This professional guide demonstrates a compact, repeatable workflow for common MongoDB tasks using a Zomato-style restaurants dataset. It covers:
- inserting sample documents,
- computing top restaurants by rating,
- counting reviews that mention the word “good”,
- fetching all reviews for a restaurant,
- updating and deleting records, and
- exporting results as JSON/CSV.
All commands are ready to paste into mongosh (except mongoexport, which runs in your system terminal). The examples assume a database named zomato and a collection named restaurants.
Dataset shape (example document)
Typical document structure used in the examples:
Note: the rate field in this dataset is a string of the form "4.1/5". Queries below show how to parse the numeric portion.
1. Insert at least 10 sample records
Use insertMany to add sample documents if you don’t already have data.
2. Top 5 restaurants by rating
Because rate is stored as "4.1/5", use $split + $toDouble to convert it to a numeric value, then aggregate and sort. The pipeline below computes an average per restaurant (useful if a restaurant has multiple documents).
What this returns: top 5 restaurant names with their average numeric rating and total votes.
3. Count reviews that contain the word “good” (case-insensitive)
If reviews_list is an array of subdocuments with a review text field:
Notes:
This counts documents (restaurants) where at least one review contains “good”.
If you need the exact count of matching reviews across all documents, use $unwind and $match then $count:
4. Retrieve all reviews for a specific restaurant (by name)
Replace "Jalsa" with the restaurant of interest.
5. Update a review and delete a record
Update one matching review (in-place)
Example: update a review in San Churro Cafe whose text mentions “Ambience”.
6. Delete a record (example: remove "Spice Elephant")
7. Export results as JSON/CSV
Important: mongoexport is a command-line tool. Run these commands in your system terminal (PowerShell, Bash, or Command Prompt), not inside mongosh.
Export entire collection as JSON
mongoexport --db=zomato --collection=restaurants --out=restaurants.json
Export selected fields as CSV
mongoexport --db=zomato --collection=restaurants \
--type=csv --fields=name,location,rate,votes,cuisines \
--out=restaurants.csv
Export the Top 5 aggregation result (recommended approach: write the aggregation to a temporary collection, then export):
Write the aggregation to a collection named top5_restaurants:
Export that collection:
mongoexport --db=zomato --collection=top5_restaurants --out=top5_restaurants.json
mongoexport --db=zomato --collection=top5_restaurants --type=csv --fields=_id,location,avgRating,totalVotes --out=top5_restaurants.csv
Authentication / remote server: if your MongoDB uses authentication or is remote, supply --uri with the full connection string:
mongoexport --uri="mongodb://user:pass@host:27017/zomato" --collection=restaurants --out=restaurants.json
Practical notes & best practices
- Run mongoexport from the OS terminal — not from mongosh. Running it inside the shell causes a syntax error.
- Normalize numeric fields: if you will frequently query by rating, consider storing rate_num as a numeric field (e.g., 4.1) during ingestion to simplify queries and improve performance.
- Text search: for more advanced review searches (stemming, ranking), consider enabling text indexes and using $text.
Backups: keep JSON exports of critical collections for reproducibility and backup.
Closing
This guide supplies a compact, production-minded set of queries and commands you can use to demonstrate MongoDB skills on the Zomato dataset. Use the commands as-is in mongosh and run the mongoexport commands from your terminal to produce JSON/CSV deliverables.
Top comments (0)