CS 838 Data Science Project (Spring 2017)

Perform entity matching in heterogeneous and structured data-sets derived from top restaurant review sites (Yelp and Zomato) for extracting insights.

Stage 0

• Form Project team : Tarun Bansal (tbansal@wisc.edu), Ayush Gupta (gupta92@wisc.edu) and Rohit Damkondwar (damkondwar@wisc.edu)

Stage 1

• Define the DS problem and collect structured data
• Report of this stage can be found here.

Stage 2

• Perform information extraction (IE) from natural text documents, using a supervised learning approach.
• Link to directory containing Text documents.
• Link to directory containing Text Documents in Set I (200 documents).
• Link to directory containing Text Documents in Set J (105 documents).
• Link to directory containing the code.
• Link to a compressed file that stores all of the above directories.
• Report of this stage can be found here.

Stage 3

• Entity Matching
• Link to directory containing the data
• Link to directory containing the code
• Report of this stage can be found here

Stage 4

• Data Merging
• E.csv
• Matched Tuples csv file
• Python Script to merge data columns
• Report

Stage 5

• Data Analysis
• E.csv
• Python Script to run prediction models
• Jupyter Nodebook
• Report

TEAM MEMBERS

Tarun Bansal (tbansal@wisc.edu), Ayush Gupta (gupta92@wisc.edu) and Rohit Damkondwar (damkondwar@wisc.edu)