Structured Zomato Restaurant Metadata and Review Corpus for Sentiment and Recommendation Analysis

Open the Dataset:

• Use spreadsheet software like Excel, Google Sheets, or load using Python libraries such as Pandas.

• Example (Python):

  import pandas as pd
df = pd.read_csv("zomato_sample.csv")


Understand the Columns:

• Key columns include:

  • Restaurant Name, Location, Cuisines, Rating, Cost, Reviews, etc.

  • The Reviews column contains a list of tuples with both ratings and review text.

Review Parsing (Python):

• To extract the review data (rating + text), use:

   import ast
df['parsed_reviews'] = df['Reviews'].apply(lambda x: ast.literal_eval(x))


Sentiment Analysis (Optional):

• Use NLP libraries like TextBlob or VADER to analyze customer sentiments.

   from textblob import TextBlob
df['sentiment'] = df['parsed_reviews'].apply(lambda reviews: TextBlob(reviews[0][1]).sentiment.polarity)


Filter by Category or Cuisine:

• Example:

  df[df['Cuisines'].str.contains("North Indian", na=False)]
df[df['Zone'].str.contains("Whitefield", na=False)]


Visualization (Optional):

• Use Power BI, Tableau, or Python libraries (Matplotlib, Seaborn, Plotly) for visual exploration and storytelling.