Digital nudging with recommender systems: Survey and future directions

'Let me recommend… ': use of digital nudges or recommender systems for overweight and obesity prevention-a scoping review protocol

INTRODUCTION

Recommender systems, digital tools providing recommendations, and digital nudges increasingly affect our lives. The combination of digital nudges and recommender systems is very attractive for its application in preventing overweight and obesity. However, linking recommender systems with personalised digital nudges has a potential yet to be fully exploited. Therefore, this study aims to conduct a scoping review to identify which digital nudges or recommender systems or their combinations have been used in obesity prevention and to map these systems according to the target population, health behaviour, system classification (eg, mechanisms for developing recommendations, delivery channels, personalisation, interconnection, used combination), and system implementation.

METHODS AND ANALYSIS

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews guideline was used to inform protocol development. The eligibility criteria are based on the PCC framework (Population: any human; Concept: recommender systems or digital nudges; Context: obesity prevention). MEDLINE, PsycINFO, Web of Science, CINHAL, Scopus, ACM Digital Library and IEEE Xplore were searched until September 2023. Primary studies with any design published in peer-reviewed academic journals and peer-reviewed conference papers will be included. Data will be extracted into a pre-developed extraction sheet. Results will be synthesised descriptively and narratively.

ETHICS AND DISSEMINATION

No ethical approval is required for the scoping review, as data will be obtained from publicly available sources. The results of this scoping review will be published in a peer-reviewed journal, presented at conferences and used to inform the co-creation process and intervention adaptation in the context of a HealthyW8 project (www.healthyw8.eu).

A cascading model for nudging employees towards energy-efficient behaviour in tertiary buildings

Energy-related occupant behaviour in the built environment is considered crucial when aiming towards Energy Efficiency (EE), especially given the notion that people are most often unaware and disengaged regarding the impacts of energy-consuming habits. In order to affect such energy-related behaviour, various approaches have been employed, being the most common the provision of recommendations towards more energy-efficient actions. In this work, the authors extend prior research findings in an effort to automatically identify the optimal Persuasion Strategy (PS), out of ten pre-selected by experts, tailored to a user (i.e., the context to trigger a message, allocate a task or providing cues to enact an action). This process aims to successfully influence the employees' decisions about EE in tertiary buildings. The framework presented in this study utilizes cultural traits and socio-economic information. It is based on one of the largest survey datasets on this subject, comprising responses from 743 users collected through an online survey in four countries across Europe (Spain, Greece, Austria and the UK). The resulting framework was designed as a cascade of sequential data-driven prediction models. The first step employs a particular case of matrix factorisation to rank the ten PP in terms of preference for each user, followed by a random forest regression model that uses these rankings as a filtering step to compute scores for each PP and conclude with the best selection for each user. An ex-post assessment of the individual steps and the combined ensemble revealed increased accuracy over baseline non-personalised methods. Furthermore, the analysis also sheds light on important user characteristics to take into account for future interventions related to EE and the most effective persuasion strategies to adopt based on user data. Discussion and implications of the reported results are provided in the text regarding the flourishing field of personalisation to motivate pro-environmental behaviour change in tertiary buildings.

Behavioural 'nudging' interventions to reduce low-value care for low back pain in the emergency department (NUDG-ED): protocol for a 2×2 factorial, before-after, cluster randomised trial

INTRODUCTION

Opioids and imaging are considered low-value care for most people with low back pain. Yet around one in three people presenting to the emergency department (ED) will receive imaging, and two in three will receive an opioid. NUDG-ED aims to determine the effectiveness of two different behavioural 'nudge' interventions on low-value care for ED patients with low back pain.

METHODS AND ANALYSIS

NUDG-ED is a 2×2 factorial, open-label, before-after, cluster randomised controlled trial. The trial includes 8 ED sites in Sydney, Australia. Participants will be ED clinicians who manage back pain, and patients who are 18 years or over presenting to ED with musculoskeletal back pain. EDs will be randomly assigned to receive (i) patient nudges, (ii) clinician nudges, (iii) both interventions or (iv) no nudge control. The primary outcome will be the proportion of encounters in ED for musculoskeletal back pain where a person received a non-indicated lumbar imaging test, an opioid at discharge or both. We will require 2416 encounters over a 9-month study period (3-month before period and 6-month after period) to detect an absolute difference of 10% in use of low-value care due to either nudge, with 80% power, alpha set at 0.05 and assuming an intra-class correlation coefficient of 0.10, and an intraperiod correlation of 0.09. Patient-reported outcome measures will be collected in a subsample of patients (n≥456) 1 week after their initial ED visit. To estimate effects, we will use a multilevel regression model, with a random effect for cluster and patient, a fixed effect indicating the group assignment of each cluster and a fixed effect of time.

ETHICS AND DISSEMINATION

This study has ethical approval from Southwestern Sydney Local Health District Human Research Ethics Committee (2023/ETH00472). We will disseminate the results of this trial via media, presenting at conferences and scientific publications.

TRIAL REGISTRATION NUMBER

ACTRN12623001000695.

A Movie Recommender System Based on User Profile and Artificial Bee Colony Optimization

In this study, a new algorithm for recommending movies to viewers has been proposed. To do this, the suggested method employs data mining techniques. The proposed method includes three steps for generating recommendations: "preprocessing of user profile information," "feature extraction," and "recommendation." In the first step of proposed method, the user information will be examined and transformed into a form that can be handled in the next phases. In the second step of the proposed method, user attributes are then extracted as a collection of their individual qualities, as well as the average rating of each user for various genres. The bee colony optimization algorithm is then used to select the optimal features. Finally, in the third step of the proposed method, the ratings of similar users are utilized to offer movies to the target user, and the similarities between various users are determined using the characteristics calculated for them, as well as the Euclidean distance criteria. The proposed method was evaluated using the MovieLens database, and its output was assessed in terms of precision and recall criteria; these results show that the proposed method will increase the precision by an average of 1.39% and the recall by 0.8% compared to the compared algorithms.

Enhancing e-commerce recommendation systems through approach of buyer's self-construal: necessity, theoretical ground, synthesis of a six-step model, and research agenda

The current recommendation system predominantly relies on evidential factors such as behavioral outcomes and purchasing history. However, limited research has been conducted to explore the use of psychological data in these algorithms, such as consumers' self-perceived identities. Based on the gap identified and the soaring significance of levering the non-purchasing data, this study presents a methodology to quantify consumers' self-identities to help examine the relationship between these psychological cues and decision-making in an e-commerce context, focusing on the projective self, which has been overlooked in previous research. This research is expected to contribute to a better understanding of the cause of inconsistency in similar studies and provide a basis for further exploration of the impact of self-concepts on consumer behavior. The coding method in grounded theory, in conjunction with the synthesis of literature analysis, was employed to generate the final approach and solution in this study as they provide a robust and rigorous basis for the findings and recommendations presented in this study.

The Decoy Effect and Recommendation Systems

Recommendation systems and the decoy effect are two popular marketing techniques that have been used for facilitating decision making. Practitioners often use decoys to help drive demand for specific items, and prior research has shown the decoy effect to be robust in traditional choice settings, with consistent reporting of an overall positive impact. Recommendation systems are also increasingly being used to present item choice sets to customers and users, assisting users in their decision-making process. However, previous work has not examined the decoy effect in the context of recommendations. The decoy effect may facilitate consumer decision making and positively impact user behavior when used with recommendation systems. However, in the recommendation context, customers often have different expectations for the reliability and quality of the presented information. Hence, a decoy as a recommendation could signal issues in system reliability, resulting in a negative effect. Our study demonstrates that depending on the recommendation context, the decoy effect can be beneficial or counterproductive. Specifically, we find in the personalized context, including a decoy minimizes the demand for the target option and pushes consumers to opt out of purchase, which deviates from the traditional decoy effect. However, a decoy increases the target item’s demand in the nonpersonalized context, following the conventional decoy effect.

Transparent by choice: Proactive disclosures increase compliance with digital defaults

Default nudges successfully guide choices across multiple domains. Online use cases for defaults range from promoting sustainable purchases to inducing acceptance of behavior tracking scripts, or “cookies.” However, many scholars view defaults as unethical due to the covert ways in which they influence behavior. Hence, opt-outs and other digital decision aids are progressively being regulated in an attempt to make them more transparent. The current practice of transparency boils down to saturating the decision environment with convoluted legal information. This approach might be informed by researchers, who hypothesized that nudges could become less effective once they are clearly laid out: People can retaliate against influence attempts if they are aware of them. A recent line of research has shown that such concerns are unfounded when the default-setters proactively discloses the purpose of the intervention. Yet, it remained unclear whether the effect persists when defaults reflect the current practice of such mandated transparency boils down to the inclusion of information disclosures, containing convoluted legal information. In two empirical studies (N = 364), respondents clearly differentiated proactive from mandated transparency. Moreover, they choose the default option significantly more often when the transparency disclosure was voluntary, rather than mandated. Policy implications and future research directions are discussed.

Incorporating a Topic Model into a Hypergraph Neural Network for Searching-Scenario Oriented Recommendations

The personalized recommendation system is a useful tool adopted by e-retailers to help consumers to find items in line with their preferences. Existing methods focus on learning user preferences from a user-item matrix or online reviews after purchasing, and they ignore the interactive features in the process of users’ learning about product information through search queries before they make a purchase. To this end, this study develops a topic augmented hypergraph neural network framework to predict the user’s purchase intention by connecting the latent topics embedded in a consumer’s online queries to their click, purchase, and online review behavior, which aims at mining the connection information existing in the interaction graph domain. Meanwhile, in order to reduce the influence of text noise words by fusing topic information, we integrate the topic distribution and convolutional embedding to better represent each user and item, which can make up for the lack of topic information in traditional convolutional neural networks. Extensive empirical evaluations on real-world datasets demonstrate that the proposed framework improves the novelty of recommendation items as well as accuracy. From a managerial perspective, recommending diversified and novel items to consumers may increase the users’ satisfaction, which is conducive to the sustainable development of e-commerce enterprises.

Electronic nudge tool technology used in the critical care and peri-anaesthetic setting: a scoping review protocol

INTRODUCTION

Electronic clinical decision support (eCDS) tools are used to assist clinical decision making. Using computer-generated algorithms with evidence-based rule sets, they alert clinicians to events that require attention. eCDS tools generating alerts using nudge principles present clinicians with evidence-based clinical treatment options to guide clinician behaviour without restricting freedom of choice. Although eCDS tools have shown beneficial outcomes, challenges exist with regard to their acceptability most likely related to implementation. Furthermore, the pace of progress in this field has allowed little time to effectively evaluate the experience of the intended user. This scoping review aims to examine the development and implementation strategies, and the impact on the end user of eCDS tools that generate alerts using nudge principles, specifically in the critical care and peri-anaesthetic setting.

METHODS AND ANALYSIS

This review will follow the Arksey and O'Malley framework. A search will be conducted of literature published in the last 15 years in MEDLINE, EMBASE, CINAHL, CENTRAL, Web of Science and SAGE databases. Citation screening and data extraction will be performed by two independent reviewers. Extracted data will include context, e-nudge tool type and design features, development, implementation strategies and associated impact on end users.

ETHICS AND DISSEMINATION

This scoping review will synthesise published literature therefore ethical approval is not required. Review findings will be published in topic relevant peer-reviewed journals and associated conferences.

A Systematic Study on a Customer’s Next-Items Recommendation Techniques

A customer’s next-items recommender system (NIRS) can be used to predict the purchase list of a customer in the next visit. The recommendations made by these systems support businesses by increasing their revenue and providing a more personalized shopping experience to customers. The main objective of this paper is to provide a systematic literature review of the domain to analyze the recent techniques and assist future research. The paper examined 90 selected studies to answer the research questions concerning the key aspects of NIRSs. To this end, the main contribution of the paper is that it provides detailed insight into the use of conventional and deep learning techniques, the popular datasets, and specialized metrics for developing and evaluating these systems. The study reveals that conventional machine learning techniques have been quite popular for developing NIRSs in the past. However, more recent works have mainly focused on deep learning techniques due to their enhanced ability to learn sequential and temporal information. Some of the challenges in developing NIRSs that need further investigation are related to cold start, data sparsity, and cross-domain recommendations.

The Issue of Proxies and Choice Architectures. Why EU Law Matters for Recommender Systems

Recommendations are meant to increase sales or ad revenue, as these are the first priority of those who pay for them. As recommender systems match their recommendations with inferred preferences, we should not be surprised if the algorithm optimizes for lucrative preferences and thus co-produces the preferences they mine. This relates to the well-known problems of feedback loops, filter bubbles, and echo chambers. In this article, I discuss the implications of the fact that computing systems necessarily work with proxies when inferring recommendations and raise a number of questions about whether recommender systems actually do what they are claimed to do, while also analysing the often-perverse economic incentive structures that have a major impact on relevant design decisions. Finally, I will explain how the choice architectures for data controllers and providers of AI systems as foreseen in the EU's General Data Protection Regulation (GDPR), the proposed EU Digital Services Act (DSA) and the proposed EU AI Act will help to break through various vicious circles, by constraining how people may be targeted (GDPR, DSA) and by requiring documented evidence of the robustness, resilience, reliability, and the responsible design and deployment of high-risk recommender systems (AI Act).

Promoting Energy-Efficient Behavior by Depicting Social Norms in a Recommender Interface

How can recommender interfaces help users to adopt new behaviors? In the behavioral change literature, social norms and other nudges are studied to understand how people can be convinced to take action (e.g., towel re-use is boosted when stating that “75% of hotel guests” do so), but most of these nudges are not personalized. In contrast, recommender systems know what to recommend in a personalized way, but not much human-computer interaction ( HCI ) research has considered how personalized advice should be presented to help users to change their current habits.

We examine the value of depicting normative messages (e.g., “75% of users do X”), based on actual user data, in a personalized energy recommender interface called “Saving Aid.” In a study among 207 smart thermostat owners, we compared three different normative explanations (“Global.” “Similar,” and “Experienced” norm rates) to a non-social baseline (“kWh savings”). Although none of the norms increased the total number of chosen measures directly, we show that depicting high peer adoption rates alongside energy-saving measures increased the likelihood that they would be chosen from a list of recommendations. In addition, we show that depicting social norms positively affects a user’s evaluation of a recommender interface.

Digital Nudging for Online Food Choices

When people search for what to cook for the day, they increasingly use online recipe sites to find inspiration. Such recipe sites often show popular recipes to make it easier to find a suitable choice. However, these popular recipes are not always the healthiest options and can promote an unhealthy lifestyle. Our goal is to understand to what extent it is possible to steer the food selection of people through digital nudging. While nudges have been shown to affect humans' behavior regarding food choices in the physical world, there is little research on the impact of nudges on online food choices. Specifically, it is unclear how different nudges impact (i) the behavior of people, (ii) the time they need to make a decision, and (iii) their satisfaction and confidence with their selection. We investigate the effects of highlighting, defaults, social information, and warnings on the decision-making of online users through two consecutive user studies. Our results show that a hybrid nudge, which both involves setting a default and adding social information, significantly increases the likelihood that a nudged item is selected. Moreover, it may help decreasing the required decision time for participants while having no negative effects on the participant's satisfaction and confidence. Overall, our work provides evidence that nudges can be effective in this domain, but also that the type of a digital nudge matters. Therefore, different nudges should be evaluated in practical applications.

Towards psychology-aware preference construction in recommender systems: Overview and research issues

User preferences are a crucial input needed by recommender systems to determine relevant items. In single-shot recommendation scenarios such as content-based filtering and collaborative filtering, user preferences are represented, for example, as keywords, categories, and item ratings. In conversational recommendation approaches such as constraint-based and critiquing-based recommendation, user preferences are often represented on the semantic level in terms of item attribute values and critiques. In this article, we provide an overview of preference representations used in different types of recommender systems. In this context, we take into account the fact that preferences aren’t stable but are rather constructed within the scope of a recommendation process. In which way preferences are determined and adapted is influenced by various factors such as personality traits, emotional states, and cognitive biases. We summarize preference construction related research and also discuss aspects of counteracting cognitive biases.