Understanding Echo Chambers in E-commerce Recommender Systems

SIGIR '20: The 43rd International ACM SIGIR conference on research and development in Information Retrieval Virtual Event China July, 2020, pp. 2261-2270, 2020.

Cited by: 0|Bibtex|Views89|DOI:https://doi.org/10.1145/3397271.3401431
EI
Other Links: arxiv.org|dl.acm.org|dblp.uni-trier.de|academic.microsoft.com
Weibo:
We examine the presence of echo chamber in two steps

Abstract:

Personalized recommendation benefits users in accessing contents of interests effectively. Current research on recommender systems mostly focuses on matching users with proper items based on user interests. However, significant efforts are missing to understand how the recommendations influence user preferences and behaviors, e.g., if and...More

Code:

Data:

0
Introduction
  • Recommender systems (RS) comes into play with the rise of online platforms, e.g., social networking sites, online media, and ecommerce[16, 18, 19].
  • Extensive attention has been drawn at this front, arriving at the two coined terms, echo chamber and filter bubble.
  • Both effects might occur after the use of personalized recommenders and entail far-reaching implications.
  • Echo chamber describes the rising up of social communities who share similar opinions within the group [41], while filter bubble [36], as the phenomenon of an overly narrow set of recommenders, was blamed for isolating users in information echo chambers [1]
Highlights
  • Recommender systems (RS) comes into play with the rise of online platforms, e.g., social networking sites, online media, and ecommerce[16, 18, 19]
  • The purpose of our work is to explore the temporal effect of recommendation systems on users, especially to investigate the existence and the characteristics of the echo chamber effect
  • As the user interests are closely related to the user interactions with items, we argue that the items that the user clicked can reflect his/her click interests, and the items that the user purchased can represent his/her purchase interests
  • We examine and analyze echo chamber effect in a realworld e-commerce platform
  • We found that the tendency of echo chamber exists in personalized e-commerce RS in terms of user click behaviors, while on user purchase behaviors, this tendency is mitigated
  • We further analyzed the underlying reason for the observations and found that the feedback loop exists between users and RS, which means that the continuous narrowed exposure of items raised by personalized recommendation algorithms brings consistent content to the Following Group, resulting in the echo chamber effect as a reinforcement of user interests
Results
  • Results Analysis

    RQ1: Does RS reinforce user click or purchase interests? 5.3.1 Internal Validity Indexes.
  • User interest can vary a lot, along with changes in external conditions in e-commerce, such as sales campaigns.
  • As a result, these temporal shifts of user embeddings in the latent space are reflected as the decrease in CH
Conclusion
  • CONCLUSIONS AND FUTURE WORK

    In this paper, the authors examine and analyze echo chamber effect in a realworld e-commerce platform.
  • The authors further analyzed the underlying reason for the observations and found that the feedback loop exists between users and RS, which means that the continuous narrowed exposure of items raised by personalized recommendation algorithms brings consistent content to the Following Group, resulting in the echo chamber effect as a reinforcement of user interests
  • This is one of the first steps towards socially responsible AI in online e-commerce environments.
  • Based on the observations and findings, in the future, the authors will develop refined e-commerce recommendation algorithms to mitigate the echo chamber effects, so as to benefit online users for more informed, effective, and friendly recommendations
Summary
  • Introduction:

    Recommender systems (RS) comes into play with the rise of online platforms, e.g., social networking sites, online media, and ecommerce[16, 18, 19].
  • Extensive attention has been drawn at this front, arriving at the two coined terms, echo chamber and filter bubble.
  • Both effects might occur after the use of personalized recommenders and entail far-reaching implications.
  • Echo chamber describes the rising up of social communities who share similar opinions within the group [41], while filter bubble [36], as the phenomenon of an overly narrow set of recommenders, was blamed for isolating users in information echo chambers [1]
  • Objectives:

    The authors aim to analyze the echo chamber phenomenon in Alibaba Taobao — one of the largest e-commerce platforms in the world.
  • With the desire to address the limitations, the authors aim to explore the existence of echo chamber in real-world e-commerce systems while investigating filter bubble via differentiating it as the potential cause of echo chamber.
  • The purpose of the work is to explore the temporal effect of recommendation systems on users, especially to investigate the existence and the characteristics of the echo chamber effect
  • Results:

    Results Analysis

    RQ1: Does RS reinforce user click or purchase interests? 5.3.1 Internal Validity Indexes.
  • User interest can vary a lot, along with changes in external conditions in e-commerce, such as sales campaigns.
  • As a result, these temporal shifts of user embeddings in the latent space are reflected as the decrease in CH
  • Conclusion:

    CONCLUSIONS AND FUTURE WORK

    In this paper, the authors examine and analyze echo chamber effect in a realworld e-commerce platform.
  • The authors further analyzed the underlying reason for the observations and found that the feedback loop exists between users and RS, which means that the continuous narrowed exposure of items raised by personalized recommendation algorithms brings consistent content to the Following Group, resulting in the echo chamber effect as a reinforcement of user interests
  • This is one of the first steps towards socially responsible AI in online e-commerce environments.
  • Based on the observations and findings, in the future, the authors will develop refined e-commerce recommendation algorithms to mitigate the echo chamber effects, so as to benefit online users for more informed, effective, and friendly recommendations
Tables
  • Table1: Statistics of each user group
  • Table2: Statistics of experiment data. clicked and purchased items as “click embedding” and “purchase embedding” respectively for simplicity
  • Table3: Hopkins statistic
  • Table4: Decreases in Calinski-Harabasz score. The corresponding Ks for each groups are introduced in Section 5.2.2
  • Table5: ARI scores. Same as CHK , the corresponding Ks for each groups are introduced in Section 5.2.2 in purchase embedding, the differences among half of the Ks in [K∗ − 5, K∗ + 5] are not significant
  • Table6: The content diversity of recommended items
Download tables as Excel
Related work
  • Today’s recommender systems are criticized for bringing dangerous byproducts of echo chamber and filter bubble. Sunstein argued that personalized recommenders would fragment users, making likeminded users aggregate [41]. The existing views or interests of these users would be reinforced and amplified since “group polarization often occurs because people are telling one another what they know” [41, 42]. Pariser later described filter bubble, as the effect of recommenders making users isolated from diverse content, and trapping them in an unchanging environment [36]. Though both are concerned with the malicious effect that recommenders would pose, echo chamber emphasizes the polarized environment, while filter bubble lays stress on the undiversified environment.

    Researchers are expressing their concerns of the two effects, and attempting to formulate a richer understanding of the potential characteristics [4, 10, 23, 32, 39]. Considering echo chamber as a significant threat to modern society as they might lead to polarization and radicalization [11], Risius et al analyzed news “likes” on Facebook, and distinguished different types of echo chambers [37]. Mohseni et al reviewed news feed algorithms as well as methods for fake news detection and focused on the unwanted outcomes of echo chamber and filter bubble after using personalized content selection algorithms [30]. They argued that personalized newsfeed might cause polarized social media and the spread of fake content.
Reference
  • David Paul Allen, Henry Jacob Wheeler-Mackta, and Jeremy R Campo. 2017. The Effects of Music Recommendation Engines on the Filter Bubble Phenomenon. (2017).
    Google ScholarFindings
  • Arda Antikacioglu and R Ravi. 2017. Post processing recommender systems for diversity. In KDD. ACM, 707–716.
    Google ScholarLocate open access versionFindings
  • Mahsa Badami, Olfa Nasraoui, and Patrick Shafto. 2018. PrCP: Prerecommendation Counter-Polarization. In KDIR. 280–287.
    Google ScholarLocate open access versionFindings
  • E. Bakshy, S. Messing, and L. A. Adamic. 2015. Exposure to ideologically diverse news and opinion on Facebook. Science 348, 6239 (may 2015), 1130–1132.
    Google ScholarLocate open access versionFindings
  • Amit Banerjee and Raiesh N. Dave. 2004. Validating clusters using the Hopkins statistic. IEEE International Conference on Fuzzy Systems 1 (2004), 149–153.
    Google ScholarLocate open access versionFindings
  • Dimitrios Bountouridis, Jaron Harambam, Mykola Makhortykh, Mónica Marrero, Nava Tintarev, and Claudia Hauff. 2019. SIREN: A Simulation Framework for Understanding the Effects of Recommender Systems in Online News Environments. In ACM Conference on Fairness, Accountability, and Transparency. ACM, 150–159.
    Google ScholarLocate open access versionFindings
  • Laura Burbach, Patrick Halbach, Martina Ziefle, and André Calero Valdez. 2019. Bubble Trouble: Strategies Against Filter Bubbles in Online Social Networks. In International Conference on Human-Computer Interaction. Springer, 441–456.
    Google ScholarLocate open access versionFindings
  • Tadeusz CaliÅĎski and Harabasz JA. 1974. A Dendrite Method for Cluster Analysis. Communications in Statistics - Theory and Methods 3 (01 1974), 1–27.
    Google ScholarLocate open access versionFindings
  • Allison J.B. Chaney, Brandon M. Stewart, and Barbara E. Engelhardt. 2018. How Algorithmic Confounding in Recommendation Systems Increases Homogeneity and Decreases Utility. RecSys (2018).
    Google ScholarLocate open access versionFindings
  • James N Cohen. 2018. Exploring Echo-Systems: How Algorithms Shape Immersive Media Environments. Journal of Media Literacy Education 10, 2 (2018), 139–151.
    Google ScholarLocate open access versionFindings
  • Pranav Dandekar, Ashish Goel, and David T Lee. 2013. Biased assimilation, homophily, and the dynamics of polarization. Proceedings of the National Academy of Sciences 110, 15 (2013), 5791–5796.
    Google ScholarLocate open access versionFindings
  • Abhisek Dash, Animesh Mukherjee, and Saptarshi Ghosh. 2019. A Networkcentric Framework for Auditing Recommendation Systems. In IEEE INFOCOM 2019-IEEE Conference on Computer Communications. IEEE, 1990–1998.
    Google ScholarLocate open access versionFindings
  • Carlos Manuel Moita de Figueiredo. 2014. Emotions and recommender systems: A social network approach. (2014).
    Google ScholarFindings
  • Seth Flaxman, Sharad Goel, and Justin M Rao. 2016. Filter bubbles, echo chambers, and online news consumption. Public opinion quarterly 80, S1 (2016), 298–320.
    Google ScholarLocate open access versionFindings
  • Daniel M Fleder and Kartik Hosanagar. 2007. Recommender systems and their impact on sales diversity. In ACM conference on Electronic commerce. ACM.
    Google ScholarLocate open access versionFindings
  • Zuohui Fu, Yikun Xian, Ruoyuan Gao, Jieyu Zhao, Qiaoying Huang, Yingqiang Ge, Shuyuan Xu, Shijie Geng, Chirag Shah, Yongfeng Zhang, and Gerard de Melo. 2020. Fairness-Aware Explainable Recommendation over Knowledge Graphs. In Proceedings of SIGIR 2020. ACM, New York, NY, USA.
    Google ScholarLocate open access versionFindings
  • Mingkun Gao, Hyo Jin Do, and Wai-Tat Fu. 2018. Burst Your Bubble! An Intelligent System for Improving Awareness of Diverse Social Opinions. In 23rd International Conference on Intelligent User Interfaces. ACM, 371–383.
    Google ScholarLocate open access versionFindings
  • Yingqiang Ge, Shuyuan Xu, Shuchang Liu, Zuohui Fu, Fei Sun, and Yongfeng Zhang. 2020. Learning Personalized Risk Preferences for Recommendation. SIGIR (2020).
    Google ScholarLocate open access versionFindings
  • Yingqiang Ge, Shuyuan Xu, Shuchang Liu, Shijie Geng, Zuohui Fu, and Yongfeng Zhang. 20Maximizing marginal utility per dollar for economic recommendation. In The World Wide Web Conference. 2757–2763.
    Google ScholarFindings
  • Daniel Geschke, Jan Lorenz, and Peter Holtz. 2019. The triple-filter bubble: Using agent-based modelling to test a meta-theoretical framework for the emergence of filter bubbles and echo chambers. British Journal of Social Psychology 58, 1 (2019), 129–149.
    Google ScholarLocate open access versionFindings
  • Julia Handl, Joshua Knowles, and Douglas B. Kell. 2005. Computational cluster validation in post-genomic data analysis. Bioinformatics 21, 15 (08 2005), 3201– 3212.
    Google ScholarLocate open access versionFindings
  • Natali Helberger, Kari Karppinen, and Lucia DâĂŹacunto. 2018. Exposure diversity as a design principle for recommender systems. Information, Communication & Society 21, 2 (2018), 191–207.
    Google ScholarLocate open access versionFindings
  • Martin Hilbert, Saifuddin Ahmed, Jaeho Cho, Billy Liu, and Jonathan Luu. 2018. Communicating with algorithms: a transfer entropy analysis of emotions-based escapes from online echo chambers. Communication Methods and Measures 12, 4 (2018), 260–275.
    Google ScholarLocate open access versionFindings
  • Kartik Hosanagar, Daniel Fleder, Dokyun Lee, and Andreas Buja. 2013. Will the global village fracture into tribes? Recommender systems and their effects on consumer fragmentation. Management Science 60, 4 (2013), 805–823.
    Google ScholarLocate open access versionFindings
  • Lawrence Hubert and Phipps Arabie. 1985. Comparing partitions. Journal of Classification 2, 1 (01 Dec. 1985), 193–218.
    Google ScholarLocate open access versionFindings
  • Ray Jiang, Silvia Chiappa, Tor Lattimore, András György, and Pushmeet Kohli. 2019. Degenerate feedback loops in recommender systems. In Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society. 383–390.
    Google ScholarLocate open access versionFindings
  • Xuehan Jiang. 2018. A community-evolution based approach for detecting the echo chamber effect in recommender systems. (2018).
    Google ScholarFindings
  • Lanu Kim, JD West, and Katherine Stovel. 2017. Echo chambers in science?. In American Sociological Association (ASA) Annual Meeting.
    Google ScholarLocate open access versionFindings
  • Richard G. Lawosn and Peter C. Jurs. 1990. New index for clustering tendency and its application to chemical problems. Journal of Chemical Information and Computer Sciences 30, 1 (1990), 36–41.
    Google ScholarLocate open access versionFindings
  • Sina Mohseni and Eric Ragan. 2018. Combating Fake News with Interpretable News Feed Algorithm. arXiv preprint arXiv:1811.12349 (2018).
    Findings
  • Judith Möller, Damian Trilling, Natali Helberger, and Bram van Es. 2018. Do not blame it on the algorithm: an empirical assessment of multiple recommender systems and their impact on content diversity. Information, Communication & Society 21, 7 (2018), 959–977.
    Google ScholarLocate open access versionFindings
  • CHA Namjun, CHO Hosoo, LEE Sangman, and Junseok HWANG. 2019. Effect of AI Recommendation System on the Consumer Preference Structure in e-Commerce: Based on Two types of Preference. In 2019 21st International Conference on Advanced Communication Technology (ICACT). IEEE, 77–80.
    Google ScholarLocate open access versionFindings
  • Tien T Nguyen, Pik-Mai Hui, F Maxwell Harper, Loren Terveen, and Joseph A Konstan. 2014. Exploring the filter bubble: the effect of using recommender systems on content diversity. In Proceedings of the 23rd WWW. ACM, 677–686.
    Google ScholarLocate open access versionFindings
  • Malay Kumar Pakhira, Sanghamitra Bandyopadhyay, and Ujjwal Maulik. 2004. Validity index for crisp and fuzzy clusters. Pattern Recognit. 37 (2004), 487–501.
    Google ScholarLocate open access versionFindings
  • Zachary A. Pardos and Weijie Jiang. 2019. Combating the Filter Bubble: Designing for Serendipity in a University Course Recommendation System. IntRS’19: Joint Workshop on Interfaces and Human Decision Making for Recommender Systems (2019).
    Google ScholarFindings
  • Eli Pariser. 2011. The filter bubble: What the Internet is hiding from you. Penguin UK.
    Google ScholarLocate open access versionFindings
  • Marten Risius, Okan Aydinguel, and Maximilian Haug. 2019. TOWARDS AN UNDERSTANDING OF CONSPIRACY ECHO CHAMBERS ON FACEBOOK. (2019).
    Google ScholarFindings
  • Kazutoshi Sasahara, Wen Chen, Hao Peng, Giovanni Luca Ciampaglia, Alessandro Flammini, and Filippo Menczer. 2019. On the Inevitability of Online Echo Chambers. arXiv preprint arXiv:1905.03919 (2019).
    Findings
  • Ana-Andreea Stoica and Augustin Chaintreau. 2019. Hegemony in Social Media and the effect of recommendations. In Companion Proceedings of The 2019 World Wide Web Conference. ACM, 575–580.
    Google ScholarLocate open access versionFindings
  • Cass R. Sunstein. 2007. Republic.com 2.0. Princeton University Press. http://www.jstor.org/stable/j.ctt7tbsw
    Findings
  • Cass R Sunstein. 2009. Going to extremes: How like minds unite and divide. Oxford University Press.
    Google ScholarFindings
  • Cass R Sunstein. 2018. Republic: Divided democracy in the age of social media. Princeton University Press.
    Google ScholarFindings
  • Nava Tintarev, Shahin Rostami, and Barry Smyth. 2018. Knowing the unknown: visualising consumption blind-spots in recommender systems. In Proceedings of the 33rd Annual ACM Symposium on Applied Computing. ACM, 1396–1399.
    Google ScholarLocate open access versionFindings
  • Michalis Vazirgiannis. 2009. Clustering Validity. Springer US, Boston, MA, 388– 393. https://doi.org/10.1007/978-0-387-39940-9_616
    Findings
  • Nguyen Xuan Vinh, Julien Epps, and James Bailey. 2009. Information Theoretic Measures for Clusterings Comparison: Is a Correction for Chance Necessary?. In Proceedings of the 26th ICML (ICML ’09). ACM, New York, NY, USA, 1073–1080.
    Google ScholarLocate open access versionFindings
  • Jizhe Wang, Pipei Huang, Huan Zhao, Zhibo Zhang, Binqiang Zhao, and Dik Lun Lee. 2018. Billion-scale Commodity Embedding for E-commerce Recommendation in Alibaba. Proceedings of the 24th ACM SIGKDD (2018).
    Google ScholarLocate open access versionFindings
  • Weina Wang and Yunjie Zhang. 2007. On fuzzy cluster validity indices. Fuzzy Sets and Systems 158 (10 2007), 2095–2117. https://doi.org/10.1016/j.fss.2007.03.004
    Locate open access versionFindings
  • Junjie Wu, Jian Chen, Hui Xiong, and Ming Xie. 2009. External validation measures for K-means clustering: A data distribution perspective. 36 (04 2009), 6050–6061.
    Google ScholarFindings
  • Yongfeng Zhang, Min Zhang, Yiqun Liu, Shaoping Ma, and Shi Feng. 2013. Localized matrix factorization for recommendation based on matrix block diagonal forms. In Proceedings of the 22nd international conference on World Wide Web. ACM, 1511–1520.
    Google ScholarLocate open access versionFindings
  • Guanjie Zheng, Fuzheng Zhang, Zihan Zheng, Yang Xiang, Nicholas Jing Yuan, Xing Xie, and Zhenhui Li. 2018. DRN: A deep reinforcement learning framework for news recommendation. In Proceedings of the 2018 World Wide Web Conference. 167–176.
    Google ScholarLocate open access versionFindings
  • C.-N. Ziegler, S. M. McNee, J. A. Konstan, and G. Lausen. 2005. Improving recommendation lists through topic diversification.. In Proceedings of the 14th international conference on World Wide Web. ACM, 23–32.
    Google ScholarLocate open access versionFindings
Full Text
Your rating :
0

 

Tags
Comments