Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/kirbytes/ecommerce-analytic-platform
DEMO for ISOM671 course project
https://github.com/kirbytes/ecommerce-analytic-platform
aws dataengineering frontend-web nosql python sql
Last synced: 12 days ago
JSON representation
DEMO for ISOM671 course project
- Host: GitHub
- URL: https://github.com/kirbytes/ecommerce-analytic-platform
- Owner: Kirbytes
- Created: 2024-12-09T17:14:22.000Z (23 days ago)
- Default Branch: main
- Last Pushed: 2024-12-09T17:22:24.000Z (23 days ago)
- Last Synced: 2024-12-20T17:14:32.123Z (12 days ago)
- Topics: aws, dataengineering, frontend-web, nosql, python, sql
- Language: Python
- Homepage:
- Size: 1.94 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
Awesome Lists containing this project
README
# Alibaba E-commerce Analytics Repository
## Project Description
Data warehouse and analytics dashboard for Alibaba e-commerce transaction data.
## Directory Structure
```
.
├── data/ # Data files
│ ├── raw/ # Raw data
│ │ └── round1_ijcai_18_train_20180301.txt
│ └── processed/ # Cleaned data
│ ├── clicked_sample.csv
│ ├── advertising_item.csv
│ ├── user_table.csv
│ ├── context_table.csv
│ └── shop_table.csv
├── sql/ # SQL scripts
│ ├── dataloader.sql
│ └── transform_data_for_dashboard.sql
├── src/ # Python source code
│ ├── datacleaner.py
│ ├── fact_sales_load.py
│ └── dashboard.py
└── README.md
```
## Data Pipeline
### 1. Data Cleaning
```python
# src/datacleaner.py
- Split raw data into normalized tables
- Handle special delimiters
- Export clean CSVs
```
### 2. Database Setup
```sql
-- sql/dataloader.sql
- Create initial tables
- Load dimensions
- Set constraints
```
### 3. Fact Table ETL
Due to the limited computational resources of the MySQL database, the fact table is generated using Python, then loaded into the database.
```python
# src/fact_sales_load.py
- Aggregate metrics
- Calculate conversions
- Export fact table
```
### 4. Star Schema
```sql
-- sql/transform_data_for_dashboard.sql
- Create dimensional model
- Time dimension handling
- Fact table structure
```
### 5. Dashboard
The dashboard in constructed using Tableau, with data source connected to MySQL.
## Database Schema
### Fact Table
- **fact_sales**
- time_key (FK)
- item_id (FK)
- user_id (FK)
- shop_id (FK)
- total_views
- total_purchases
- conversion_rate
### Dimension Tables
- **dim_time**
- time_key (PK)
- full_datetime
- date_only
- hour_of_day
- is_peak_hour
- is_night
- day_of_week
- is_weekend
- time_segment
- **dim_product**
- item_id (PK)
- category_list
- property_list
- brand_id
- city_id
- price_level
- sales_level
- collected_level
- pv_level
- **dim_user**
- user_id (PK)
- gender_id
- age_level
- occupation_id
- star_level
- **dim_shop**
- shop_id (PK)
- review_num_level
- review_positive_rate
- star_level
- score_service
- score_delivery
- score_description
## Pipeline
Note that file paths need to be adjusted to the local environment.
```bash
# Clean data
python src/datacleaner.py
# Generate fact table
python src/fact_sales_load.py
# Load database
mysql -u root -p < sql/dataloader.sql
mysql -u root -p < sql/transform_data_for_dashboard.sql
# Launch dashboard
streamlit run src/dashboard.py
```