Chatbot - beauty skin care products recommendations
Skincare products work differently depending on the type of skin. Therefore, the project proposes a context-aware chatbot for skincare product recommendations based on skin types. Firstly, we collect genuine product reviews dataset using a custom web crawler on cosmetic websites. The dataset is p...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project / Dissertation / Thesis |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://eprints.utar.edu.my/4739/1/fyp_IB_2021_LYK.pdf http://eprints.utar.edu.my/4739/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Skincare products work differently depending on the type of skin. Therefore, the project
proposes a context-aware chatbot for skincare product recommendations based on skin
types. Firstly, we collect genuine product reviews dataset using a custom web crawler
on cosmetic websites. The dataset is preprocessed to remove noises like null value,
incomplete reviews, and unverified reviews. Then, we built a sentiment analyzer based
on DistilBERT to rate beauty products based on the positive and negative scores from
the products reviews. Next, we train a skin type model to detect four skin types: dry,
oily, combination and natural using a CNN. Then, we trained a recommendation system
using a factorization machine to automatically recommend skincare products to users
based on the skin types. Lastly, we built a chatbot in Telegram for users to input their
facial image for skin detection and product recommendations. |
|---|