Crowd-enabled Processing of Trustworthy, Privacy-Enhanced and Personalised Location Based Services with Quality Guarantee

Abstract

We propose a novel approach for enabling trustworthy, privacy-enhanced and personalised location based services (LBSs) that find nearby points of interests (POIs) such as restaurants, ATM booths, and hospitals in a crowdsourced manner. In our crowdsourced approach, a user forms a group from the crowd and processes the LBS using the POI knowledge of the group members without involving an external service provider. We use personalised rating in addition to the distance of a POI for finding the answers of the location based queries. The personalised rating of a POI is computed using individual POI ratings given by the group members and the query requestor's trust and similarity scores for the group members. The major challenges for the crowdsourced data are incompleteness and inaccuracy, which may result in lower quality answer for the LBS. In this paper, we first present techniques to select knowledgeable group members for processing LBSs and thereby increase the accuracy and the confidence level of the query answers. We then develop efficient algorithms to process LBSs in real time and enhance privacy by reducing the number of the group members' POIs shared with the query requestor. Finally, we run extensive experiments using real datasets to show the efficiency and effectiveness of our approach.

Accountable AI for Healthcare IoT Systems

“…Reputation, here refers to how well something is known in a community of users. Similar methods are used to find the credibility of the source in crowd-sourced data [38]. Blockchain, one of the highly successful methods in securing data for financial as well as many other domains, runs along the similar line.…”