MIS 354 Final Project Walkthrough | NLP Sentiment Analysis & Market Basket Analysis | No-Code AI

MIS 354 – No-Code AI & Machine Learning final project walkthrough.

This video covers two complete analytics pipelines built without writing a single line of code.

——————————————————
Project 1 – NLP Sentiment Analysis & Topic Modeling (Orange)
Dataset: Women's E-Commerce Clothing Reviews (Kaggle)

- 6,793 reviews analyzed with a 6-step preprocessing pipeline
- LDA topic modeling with 5 topics, Coherence 0.358
- VADER sentiment analysis — unexpected finding: 1-star reviews average a compound score of 0.238, revealing that fashion customers express dissatisfaction in polite language that standard sentiment tools misclassify as neutral

——————————————————
Project 2 – Market Basket Analysis (KNIME)
Dataset: Online Retail Dataset (Kaggle)

- 397,924 transactions across 18,536 unique baskets
- Apriori algorithm, 30 rules generated, all Lift above 1
- Unexpected finding: The Regency Teacup series (Pink, Green, Roses) shows full six-way bidirectional symmetry with Lift up to 24.033 — customers buy all three variants simultaneously in a single gift session

——————————————————
Tools: Orange Data Mining, KNIME Analytics Platform
Data: Kaggle external datasets only

Furkan Ay | Student ID: 230106003
Ostim Technical University — MIS 354, May 2026
——————————————————

Видео MIS 354 Final Project Walkthrough | NLP Sentiment Analysis & Market Basket Analysis | No-Code AI канала Furkan Ay