DA 201 Introduction to Data Analytics
Fall2020
Assignment:Text and Sentiments
Total Points available =100
Goethe Industries.goes all Semantic..
You've done a bunch of work looking at the many traditional structured data sources to try and get a handle on where to look for answers, but somebody toldParsifalGoethethat it is often more valuable to look at the less structured data sources the company has access to. Principally the company wants to know what the world thinks about Goethe Industries ,including how many customers are considering churning or are very unhappy. This brings into play the related areas of text analytics and Sentiment Analysis.
Thankfully you do have some starting point you have the new Customer Outreach Record which does include some unstructured text (doesn't it?. But clearly that will no be enough.
For this last assignment you will come up with a plan to gather and analyze customer opinions.
Part 1: The COR
What will you do to extract opinions from the Customer Outreach Record data. Talk me through the steps you will take to get to customer sentiments. Issues to think about may include cleaning, extracting, parsing, topic clustering and performing sentiment analysis on this data.
Part2: TheOutside World
Now the tough part. What about the opinions of customers you have not had contact with or just people who may not be customers but may hold some kind of opinion about the companygood or bad.
Where will you look for this valuable information? Will you have to create new channels where customers/pundits can volunteer opinions? If so where will youplace these things and whhat will they be. Once suitable data channels are in place how will you farm them for information and (like part one) how will you extract vakuable opinion data from them... I think that's enough for this part
TheSentimentalPlan:100Points
Page length –6to1500pages excluding title page
Report Title:Report Title page including Group Name, authors and date.