Loading...

Design a Recommender System for Purchasing Cosmetics using Text Mining Methods

Ramezani Khozestani, Fatemeh | 2023

40 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56231 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Rafiee, Majid
  7. Abstract:
  8. In recent years, the cosmetics industry has dramatically grown in e-commerce. In e-commerce platforms, where multiple choices are available, an efficient recommender system is required to sort, order, and effectively transfer relevant content or product information to users. Recommender systems have attracted a lot of attention from retailers because they provide consumers with a personalized shopping experience. With technological advancements, this branch of artificial intelligence exhibits great potential in imaging, analysis, classification, and segmentation. Despite the great potential, the academic articles in this field are limited. Therefore, we conducted research in this context, in which we first examine the machine learning algorithms and matrix factorization. Then, we refer to the studies conducted based on recommender systems according to their applications, and due to the limited background, we examine some university articles in the literature review section. Finally, we implement XG Boost algorithm, Logistic Regression, Naive Bayes, Decision Tree, and Random Forest to design recommender systems for collaborative filtering. The accuracy obtained from these methods is respectively: 0.78, 0.82, 0.85, 0.87, and 0.82. Then, by using the matrix factorization method, user ratings, and cosmetics ingredients, design systems to provide appropriate recommendations to users in the least time. Also, we will use the effect of the Gray Wolf optimizer on machine learning algorithms so that the designed systems can be more reliable. After applying this optimizer, the outputs show that in all learning methods accuracy will reach 100% on the dataset.
  9. Keywords:
  10. Recommender System ; Machine Learning ; Matrix Factorization ; Collaborative Filtering ; Content Base Filtering ; Hybrid Filtering ; Gray Wolf Algorithm ; Text Mining

 Digital Object List

 Bookmark

No TOC