SIT719 Security and Privacy Issues in Analytics

AssignmentTutorOnline

Distinction/High Distinction Task 9.1: Location-based Privacy Protection

Overview

Trajectory data is powerful to many crowdsourcing tasks. For example, Uber and other services use the drivers’ geolocation to match the client’s requests. However, there are serious concerns about the privacy of publishing the geolocation data.

In this Distinction/Higher Distinction Task, you will experiment with machine learning classification algorithms. Please see more details in the Task description. Before attempting this task, please make sure you are already up to date with all previous Credit and Pass tasks.

Task Description

Instructions:

Suppose that you are hired by a large company that uses the user’s geolocation data as references for allocating crowdsourcing tasks. The company has developed good algorithms for allocating tasks based on accurate geolocation data. One of the new business requirements is that each client can visualize a few nearby crowdsourcing workers before the client finalizes the order. Displaying the geolocation on Google Maps and alike services is doable. However, the accurate geolocation data is sensitive and cannot be directly disclosed to the clients. Therefore, the boss has requested you to develop a demo system to protect the privacy of the geolocation data.

Since this is a demo system, the famous dataset named “Gowalla” is provided to simulate the crowdsourcing workers, which is available at https://snap.stanford.edu/data/loc-gowalla.html. The Gowalla dataset consists of multiple users’ check-in data with timestamps in five columns, some sample data look like this:

[user] [check-in time]                      [latitude]         [longitude]     [location id] 196514 2010-07-24T13:45:06Z                       53.3648119             -2.2723465833                                   145064

196514	2010-07-24T13:44:58Z	53.360511233	-2.276369017	1275991
196514	2010-07-24T13:44:46Z	53.3653895945	-2.2754087046	376497
196514	2010-07-24T13:44:38Z	53.3663709833	-2.2700764333	98503
196514	2010-07-24T13:44:26Z	53.3674087524	-2.2783813477	1043431
196514	2010-07-24T13:44:08Z	53.3675663377	-2.278631763	881734
196514	2010-07-24T13:43:18Z	53.3679640626	-2.2792943689	207763

196514 2010-07-24T13:41:10Z     53.364905        -2.270824        1042822

You will need to download the dataset, familiarize with it before performing the following actions:

1. Find three privacy protection methods to protect location-based data from at least three published papers, including the papers published on arxiv.org.
2. Write a short literature review (approximately 500 words) to compare the identified methods in at least three aspects that are relevant to data privacy and utility.
3. Identify meaningful performance metrics based on your critical literature review and comparison before proposing how to measure these metrics.
4. Implement or apply the existing implementations of privacy protection methods on the Gowalla dataset.
5. Report the performance metrics that are identified in Step 3.
6. Demonstrate the proposed solutions with a few case studies using Google maps.

A simple illustration of your demo may look like the following, where Pi are the crowdsourcing workers, ai are the clients:

Once you have completed the above steps of the project, you need to deliver the outcome. In real-world, results are typically delivered as a product/tool/web-app or through a presentation or by submitting the report. However, in our unit, we will consider a report and a demo only.

Here, you need to write a report (at least 2,000 words including the above-mentioned literature review) based on the outcome and results you obtained by performing the above steps. The report will describe the literature review, the algorithms used, their working principle, key parameters, and the results. Results should consider all the key performance measures and comparative results in the form of tables, graphs, etc. The demo should be a 5-minutes long pre- recorded presentation.

Submit the PDF report and the demo PPT through OnTrack. You also need to submit the code separately (within the “Code for task 9.1” folder) under the assignment tab of the CloudDeakin Python script(s) during submission.

Marking Rubric:

Criteria	Unsatisfactory – Beginning	Developing	Accomplished	Exemplary	Total
Report Focus: Purpose/ Position Stateme nt	0-7 points	8-11 points	12-15 points	16-20 points	/20
Fails to clearly relate the report topic or is not clearly defined and/or the report lacks focus throughout.	The report is too broad in scope (outside of the title topic) and/or the report is somewhat unclear and needs to be developed further. Focal point is not consistently maintained throughout the report.	The report provides adequate direction with some degree of interest for the reader. The report states the position, and maintains the focal point of the analysis for the most part.	The report provides direction for the discussion part of the analysis that is engaging and thought provoking, The report clearly and concisely states the position, and consistently maintain the focal point.
Compar ative analysis and Discussi on	0-15 points	16-20 points	21-24 points	25-30 points	/30
Demonstrates a lack of understanding and inadequate knowledge of the topic. Analysis is very superficial and	Demonstrates general understanding of python scripting. Analysis is good and has addressed all criteria. Comparative	Demonstrates good level of understanding of python scripting. Algorithms are fine-tuned and comprise good	Demonstrates superior level of understanding of python scripting and algorithms. Algorithms are fine-tuned with some
	contains flaws. The	analysis is presented.	selection of algorithms.	novelty or hybridization or
	report is also not clear.	Sufficient discussion is	Comparative results are	advanced and/or recent
		also presented.	presented using standard	algorithm. Comparative
			performance measures.	results are presented
				using performance
				measures in a way that it
				provides very clear and
				meaningful insights of the
				output.
Demonst ration	0-6 points	7-11 points	12-15 points	16-20 points	/20
Demonstration lacks	Demonstration includes a	Demo is working and	Professionally conducted
	coherent ideas and fails	working system,	clearly explained,	demo, free from errors,
	to demonstrate a	however, the benefits of	however, there might be	excellent talk with deep
	working system.	privacy protection are not	occasional mistakes or	knowledge on privacy
		clearly presented.	difficult points to	protection for location
			understand.	data.
Writing Quality & Adheren ce to Format Guidelin es	0-10 points	11-17 points	18-21 points	22-30 points	/30
Report shows a below average/poor writing style lacking in elements of appropriate standard English. Frequent errors in spelling, grammar, punctuation, spelling, usage, and/or formatting.	Report shows an average and/or casual writing style using standard English. Some errors in spelling, grammar, punctuation, usage, and/or formatting.	Report shows above average writing style (can be considered good) and clarity in writing using standard English. Minor errors in grammar, punctuation, spelling,	Article is well written and clear and standard English characterized by elements of a strong writing style. Basically free from grammar, punctuation, spelling,
	usage, and/or formatting.	usage, or formatting
	Author has demonstrated	errors.
	the use of scientific	Author has demonstrated
	language and results are	advanced use of
	well explained.	scientific language and
		results are well explained
		with insights.

Rubric adopted from: Denise Kreiger, Instructional Design and Technology Services, SC&I, Rutgers University, 4/2014