E. P. Nichols, Ruomeng Xu, Balasubramanian Thiagarajan, Shruti Kamath
This talk describes the system used at Zillow to govern the quantity of email and push messages sent to users. Emphasis is given to practical issues and lessons learned in running the system in production.
{"title":"Zillow: Volume Governing for Email and Push Messages","authors":"E. P. Nichols, Ruomeng Xu, Balasubramanian Thiagarajan, Shruti Kamath","doi":"10.1145/3523227.3547399","DOIUrl":"https://doi.org/10.1145/3523227.3547399","url":null,"abstract":"This talk describes the system used at Zillow to govern the quantity of email and push messages sent to users. Emphasis is given to practical issues and lessons learned in running the system in production.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"65 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127075742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sequential recommender systems are becoming widespread in the online retail and streaming industry. These systems are often trained to predict the next item given a sequence of a user’s recent actions, and standard evaluation metrics reward systems that can identify the most probable items that might appear next. However, some recent papers instead evaluate recommendation systems with popularity-sampled metrics, which measure how well the model can find a user’s next item when hidden amongst generally-popular items. We argue that these popularity-sampled metrics are more appropriate for recommender systems, because the most probable items for a user often include generally-popular items. If the probability that a customer will watch Toy Story is not much more probable than for the average customer, then the movie isn’t especially relevant for them and we should not recommend it. This paper shows that optimizing popularity-sampled metrics is closely related to estimating point-wise mutual information (PMI). We propose and compare two techniques to fit PMI directly, which both improve popularity-sampled metrics for state-of-the-art recommender systems. The improvements are large compared to differences between recently-proposed model architectures.
{"title":"Don’t recommend the obvious: estimate probability ratios","authors":"Roberto Pellegrini, Wenjie Zhao, Iain Murray","doi":"10.1145/3523227.3546753","DOIUrl":"https://doi.org/10.1145/3523227.3546753","url":null,"abstract":"Sequential recommender systems are becoming widespread in the online retail and streaming industry. These systems are often trained to predict the next item given a sequence of a user’s recent actions, and standard evaluation metrics reward systems that can identify the most probable items that might appear next. However, some recent papers instead evaluate recommendation systems with popularity-sampled metrics, which measure how well the model can find a user’s next item when hidden amongst generally-popular items. We argue that these popularity-sampled metrics are more appropriate for recommender systems, because the most probable items for a user often include generally-popular items. If the probability that a customer will watch Toy Story is not much more probable than for the average customer, then the movie isn’t especially relevant for them and we should not recommend it. This paper shows that optimizing popularity-sampled metrics is closely related to estimating point-wise mutual information (PMI). We propose and compare two techniques to fit PMI directly, which both improve popularity-sampled metrics for state-of-the-art recommender systems. The improvements are large compared to differences between recently-proposed model architectures.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125493355","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Previous studies on exploration have shown that users can be nudged to explore further away from their current preferences. However, these effects were shown in a single session study, while it often takes time to explore new tastes and develop new preferences. In this work, we present a longitudinal study on users’ exploration behavior and behavior change over time after they have used a music genre exploration tool for four sessions in six weeks. We test two relevant nudges to help them explore more: the starting point (the personalization of the default initial playlist) and the visualization of users’ previous position(s). Our results show that the personalization level of the default initial playlist in the first session influences the preferred personalization level users set in the second session but fades away in later sessions as users start exploring in different directions. Visualization of users’ previous positions did not anchor users to stay closer to the initial defaults. Over time, users perceived the playlist to be more personalized to their tastes and helpful to explore the genre. Perceived helpfulness increased more when users explored further away from their current preferences. Apart from differences in self-reported measures, we also find some objective evidence for preference change in users’ top tracks from their Spotify profile, that over the period of 6 weeks moved somewhat closer to the genre that users selected to explore with the tool.
{"title":"Exploring the longitudinal effects of nudging on users’ music genre exploration behavior and listening preferences","authors":"Yu Liang, M. Willemsen","doi":"10.1145/3523227.3546772","DOIUrl":"https://doi.org/10.1145/3523227.3546772","url":null,"abstract":"Previous studies on exploration have shown that users can be nudged to explore further away from their current preferences. However, these effects were shown in a single session study, while it often takes time to explore new tastes and develop new preferences. In this work, we present a longitudinal study on users’ exploration behavior and behavior change over time after they have used a music genre exploration tool for four sessions in six weeks. We test two relevant nudges to help them explore more: the starting point (the personalization of the default initial playlist) and the visualization of users’ previous position(s). Our results show that the personalization level of the default initial playlist in the first session influences the preferred personalization level users set in the second session but fades away in later sessions as users start exploring in different directions. Visualization of users’ previous positions did not anchor users to stay closer to the initial defaults. Over time, users perceived the playlist to be more personalized to their tastes and helpful to explore the genre. Perceived helpfulness increased more when users explored further away from their current preferences. Apart from differences in self-reported measures, we also find some objective evidence for preference change in users’ top tracks from their Spotify profile, that over the period of 6 weeks moved somewhat closer to the genre that users selected to explore with the tool.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116082599","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recommender systems are essential tools to support human decision-making in online information spaces. Many state-of-the-art recommender systems adopt advanced machine learning techniques to model and predict user preferences from behavioral data. While such systems can provide useful and effective recommendations, their algorithmic design commonly neglects underlying psychological mechanisms that shape user preferences and behavior. In this tutorial, we offer a comprehensive review of the state of the art and progress in psychology-informed recommender systems, i.e., recommender systems that incorporate human cognitive processes, personality, and affective cues into recommendation models, along with definitions, strengths and weaknesses. We show how such systems can improve the recommendation process in a user-centric fashion. With this tutorial, we aim to stimulate more ideas and discussion with the audience on core issues of this topic such as the identification of suitable psychological models, availability of datasets, or the suitability of existing performance metrics to evaluate the efficacy of psychology-informed recommender systems. Besides, we present takeaways to recommender systems practitioners how to build psychology-informed recommender systems. Previous versions of this tutorial were presented, among others, at The ACM Web Conference 2022 and the ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR) 2022.
推荐系统是在线信息空间中支持人类决策的重要工具。许多最先进的推荐系统采用先进的机器学习技术，从行为数据中建模和预测用户偏好。虽然这样的系统可以提供有用和有效的建议，但它们的算法设计通常忽略了影响用户偏好和行为的潜在心理机制。在本教程中，我们全面回顾了基于心理学的推荐系统的现状和进展，即将人类认知过程、个性和情感线索纳入推荐模型的推荐系统，以及定义、优势和劣势。我们展示了这样的系统如何以用户为中心的方式改进推荐过程。在本教程中，我们的目标是激发更多的想法，并与观众讨论这一主题的核心问题，如识别合适的心理模型，数据集的可用性，或现有性能指标的适用性，以评估心理信息推荐系统的有效性。此外，我们还向推荐系统从业者提出了如何构建基于心理学的推荐系统的建议。本教程的前几个版本在ACM Web Conference 2022和ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR) 2022上发布。
{"title":"Psychology-informed Recommender Systems Tutorial","authors":"E. Lex, M. Schedl","doi":"10.1145/3523227.3547375","DOIUrl":"https://doi.org/10.1145/3523227.3547375","url":null,"abstract":"Recommender systems are essential tools to support human decision-making in online information spaces. Many state-of-the-art recommender systems adopt advanced machine learning techniques to model and predict user preferences from behavioral data. While such systems can provide useful and effective recommendations, their algorithmic design commonly neglects underlying psychological mechanisms that shape user preferences and behavior. In this tutorial, we offer a comprehensive review of the state of the art and progress in psychology-informed recommender systems, i.e., recommender systems that incorporate human cognitive processes, personality, and affective cues into recommendation models, along with definitions, strengths and weaknesses. We show how such systems can improve the recommendation process in a user-centric fashion. With this tutorial, we aim to stimulate more ideas and discussion with the audience on core issues of this topic such as the identification of suitable psychological models, availability of datasets, or the suitability of existing performance metrics to evaluate the efficacy of psychology-informed recommender systems. Besides, we present takeaways to recommender systems practitioners how to build psychology-informed recommender systems. Previous versions of this tutorial were presented, among others, at The ACM Web Conference 2022 and the ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR) 2022.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116488280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Today, most recommender systems employ Machine Learning to recommend posts, products, and other items, usually produced by the users. Although the impressive progress in Deep Learning and Reinforcement Learning, we observe that recommendations made by such systems still do not correlate with actual human preferences. In our tutorial, we will bridge the gap between crowdsourcing and recommender systems communities by showing how one can incorporate human-in-the-loop into their recommender system to gather the real human feedback on the ranked recommendations. We will discuss the ranking data lifecycle and run through it step-by-step. A significant portion of tutorial time is devoted to a hands-on practice, when the attendees will, under our guidance, sample recommendations and build the ground truth dataset using crowdsourced data, and compute the offline evaluation scores.
{"title":"Improving Recommender Systems with Human-in-the-Loop","authors":"Dmitry Ustalov, N. Fedorova, Nikita Pavlichenko","doi":"10.1145/3523227.3547373","DOIUrl":"https://doi.org/10.1145/3523227.3547373","url":null,"abstract":"Today, most recommender systems employ Machine Learning to recommend posts, products, and other items, usually produced by the users. Although the impressive progress in Deep Learning and Reinforcement Learning, we observe that recommendations made by such systems still do not correlate with actual human preferences. In our tutorial, we will bridge the gap between crowdsourcing and recommender systems communities by showing how one can incorporate human-in-the-loop into their recommender system to gather the real human feedback on the ranked recommendations. We will discuss the ranking data lifecycle and run through it step-by-step. A significant portion of tutorial time is devoted to a hands-on practice, when the attendees will, under our guidance, sample recommendations and build the ground truth dataset using crowdsourced data, and compute the offline evaluation scores.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128568161","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The modern e-commerce web pages have brought better customer experience and more profitable services by whole page optimization at different granularity, e.g., page layout optimization, item ranking optimization, etc. Generating the proper page layout per customer’s request is one of the vital tasks during the web page rendering process, which can directly impact customers’ shopping experience and their decision-making. In this paper, we formulate the request-rendering interactions as a Markov decision process (MDP) and solve it by deep reinforcement learning (RL). Specifically, we present the design and implementation of applying offline Deep Q-Learning (DQN) to the contextual page layout optimization problem. Through the offline evaluation method, we demonstrate the effectiveness of the proposed framework, i.e., the RL agent has the potential to perform better than the baseline ranker by learning from the offline data set, e.g., the RL agent can improve the average cumulative rewards up to 36.69% comparing to the baseline ranker.
{"title":"Automate Page Layout Optimization: An Offline Deep Q-Learning Approach","authors":"Zhou Qin, Wenyang Liu","doi":"10.1145/3523227.3547400","DOIUrl":"https://doi.org/10.1145/3523227.3547400","url":null,"abstract":"The modern e-commerce web pages have brought better customer experience and more profitable services by whole page optimization at different granularity, e.g., page layout optimization, item ranking optimization, etc. Generating the proper page layout per customer’s request is one of the vital tasks during the web page rendering process, which can directly impact customers’ shopping experience and their decision-making. In this paper, we formulate the request-rendering interactions as a Markov decision process (MDP) and solve it by deep reinforcement learning (RL). Specifically, we present the design and implementation of applying offline Deep Q-Learning (DQN) to the contextual page layout optimization problem. Through the offline evaluation method, we demonstrate the effectiveness of the proposed framework, i.e., the RL agent has the potential to perform better than the baseline ranker by learning from the offline data set, e.g., the RL agent can improve the average cumulative rewards up to 36.69% comparing to the baseline ranker.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128666616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Raul Gomez Bruballa, Lauren Burnham-King, Alessandra Sala
Providing meaningful recommendations in a content marketplace is challenging due to the fact that users are not the final content consumers. Instead, most users are creatives whose interests, linked to the projects they work on, change rapidly and abruptly. To address the challenging task of recommending images to content creators, we design a RecSys that learns visual styles preferences transversal to the semantics of the projects users work on. We analyze the challenges of the task compared to content-based recommendations driven by semantics, propose an evaluation setup, and explain its applications in a global image marketplace.
{"title":"Learning Users’ Preferred Visual Styles in an Image Marketplace","authors":"Raul Gomez Bruballa, Lauren Burnham-King, Alessandra Sala","doi":"10.1145/3523227.3547382","DOIUrl":"https://doi.org/10.1145/3523227.3547382","url":null,"abstract":"Providing meaningful recommendations in a content marketplace is challenging due to the fact that users are not the final content consumers. Instead, most users are creatives whose interests, linked to the projects they work on, change rapidly and abruptly. To address the challenging task of recommending images to content creators, we design a RecSys that learns visual styles preferences transversal to the semantics of the projects users work on. We analyze the challenges of the task compared to content-based recommendations driven by semantics, propose an evaluation setup, and explain its applications in a global image marketplace.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129122852","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Modern e-commerce platforms make use of promotional offers such as discounts and rewards to encourage customers to complete purchases. While offering the promotions has a great effect on the sales, it also generates a monetary loss. By utilizing causal machine learning and optimization, our team at Booking.com was able to personalize the promotions allocation to customers, while efficiently controlling the spend within a given budget. In this talk we’ll share the personalized promotion assignment techniques, such as uplift modeling and constrained optimization, which helped us to predict the outcomes of discounts offering and allocate them efficiently. This solution allowed us to unlock promotional campaigns to bring more value to the customers and grow our business.
{"title":"Personalizing Benefits Allocation Without Spending Money: Utilizing Uplift Modeling in a Budget Constrained Setup","authors":"Dmitri Goldenberg, Javier Albert","doi":"10.1145/3523227.3547381","DOIUrl":"https://doi.org/10.1145/3523227.3547381","url":null,"abstract":"Modern e-commerce platforms make use of promotional offers such as discounts and rewards to encourage customers to complete purchases. While offering the promotions has a great effect on the sales, it also generates a monetary loss. By utilizing causal machine learning and optimization, our team at Booking.com was able to personalize the promotions allocation to customers, while efficiently controlling the spend within a given budget. In this talk we’ll share the personalized promotion assignment techniques, such as uplift modeling and constrained optimization, which helped us to predict the outcomes of discounts offering and allocate them efficiently. This solution allowed us to unlock promotional campaigns to bring more value to the customers and grow our business.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129239765","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
1 MOTIVATION AND GOAL Diseases such as diabetes, cancer and heart disease demand that patients take an active role in disease management and seek health information for decision-making and self-management [1]. The most common methods of providing reliable information to patients include health literacy workshops and patient education materials [12, 15]. However, these materials are prepared for the general patient population and not always tailored to each patient’s specific needs [2, 12]. In addition, patients seek for their information needs through search engines. Previous studies have reported that search engines don’t always support information needs of patients [11]. Consequently, people seek information on disease specific online health communities (OHCs) [9]. But the risk of propagating misinformation still exists because existing OHCs do not provide an infrastructure to help patients find relevant and trustworthy information [8, 19].Existing patient focused health search engines and health recommender systems can accommodate better support of patients’ information needs by providing them trustworthy resources. However, current PHRS are personalized to patients’ interests and so information provided by these layperson-oriented systems is rather general [5, 6]. In fact, previous study has shown that a patient’s personal knowledge about their disease becomes more sophisticated over the course of disease [6, 7]. Therefore, our principal motivation is to fill this gap in current PHRS, by investigating ways to suggest individualized health information that not only adapts to patients’ current information needs but also patients’ knowledge-level across the disease trajectory. The health materials recommended at the level of patients’ knowledge will not only help them engage with materials but also help in informed decision making and self management [16].
{"title":"KA-Recsys: Knowledge Appropriate Patient Focused Recommendation Technologies","authors":"Khushboo Thaker","doi":"10.1145/3523227.3547422","DOIUrl":"https://doi.org/10.1145/3523227.3547422","url":null,"abstract":"1 MOTIVATION AND GOAL Diseases such as diabetes, cancer and heart disease demand that patients take an active role in disease management and seek health information for decision-making and self-management [1]. The most common methods of providing reliable information to patients include health literacy workshops and patient education materials [12, 15]. However, these materials are prepared for the general patient population and not always tailored to each patient’s specific needs [2, 12]. In addition, patients seek for their information needs through search engines. Previous studies have reported that search engines don’t always support information needs of patients [11]. Consequently, people seek information on disease specific online health communities (OHCs) [9]. But the risk of propagating misinformation still exists because existing OHCs do not provide an infrastructure to help patients find relevant and trustworthy information [8, 19].Existing patient focused health search engines and health recommender systems can accommodate better support of patients’ information needs by providing them trustworthy resources. However, current PHRS are personalized to patients’ interests and so information provided by these layperson-oriented systems is rather general [5, 6]. In fact, previous study has shown that a patient’s personal knowledge about their disease becomes more sophisticated over the course of disease [6, 7]. Therefore, our principal motivation is to fill this gap in current PHRS, by investigating ways to suggest individualized health information that not only adapts to patients’ current information needs but also patients’ knowledge-level across the disease trajectory. The health materials recommended at the level of patients’ knowledge will not only help them engage with materials but also help in informed decision making and self management [16].","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130547914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
After multiple years of providing automated video recommendations in the ZDFmediathek, ZDF has established a solid ground for the usage of recommender systems. Being a Public Service Media (PSM) provider, our most important driver on this journey is our Public Service Media Remit (PSMR). We are committed to cultivate PSM values such as diversity, fairness, and transparency while providing fresh and relevant content. Therefore, it is important for us to not only measure the success of our recommender systems in terms of basic business Key Performance Indicators (KPIs) such as clicks and viewing minutes but also to ensure and to measure the achievement of PSM values. While speaking about PSM values, however, it is important to keep in mind that there is no easy way to directly measure values as such. In order to be able to measure their extent in a recommender system, we need to translate these values into public value metrics. However, not only the final results are essential for the PSMR. Additionally, it is highly important to establish transparency while working towards these results, that is, while defining the data, the algorithms, and the pipelines used in recommender systems. In our talk we will provide a deeper insight into how we approach this task with Model Cards and give an overview of some models, their Model Cards, and metrics that we are currently using for ZDFmediathek.
{"title":"Translating the Public Service Media Remit into Metrics and Algorithms","authors":"Andreas Grün, Xenija Neufeld","doi":"10.1145/3523227.3547380","DOIUrl":"https://doi.org/10.1145/3523227.3547380","url":null,"abstract":"After multiple years of providing automated video recommendations in the ZDFmediathek, ZDF has established a solid ground for the usage of recommender systems. Being a Public Service Media (PSM) provider, our most important driver on this journey is our Public Service Media Remit (PSMR). We are committed to cultivate PSM values such as diversity, fairness, and transparency while providing fresh and relevant content. Therefore, it is important for us to not only measure the success of our recommender systems in terms of basic business Key Performance Indicators (KPIs) such as clicks and viewing minutes but also to ensure and to measure the achievement of PSM values. While speaking about PSM values, however, it is important to keep in mind that there is no easy way to directly measure values as such. In order to be able to measure their extent in a recommender system, we need to translate these values into public value metrics. However, not only the final results are essential for the PSMR. Additionally, it is highly important to establish transparency while working towards these results, that is, while defining the data, the algorithms, and the pipelines used in recommender systems. In our talk we will provide a deeper insight into how we approach this task with Model Cards and give an overview of some models, their Model Cards, and metrics that we are currently using for ZDFmediathek.","PeriodicalId":443279,"journal":{"name":"Proceedings of the 16th ACM Conference on Recommender Systems","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133077728","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：