Practical Crowd-sensing Applications with Privacy Protection

PhD Thesis Proposal Defence


Title: "Practical Crowd-sensing Applications with Privacy Protection"

by

Mr. Shanfeng ZHANG


Abstract:

With the phenomenal growth of smartphones, wearable devices and
sensor-quipped vehicles, mobile device users have the ability to acquire
local information such as location, traffic conditions, pollution and so
on. When the sensing information is share with the cloud where data fusion
takes place, a myriad of crowd-sensing applications has appeared,
including vehicle navigation, environment monitoring and social networks.
However, some topics remain hard to be solved.

In this thesis proposal, we address three challenging topics in this area
named urban taxi-sharing, emotion detection and location privacy
protection. In the first work, we design a QoS-aware taxi-sharing system
named QA-Share which allows occupied taxi to pick up new passengers on the
fly. By dynamically collecting location information of taxi riders and
drivers, QA-Share schedules routines for taxis to reduce waiting time for
taxi riders and increase productivity for drivers. In the second work, we
propose iSelf, which automatically detects users' emotions in cold-start
conditions. iSelf collects usage pattern of smartphones besides location.
Given only a few labeled samples, we use transfer learning technology for
emotion labeling. In the third work, we present a novel crowd-sensing
scheme PLP, which preserves privacy when collecting location information
from users. PLP aims to maximizes the amount of location data collection
by filtering a user's context stream.


Date:			Friday, 22 May 2015

Time:                   3:00pm - 5:00pm

Venue:                  Room 2126B
                         lift 19

Committee Members:	Prof. Lionel Ni (Supervisor)
  			Prof. Gary Chan (Chairperson)
 			Dr. Qiong Luo
  			Dr. Ke Yi


**** ALL are Welcome ****