Collaborating with Users in Proximity for Decentralized Mobile Recommender Systems

Beierle, Felix; Eichinger, Tobias

doi:10.1109/smartworld-uic-atc-scalcom-iop-sci.2019.00222

Cited by 8 publications

(7 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Barbosa et al [2] propose that smartphones exchange data between devices and calculate their own recommendation via collaborative filtering. Beierle and Eichinger [3] further present a mobile architecture consisting of data collection, data exchange, and a local recommender system; the data collection component gets data about the user from local device, data exchange gets data about other users from other devices, and the local recommender system utilizes all available data for recommending items to the user. Several studies have introduced federated learning [33] into the realm of recommendation, which provides a way to realize federated recommender systems.…”

Section: Federated Recommender Systemsmentioning

confidence: 99%

See 1 more Smart Citation

Meta Matrix Factorization for Federated Rating Predictions

Lin

Ren

Chen

et al. 2020

Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

Federated recommender systems have distinct advantages in terms of privacy protection over traditional recommender systems that are centralized at a data center. With the widespread use and the growing computing power of mobile devices, it is becoming increasingly feasible to store and process data locally on the devices and to train recommender models in a federated manner. However, previous work on federated recommender systems does not fully account for the limitations in terms of storage, RAM, energy and communication bandwidth in a mobile environment. The scales of the models proposed are too large to be easily run on mobile devices. Also, existing federated recommender systems need to fine-tune recommendation models on each device, which makes it hard to effectively exploit collaborative filtering information among users/devices.Our goal in this paper is to design a novel federated learning framework for rating prediction (RP) for mobile environments that operates on par with state-of-the-art fully centralized RP methods. To this end, we introduce a federated matrix factorization (MF) framework, named meta matrix factorization (MetaMF), that is able to generate private item embeddings and RP models with a meta network. Given a user, we first obtain a collaborative vector by collecting useful information with a collaborative memory module. Then, we employ a meta recommender module to generate private item embeddings and a RP model based on the collaborative vector in the server. To address the challenge of generating a large number of high-dimensional item embeddings, we devise a rise-dimensional generation strategy that first generates a low-dimensional item * Co-corresponding author.

show abstract

Section: Federated Recommender Systemsmentioning

confidence: 99%

“…1 We use the code released by the respective authors 2 for AutoRec. We use LibRec 3 for the remaining baselines.…”

Section: Implementation Detailsmentioning

confidence: 99%

Meta Matrix Factorization for Federated Rating Predictions

Lin

Ren

Chen

et al. 2020

Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

show abstract

“…The WiFi Alliance developed another WiFi mode for device-to-device communication, WiFi Direct, which is supported on Android devices with version 4.0 and higher [39]. 8 There are several related papers dealing with WiFi Direct for device-to-device communication [18], [37], [40]- [43]. While the Talk2Me prototype was developed utilizing WiFi Direct, Shu et al describe how it is not mature enough and wasn't used for the evaluation [18].…”

Section: B Wifimentioning

confidence: 99%

“…1 • An extensive evaluation of the device-to-device connection module. We sketched the general idea of MobRec in our work-inprogress paper [8]. This paper builds on the results of [8] and extends it with the details regarding device-to-device communication, and the implementation and evaluation of MobRec.…”

Section: Introductionmentioning

confidence: 99%

MobRec — Mobile Platform for Decentralized Recommender Systems

Beierle

Egger

2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

Recommender systems recommend new movies, music, restaurants, etc. Typically, service providers organize such systems in a centralized way, holding all the data. Biases in the recommender systems are not transparent to the user and lock-in effects might make it inconvenient for the user to switch providers. In this paper, we present the concept, design, and implementation of MobRec, a mobile platform that decentralizes the data collection, data storage, and recommendation process. MobRec's architecture does not need any backend and solely consists of the users' smartphones, which already contain the users' preferences and ratings. Being in proximity in public places or public transportation, data is exchanged in a device-to-device manner, building local databases that can recommend new items. One of biggest challenges of such a system is the implementation of unobtrusive device-to-device data exchange on off-the-shelf Android devices and iPhones. MobRec facilitates such data exchange, building on Google Nearby Messages with Bluetooth Low Energy. We achieve the successful exchange of data within 3 to 4 minutes, making it suitable for the described scenario. We demonstrate the feasibility of decentralized recommender systems and provide blueprints for the development of seamless multi-platform device-to-device communication.

show abstract

“…In the present work, we propose a method for disseminating recommendations epidemically in combination with an on-device filtering process for which we proposed a mobile software architecture in [5]. The most similar work is that of Barbosa et al [25] that propose device-to-device raw profile exchanges in an opportunistic networking scenario.…”

Section: Related Workmentioning

confidence: 99%

On gossip-based information dissemination in pervasive recommender systems

Eichinger

Beierle

Papke

et al. 2019

Proceedings of the 13th ACM Conference on Recommender Systems

Self Cite

View full text Add to dashboard Cite

Pervasive computing systems employ distributed and embedded devices in order to raise, communicate, and process data in an anytime-anywhere fashion. Certainly, its most prominent device is the smartphone due to its wide proliferation, growing computation power, and wireless networking capabilities. In this context, we revisit the implementation of digitalized word-of-mouth that suggests exchanging item preferences between smartphones offline and directly in immediate proximity. Collaboratively and decentrally collecting data in this way has two benefits. First, it allows to attach for instance location-sensitive context information in order to enrich collected item preferences. Second, model building does not require network connectivity. Despite the benefits, the approach naturally raises data privacy and data scarcity issues. In order to address both, we propose Propagate and Filter, a method that translates the traditional approach of finding similar peers and exchanging item preferences among each other from the field of decentralized to that of pervasive recommender systems. Additionally, we present preliminary results on a prototype mobile application that implements the proposed device-to-device information exchange. Average ad-hoc connection delays of 25.9 seconds and reliable connection success rates within 6 meters underpin the approach's technical feasibility.

show abstract

Collaborating with Users in Proximity for Decentralized Mobile Recommender Systems

Cited by 8 publications

References 26 publications

Meta Matrix Factorization for Federated Rating Predictions

Meta Matrix Factorization for Federated Rating Predictions

MobRec — Mobile Platform for Decentralized Recommender Systems

On gossip-based information dissemination in pervasive recommender systems

Contact Info

Product

Resources

About