260 likes | 522 Views
COE2013-5. The Development of Book Recommend ation System. Developers Mr. Pradipat W orrarakthara 533040453-8 Mr. Thanaphon Boonphonsri 533040713-8 Mentors Dr. Jiradej Ponsawat Assist. Prof. Wanida Kanarkard Assist. Prof. Boonyarit Kukiattikool. 1. How to Choose books in the library?.
E N D
COE2013-5 The Development of Book Recommendation System Developers Mr. PradipatWorrarakthara533040453-8 Mr. ThanaphonBoonphonsri533040713-8 Mentors Dr.JiradejPonsawat Assist. Prof.WanidaKanarkard Assist. Prof.BoonyaritKukiattikool
1 How to Choose books in the library?
2 The purpose of the project
3 How can we solve this ? Recommendation System - Item-based filtering - User-based filtering
4 Item-based filtering Basic theory involved
5 User-based filtering Basic theory involved
6 Basic theory involved Similarity measure - Log-likelihood
7 Basic theory involved Similarity measure - Tanimotocoefficient similarity(Jaccardindex)
8 Design system for recommended ขDatabase Engine Recommender System Input data User List data for recommended The basic architecture of recommendation system
9 Engine for recommended
10 Procedures • Education Mahout Engine for Create Recommended system • Collected and compared to the Recommended System concept for the model that is right for most libraries. • Manage and store user information and borrowing history. • Test recommended system with real database
11 Functions • Euclidean Distance • Pearson Correlation • Tanimoto Coefficient • Loglikelihood
12 Building a recommendation engine 1. Creating the model and defining user similarity 2. Generating recommendations 3. Output from user recommendation
13 Code Example
14 Database Example
15 Database Example
16 Database Example
17 Result User-Based Filtering
18 Result Item-Based Filtering (Tanimoto)
19 Result Item-Based Filtering(Log-likelihood)
20 Problem
21 Analysis
22 Conclusion