case study twitter feed customer sentiment analysis data curry
DESCRIPTION
Customer sentiment analysis using twitter feed and unstructured data, separating data from the noise and extracting insights....TRANSCRIPT
Extract user sentiments from Twitter data – Implementation Case Study
2
CASE STUDYMeasure the performance of a product using unstructured text from Twitter
3
The problem statement
Twitter being one of the most active and real time social networking tool, it swiftly reflects the mood of users of a product or service.
Here, we measure the performance of a newly released laptop using the free-form text from Twitter
Obtaining user emotions for specific features of laptop rather than the laptop as a whole
Take preventive actions in order to provide proactive customer services and thereby improve the product sales by end first Quarter
4
Challenges
Huge volumes of free-form data
Processing the tweets: twitter lingo constantly evolves, new names and characterizations flare up all the time, which excludes straightforward full-text analysis.
Users express their emotions on diverse issues related to a particular feature or several features of the product
Classify based on user emotions as positive, negative or neutral.
5
Solution approach/methodology
Extract Tweets
Pre-process
Algorithmic scoring of
tweets
Feature based sentiment extraction
Visualize the results
6
Results and analysis
Extracted the Key features of the product
Rating of the importance of each feature
Sentiment score on the feature is analyzed
7
Key features extracted
8
Sentiment by key features
Camera Battery RAM Memory Reader Digital Media Bluetooth Speed0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Neutral
Negative
Positive
Key features of the laptop
% se
ntim
ent f
rom
the
data
9
Summary, conclusions and/or benefits
Pin-pointed the potential causes of negative sentiments
Identified the broader trends on the perception
Direct engagement rate with customers increased
Thank You
10