Three-tier IoT-edge-cloud (3T-IEC) architectural paradigm for real-time event recommendation in event-based social networks

Autor: Pratibha Mahajan, Pankaj Deep Kaur
Rok vydání: 2020
Předmět:
Zdroj: Journal of Ambient Intelligence and Humanized Computing. 12:1363-1386
ISSN: 1868-5145
1868-5137
DOI: 10.1007/s12652-020-02202-9
Popis: With the popularization of internet, event-based social networks (EBSNs) have experienced recognition among people for planning and communicating social events. Due to plethora of events occurring over EBSNs and time varying user interests, different recommendation techniques have been employed to make suitable events suggestions to participants. The state of the art event recommendation methods have not explored the importance of real-time data while generating event recommendations. However, we strongly believe that performance of event recommendation in EBSN can significantly be improved if multiple parameters captured through IoT devices may be considered. In this work, a real-time event recommendation problem which involves monitoring user’s current location, present road traffic, and weather conditions is addressed that adopts instant event recommendation. To address this problem, a novel three-tier IoT-edge-cloud based solution for real-time context-aware event recommendation problem named as 3T-IEC has been proposed. The 3T-IEC introduces edge computing layer where IoT data is processed for deriving contextual IoT-based location information along with event recommendation generator in the cloud layer. Further, contextual information such as user’s current location, weather and temporal feasibility has been applied to filter the events. Furthermore, group, category, and economic influences are modeled on filtered events to rank them with help of multiple criteria decision making method. Moreover, personalized weights on influential factors are also learned by using distance method. For practical realization, an android based mobile application-SpotEvent has been developed. Furthermore, the qualitative and quantitative analysis of the 3T-IEC is performed on two real-world datasets acquired from Meetup. The results clearly indicate that recommendation quality of proposed system is better, when compared to its variants and other baseline methods such as VSM (260%), SVD (144%), CAER (335%), Skyline (100%) and SoCast* (16%).
Databáze: OpenAIRE
Externí odkaz: