as well as the satisfaction based on the CLV and RFMTS models. By combining all three datasets, specific and targeted promotion strategies can be developed. Thus, the company is able to anticipate customer needs and generate the most appropriate offers for them while respecting their budget, with minimum operational costs and a high probability of purchase transformation.
Diversify or Not: Dynamic Diversification for Personalized Recommendation

Bin Hao,Min Zhang,Cheng Guo,Weizhi Ma,Yiqun Liu,Shaoping Ma

DOI: https://doi.org/10.1007/978-3-030-75765-6_37

2021-01-01

Abstract:Diversity is believed to be an essential factor in improving user satisfaction in recommender systems, while how to take advantage of it has long been a problem worth exploring. Existing work either ignores the influence of diversity or overlooks users’ different diversity demands in recommendations. In this study, we analyze users’ behaviors on a real-world dataset collected from an e-commerce website and find that the demand for diversity changes among different users, even the same user’s demand varies among different shopping scenarios. There is also evidence that users’ behaviors are affected by the diversity of impressions, which has been often ignored by traditional session-based recommendation models. Then, we propose a Dynamic Diversification Recommendation Model (DDRM) with the integration of both click and impression diversities to better make use of diversity for recommendations. Extensive experimental results demonstrate that DDRM outperforms all baseline methods significantly. Further studies show our model can provide more precise and reasonable recommendations.
How Do Recommender Systems Lead to Consumer Purchases? A Causal Mediation Analysis of a Field Experiment

Xitong Li,Jörn Grahl,Oliver Hinz

DOI: https://doi.org/10.1287/isre.2021.1074

IF: 4.9

2021-12-14

Information Systems Research

Abstract:The findings underscore the important role of consumers’ consideration sets in mediating the positive effects of recommender systems on consumer purchases. Practical strategies can be developed to facilitate the formation of the consideration sets. For example, to reduce consumers’ search costs and cognitive efforts, online retailers can display the recommended products in a descending order according to the predicted closeness of consumers’ preferences. Online retailers can further indicate the predicted closeness scores of consumers’ preferences for the recommended products. Given such a placement arrangement, consumers can quickly screen the recommended products and add the most relevant alternatives to their consideration sets, which should facilitate consumers’ shopping process and increase the shopping satisfaction. The findings also suggest that a larger consideration set due to the use of recommender systems could induce consumers to buy. Yet, it is difficult for consumers to manage many alternatives when the consideration set is very large. To facilitate consumers’ shopping process, online retailers need to consider strategies and tools that help consumers manage the alternatives in the consideration set in a better-organized manner and facilitate the comparison across the alternatives.

information science & library science,management
On Practical Diversified Recommendation with Controllable Category Diversity Framework

Tao Zhang,Luwei Yang,Zhibo Xiao,Wen Jiang,Wei Ning

2024-02-06

Abstract:Recommender systems have made significant strides in various industries, primarily driven by extensive efforts to enhance recommendation accuracy. However, this pursuit of accuracy has inadvertently given rise to echo chamber/filter bubble effects. Especially in industry, it could impair user's experiences and prevent user from accessing a wider range of items. One of the solutions is to take diversity into account. However, most of existing works focus on user's explicit preferences, while rarely exploring user's non-interaction preferences. These neglected non-interaction preferences are especially important for broadening user's interests in alleviating echo chamber/filter bubble effects.Therefore, in this paper, we first define diversity as two distinct definitions, i.e., user-explicit diversity (U-diversity) and user-item non-interaction diversity (N-diversity) based on user historical behaviors. Then, we propose a succinct and effective method, named as Controllable Category Diversity Framework (CCDF) to achieve both high U-diversity and N-diversity simultaneously.Specifically, CCDF consists of two stages, User-Category Matching and Constrained Item Matching. The User-Category Matching utilizes the DeepU2C model and a combined loss to capture user's preferences in categories, and then selects the top-$K$ categories with a controllable parameter $K$.These top-$K$ categories will be used as trigger information in Constrained Item Matching. Offline experimental results show that our proposed DeepU2C outperforms state-of-the-art diversity-oriented methods, especially on N-diversity task. The whole framework is validated in a real-world production environment by conducting online A/B testing.

Information Retrieval
Impact of recommender algorithms on the sales of e-commerce websites

Manish Sinha,Divyank Srivastava

DOI: https://doi.org/10.1108/ijis-09-2020-0155

2021-01-13

International Journal of Innovation Science

Abstract:Purpose With the current pandemic situation, the world is shifting to online buying and therefore the purpose of this study is to understand how the industry can improve sales based on the product recommendations shown on their online platforms. Design/methodology/approach This paper has studied content-based filtering using decision trees algorithm and collaborative filtering using K-nearest neighbour algorithm and measured their impact on sales of product of different genres on e-commerce websites and if their recommendation causes a difference in sales.This paper has conducted a field experiment to analyse the customer frequency, change in sales caused by different algorithms and also tried analysing the change in buying preferences of customers in post-pandemic situation and how this paper can improve on the search results by incorporating them in the already used algorithms. Findings This study indicates that different algorithms cause differences in sales and score over each other depending upon the category of the product sold. It also suggests that post-Covid, the buying frequency and the preferences of consumers have changed significantly. Research limitations/implications The study is limited to existing users of these sites, it also requires the sites to have a huge database of active users and products. Also, the preferences and likings of Indian subcontinent might not generally apply everywhere else. Originality/value This study enables better insight into consumer behaviour, thus enabling the data scientists to design better algorithms and help the companies improve their product sales.

English Else
Addressing Marketing Bias in Product Recommendations

Mengting Wan,Jianmo Ni,Rishabh Misra,Julian McAuley

DOI: https://doi.org/10.48550/arXiv.1912.01799

2019-12-04

Abstract:Modern collaborative filtering algorithms seek to provide personalized product recommendations by uncovering patterns in consumer-product interactions. However, these interactions can be biased by how the product is marketed, for example due to the selection of a particular human model in a product image. These correlations may result in the underrepresentation of particular niche markets in the interaction data; for example, a female user who would potentially like motorcycle products may be less likely to interact with them if they are promoted using stereotypically 'male' images. In this paper, we first investigate this correlation between users' interaction feedback and products' marketing images on two real-world e-commerce datasets. We further examine the response of several standard collaborative filtering algorithms to the distribution of consumer-product market segments in the input interaction data, revealing that marketing strategy can be a source of bias for modern recommender systems. In order to protect recommendation performance on underrepresented market segments, we develop a framework to address this potential marketing bias. Quantitative results demonstrate that the proposed approach significantly improves the recommendation fairness across different market segments, with a negligible loss (or better) recommendation accuracy.

Information Retrieval,Social and Information Networks
SalesExplorer: Exploring sales opportunities from white-space customers in the enterprise market.

Dongsheng Li,Yaoping Ruan,Qin Lv,Li Shang

DOI: https://doi.org/10.1016/j.knosys.2016.09.011

IF: 8.139

2016-01-01

Knowledge-Based Systems

Abstract:In commercial sales and services, recommender systems have been widely adopted to predict customers’ purchase interests using their prior purchasing behaviors. Cold-start is a known challenge to existing recommendation techniques, e.g., the popular collaborative filtering method is not applicable to predict the interests of “white-space” customers since they have no prior purchasing history in the targeted product categories.
Interactive Content Diversity and User Exploration in Online Movie Recommenders: A Field Experiment

Ruixuan Sun,Avinash Akella,Ruoyan Kong,Moyan Zhou,Joseph A. Konstan

DOI: https://doi.org/10.1080/10447318.2023.2262796

2023-09-23

Abstract:Recommender systems often struggle to strike a balance between matching users' tastes and providing unexpected recommendations. When recommendations are too narrow and fail to cover the full range of users' preferences, the system is perceived as useless. Conversely, when the system suggests too many items that users don't like, it is considered impersonal or ineffective. To better understand user sentiment about the breadth of recommendations given by a movie recommender, we conducted interviews and surveys and found out that many users considered narrow recommendations to be useful, while a smaller number explicitly wanted greater breadth. Additionally, we designed and ran an online field experiment with a larger user group, evaluating two new interfaces designed to provide users with greater access to broader recommendations. We looked at user preferences and behavior for two groups of users: those with higher initial movie diversity and those with lower diversity. Among our findings, we discovered that different level of exploration control and users' subjective preferences on interfaces are more predictive of their satisfaction with the recommender.

Human-Computer Interaction,Information Retrieval
A Joint Optimization Approach for Personalized Recommendation Diversification.

Xiaojie Wang,Jianzhong Qi,Kotagiri Ramamohanarao,Yu Sun,Bo Li,Rui Zhang

DOI: https://doi.org/10.1007/978-3-319-93040-4_47

2018-01-01

Abstract:In recommendation systems, items of interest are often classified into categories such as genres of movies. Existing research has shown that diversified recommendations can improve real user experience. However, most existing methods do not consider the fact that users’ levels of interest (i.e., user preferences) in different categories usually vary, and such user preferences are not reflected in the diversified recommendations. We propose an algorithm that considers user preferences for different categories when recommending diversified results, and refer to this problem as personalized recommendation diversification. In the proposed algorithm, a model that captures user preferences for different categories is optimized jointly toward both relevance and diversity. To provide the proposed algorithm with informative training labels and effectively evaluate recommendation diversity, we also propose a new personalized diversity measure. The proposed measure overcomes limitations of existing measures in evaluating recommendation diversity: existing measures either cannot effectively handle user preferences for different categories, or cannot evaluate both relevance and diversity at the same time. Experiments using two real-world datasets confirm the superiority of the proposed algorithm, and show the effectiveness of the proposed measure in capturing user preferences.
Choice-Based Recommender Systems: A Unified Approach to Achieving Relevancy and Diversity

Hai Jiang,Xin Qi,He Sun

DOI: https://doi.org/10.1287/opre.2014.1292

IF: 2.7

2014-10-01

Operations Research

Abstract:Recommender systems have been widely used by online stores to suggest items of interest to users. These systems often identify a subset of items from a much larger set that best matches the user's interest. A key concern with existing approaches is overspecialization, which results in returning items that are too similar to each other. Unlike existing solutions that rely on diversity metrics to reduce similarity among recommended items, we propose using choice probability to measure the overall quality of a recommendation list, which unifies the desire to achieve both relevancy and diversity in recommendation. We first define the recommendation problem from the discrete choice perspective. We then model the problem under the multilevel nested logit model, which is capable of handling similarities between alternatives along multiple dimensions. We formulate the problem as a nonlinear binary integer programming problem and develop an efficient dynamic programming algorithm that solves the problem to optimum in O(nKSR 2 ) time, where n is the number of levels and K is the maximum number of children nests a nest can have in the multilevel nested logit model, S is the total number of items in the item pool, and R is the number of items wanted in recommendation.

management,operations research & management science

How Do Recommender Systems Affect Sales Diversity? A Cross-Category Investigation via Randomized Field Experiment

Covering Diversification and Fairness for Better Recommendation (Short Paper).

Recommender systems and their impact on sales diversity

Blockbuster culture ’ s next rise or fall : The effect of recommender systems on sales diversity

Empirical Analysis of the Impact of Recommender Systems on Sales

Empirical Research on the Impact of Personalized Recommendation Diversity.

A Personalized Recommendation Algorithm Via Biased Random Walk

Divrec: A Framework For Top-N Recommendation With Diversification In E-Commerce

Making Recommendations Better: The Role of User Online Purchase Intention Identification

Impact of recommender systems on unplanned purchase behaviours in e-commerce

A New Marketing Recommendation System Using a Hybrid Approach to Generate Smart Offers

Diversify or Not: Dynamic Diversification for Personalized Recommendation

How Do Recommender Systems Lead to Consumer Purchases? A Causal Mediation Analysis of a Field Experiment

On Practical Diversified Recommendation with Controllable Category Diversity Framework

Impact of recommender algorithms on the sales of e-commerce websites

Addressing Marketing Bias in Product Recommendations

SalesExplorer: Exploring sales opportunities from white-space customers in the enterprise market.

Interactive Content Diversity and User Exploration in Online Movie Recommenders: A Field Experiment

A Joint Optimization Approach for Personalized Recommendation Diversification.

Choice-Based Recommender Systems: A Unified Approach to Achieving Relevancy and Diversity