1. Big Data Analytics: HW#3
By J. H. Wang
May 16, 2017

2. Programming Exercise: MapReduce
Goal: A MapReduce program for analyzing the check-in records
Input: Check-in records in social networking site foursquare
Check-in records: <user_id, venue_id, checkin_time>
Venues: <venue_id, category, latitude, longitude>
Output: analysis results (to be detailed later)

3. Output Detailed analysis including the following tasks:
Lists the top checked-in venues (most popular)
Lists the top checked-in users
Lists the most popular categories
Lists the most popular time for check-ins (in time slots in hours, for example, 7:00-8:00 or 18:00-19:00)
All of the above results *should* be sorted by the number of check-ins
You also have to output the efficiency (running time) of each task

4. Note on Programming Exercises
Programming exercises can be done as a team (at most two persons per team)
You can use any programming language in Hadoop or Spark to implement
Java, Scala, Python, or R

5. Optional functions More analysis Further extension
Using other attributes: latitude, longitude, …
Further extension
Automatically integrating more information by your program
venue names and information from the <venue_id>, <latitude>, <longitude>, …
More information about venue_id:
Geographic coordinates: for example, converting latitude, longitude to address …

6. Homework Submission
For implementation projects, please submit a compressed file containing:
Your cluster environment setup
How many PCs, what spec, network setup, …
Your source codes
Documentation on how to compile, install, or configure the environment.
Due: 3 weeks (Jun. 6, 2017)

7. Evaluation
In completion of each of the four tasks, you get part of the scores
The efficiency of implemented algorithm will also be counted
Optional functions will get extra credits
You might need to demo if the program was unable to run

8. Thanks for Your Attention!