[大学院ーpython-base learning3: python and recommendation system ]

[python-base learning3: python and recommendation system ]

1:introduction
[2:computer setting development environment & Jupyter Notebook](#2:computer setting development environment & Jupyter Notebook)
- [(1):computer setting development environment](#(1):computer setting development environment)
[3:basic concept of Recommendation System](#3:basic concept of Recommendation System)
[4:Types of Recommendation Methods](#4:Types of Recommendation Methods)
- [(1):checkbox/ Profile-Based Method](#(1):checkbox/ Profile-Based Method)
- [(2):Rule-based method](#(2):Rule-based method)
- [(3):Recommendation System Method](#(3):Recommendation System Method)
5:Method
- [1--Collaborative Filtering(Main Topic)](#1--Collaborative Filtering(Main Topic))
- - (1):Idea
  - (2):Process:
  - [(3):Output :](#(3):Output :)
  - (4):Example:
- [2--Content-Based Filtering](#2--Content-Based Filtering)
- [3--Hybrid Method](#3--Hybrid Method)
[6:Internal Mechanism](#6:Internal Mechanism)
7:Example
8:Conclusion

1:introduction

In this lecture , we learned about about recommendation systems, especially collaborative filtering

recommendation system are widely used in:

E-commerce (Amazon)
Video platfroms(Netfli)
Online services

Their purpose is to suggest items that users may like based on data

2:computer setting development environment & Jupyter Notebook

(1):computer setting development environment

It is ver important to ensure that your computer settings and development environment are properly configured .

If the setup is in correct or complicated, Python may not work properly , and errors may occur

Therefore we need to make sure that our computer settings and development environment are correctly prepared

System Requirements

win : 10 /11
python 3.10
Jupyter Notebook

3:basic concept of Recommendation System

A recommendation system works as follows:

(1):Input

User profile
Browsing history
purchase history

(2):Process:

Analyze user behavior
find patterns or similarities

(3):Output:

Recommended items

4:Types of Recommendation Methods

Tere are three main types:

(1):checkbox/ Profile-Based Method

Users manually select their preferences.

Features:

Simple
Direct input from users

(2):Rule-based method

Recommendations are generated using predefined rules or machine learning

(3):Recommendation System Method

this is includes:

Collaborative Filtering
Content-Based Filtering
Hybrid Method

5:Method

1--Collaborative Filtering(Main Topic)

collaborative Filtering focuses on user

(1):Idea

users with similar preferences will like similar items

(2):Process:

Find similar users
Identify items they like

(3):Output :

Recommend unseen items

(4):Example:

If User A and User B have similar tastes,

items liked by B wil be reommended to A .

Advantages:

No need to analyze item content
Can discover new items

Disadvantages:

Limited diversity
Depends on content analysis

2--Content-Based Filtering

This method focuses on item similarity.

(1):Idea:

Recommend items similar to what the user liked before

(2):Advantages:

No need for other users
works for new items

(3):Disadvantages

Limited diversity
Depends on content analysis

3--Hybrid Method

Combines:

collaborative filtering
Content-based filtering

Provides better accuracy

6:Internal Mechanism

The system processes data as follows:

High-dimensional data ->Low-dimensional representation
Similarity calculation
Generate recommendations

7:Example

liked by others + not seen by you = recommendation

bash 复制代码

# Items liked by users (simple data)
user_A = ["apple", "milk"]
user_B = ["apple", "milk", "bread"]

# Find recommendations (items B has but A does not)
recommend = []

for item in user_B:
    if item not in user_A:
        recommend.append(item)

print("Recommendation for A:", recommend)

8:Conclusion

In this part, we learned :

The concept of recommendation systems
Three main methods
Collaborative filtering as the core technique

These methods are essential in modern AI systems and data-driven applications.