Objective

This project aims to dive into LinkedIn reviews sourced from google play store via kaggle data.Using python, I explore distribution of ratings, identify common themes in user comments and employe sentiment analysis techniques. The goal is to extract actionable insights for understanding user satisfaction and areas for app improvement.

Approach:

1. Data Collection and Preparation:

I acquired the dataset from Kaggle in CSV format. I initiated the data analysis process by importing the necessary libraries for data manipulation, analysis and visualization as illustrated below:

After importing the CSV file into Jupyter Notebook, I proceeded to create a new dataset with the two columns which formed the foundation of this analysis. I isolated the 'review_text' column containing user comments and 'rating' column reflecting user-provided ratings:

2. Sentiment Analysis: Cleaning Text in the 'review_text' column

In this phase of sentiment analysis, the focus is on prepairing the text data for analysis. The 'review_text' column, undergoes a cleaning process to standardize the text for better analysis.

Here is a breakdown process of the cleaning steps I performed:

1. Lowercasing: All text is converted to lowercase to ensure uniformity and consistency.
2. Removing brackets, URLs and HTML tags: Using regular expressions (re.sub()), square brackets along with their contents, URLs, HTML tags are removed from the text.
3. Eliminating Punctuations: Punctuations characters are removed to simplify the text.
4. Removing words with numbers: Words containing numbers are filtered out of the text.
5. Stopword Removal: Stopwords (common words like "is", "the", "and" etc) are removed from the text to focus on the meaningful content.
6. Stemming: Words are reduced to their base form using the SnowballStemmer. The process aids in normalizing variations of words.

3. Analyzing Rating Distribution

The objective here is understanding the distribution of ratings provided by users for the linked in:

Results:

Observations: 58.3% of LinkedIn users rate it a five and 4.64% of the users rate it a one

4. Extracting Frequent Words

This step finds the frequent words which helps in recognizing topics discussed by users, highlight area of focus for sentiment analysis, and the visualization gives a quick understanding of users priorities and concerns.

5. Sentiment Analysis

I assigned numerical scores for positive, negative and neutral sentiments, which quantifies the emotional tone of each review textThis helps to identify if the majority of the comments are positive, negative or neutral as shown below:

Interpration:

The majority of the reviews exhibit a positive sentiment reflected in the higher positive score values (ranging 0.4-1.0) in comparison to neutral and negative scores.

There are no negative scores (all negatives scores are 0.0) indicating an absense of no negative opinions in this particular reviews.

The absence of negative sentiment indicate a generally favaroble users experience with the LinkedIn app, aligning more with satisfaction and positive feedback

6. Visualizing words used in the Positive Review

The essence of identifying common words used in positive reviews plays a role in guiding product development, marketing strategies and the overall efforts to enhance user satisfaction and app performance

7. Visualizing words used in the Negative Review

The essence of identifying common words used in negative reviews plays a role in guiding improvements

8. Conclusion

Benefits of Sentiment Analysis

Sentiment analysis offers multifaceted benefits for organizations across various domains:

1. Enhanced Customer Understanding: By analyzing sentiments, organizations gain deeper insights into user perceptions, preferences, and pain points related to their brand, services, or products . This understanding aids in tailoring offerings to meet user expectations effectively.
2. Proactive Reputation Management: Identifying and addressing negative sentiments in a timely manner enables organizations to manage their online reputation effectively. Prompt responses to concerns can mitigate reputational risks and demonstrate a commitment to user satisfaction.
3. Product and Service Improvement: Sentiment analysis guides product or service enhancements aligned with user needs. It helps in identifying specific features that resonate positively or areas requiring refinement, fostering continuous improvement cycles.
4. Strategic Decision-Making: Insights from sentiment analysis inform strategic decisions, aiding in resource allocation, marketing strategies, brand positioning, and long-term planning. It ensures that organizational strategies align with user sentiments and preferences.