Bussiness Intelligence and Data Science for HyM enterprise

Developped by

• Pablo Reina Jiménez: https://github.com/preinaj, pabreijim1 ARROBA alum.us.es
• María Lourdes Linares Barrera: https://github.com/lourdesLB, marlinbar ARROBA alum.us.es

Project objectives and structure

1. Sentiment analysis on customers opinion about the brand
   Predicting model results and inference API to test: https://huggingface.co/lourdesLB/finetuning-sentiment-model With this model we got an score of 96% in predicting the sentiment of reviews made by the customers

2. Dashboards about sales tendency
   PowerBI dashboard app link: https://app.powerbi.com/view?r=eyJrIjoiZTk1NWI4ZDQtYmU1NC00MDcyLWFlNzQtZTVhZmM2MGIxNDYxIiwidCI6ImVmNGE2ODRlLTgxYjUtNDkxYy1hOThlLWM3YjMxYmU2YzQ2OSIsImMiOjh9
   This dashboard provides an useful insight about sales tendencies and customers satisfaction

3. Marketing campaigns analysis and sales prediction
   Models and data science analysis: marketing_sales_prediction/marketing_sales_prediction.ipynb
   This study provides an useful insight about marketing campaigns, the best way to invest in marketing and the preditcion of sales with a metric of error of only 2% and a metric of explainability of R^2 of 0.99. We also provide an overview of XAI over our regression models.

Technologies

Basic data science libraries

Web scraping

Models and predictions

Visualizations