Assignment 2 - Sentiment Analysis

Topic

Sentiment analysis is contextual examination of words, and words within sentences across all

social media platforms and web sites, in order to reveal the sentiment of the writer.

All texts (unstructured texts) are mined to gain insight into various subjects of interests with

regard to the degree of positivity/negativity revealed in the text. In this manner, unstructured

texts are transformed into a quantitative scale and the results can be presented in charts and

tables so that important information can be revealed and easily digested.

The purpose of this assignment is to give you an experience of how sentiment analysis works at a

rudimentary level. In reality, the analysis is done by software applications that examine each

word within the context of the sentence, along with other metrics on the frequency/count metrics,

sources, demographics, key words, etc. The analysis is then connected to data visualization

tools to create dashboards with charts and alerts that are updated as new messages are posted.

In this assignment, learn from the example of a sentiment analysis done manually in word and

excel. Using this example, perform sentiment analysis on different product, based on reviews on

a beauty product (moisturizer) known for its natural ingredients and affordability and the

endorsement of Jessica Alba.

You are provided with a sample set of 40 product reviews on this moisturizer.

Using this data

set for your assignment,

1. create a lexicon to score each product review,

2. capture key words,

3. calculate the NPS,

4. present the results using a variety charts.

You can (and should) go further than the

example provided.

5. Finally, drawing from the results of the analysis, form conclusion and recommendations

that will help enhance customer perceptions, and thereby, customer acquisitions.