Table of contents
Preface
This project uses a matrix decomposition algorithm to conduct in-depth analysis of data that players have played. Its goal is to select the most suitable game for the player from many games to achieve a relatively accurate game recommendation system.
First, the project will collect and analyze game data that players have played, including game name, game duration, game ratings and other information. These data form a large user-game interaction matrix, in which each row represents a player, each column represents a game, and the values in the matrix represent the interaction between the player and the game.
Next, the project uses a matrix decomposition algorithm to approximately replace the sparse user-game matrix with two small matrices - the feature-game matrix and the user-feature matrix. This decomposition process maps players and games into a latent feature space, allowing potential relationships between players and games to be inferred.
Once the model is trained, the system can predict the games a player might like based on their gaming history. This prediction is based on how similar the player is to other players and how similar the game is to other games. As a result, the system can provide personalized game recommendations for each player, taking into account their gaming preferences and historical behavior.
Overall, the goal of this project is to provide a more accurate game recommendation system through matrix decomposition and latent factor models. This kind of personalized recommendation can improve players' gaming experience, and also help gaming platforms provide better game promotion and increase user stickiness.
overall design
This part includes the overall system structure diagram and system flow chart.
Overall system structure diagram
The overall structure of the system is shown in the figure.
System flow chart
The system flow is shown in the figure.
Operating environment
This part includes the Python environment, TensorFlow environment, and PyQt5 environment.
See the blog for details: https://blog.csdn.net/qq_31136513/article/details/133148686#_38
Module implementation
This project includes 4 modules: data preprocessing, model construction, model training and saving, and model testing. The function introduction and related codes of each module are given below.
1. Data preprocessing
The data set comes from Kaggle, and the link address is https://www.kaggle.com/tamber/steam-video-games . This data set includes the user’s ID, game name, whether to purchase or play, and game duration. Among them: a total of It contains 12,393 users and involves 5,155 games. steam-video-gamesPlace the dataset in a folder under the Jupyter working path .
See the blog for details: https://blog.csdn.net/qq_31136513/article/details/133148686#1__97
2. Model construction
After the data is loaded into the model, the model structure needs to be defined and the loss function optimized.
1) Define the model structure
Using the matrix decomposition algorithm, the user-game sparse matrix is approximately replaced by two small matrices - the feature-game matrix and the user-feature matrix.
See the blog for details: https://blog.csdn.net/qq_31136513/article/details/133151049#1_54
2) Optimize the loss function
The L2 norm is often used in the loss function of matrix factorization algorithms. Therefore, the L2 norm is also introduced into the loss function of this project to avoid overfitting. Optimize model parameters using Adagrad optimizer.
See the blog for details: https://blog.csdn.net/qq_31136513/article/details/133151049#2_91
3. Model training and saving
Since the data set used in this project lists the game's DLC (Downloadable Content, subsequent downloadable content) as another game, therefore, when calculating the accuracy, the DLC and the game itself are judged to be the same game and the same series. The games can also be judged to be the same one.
See the blog for details: https://blog.csdn.net/qq_31136513/article/details/133151049#3__105
1) Model training
See the blog for details: https://blog.csdn.net/qq_31136513/article/details/133151049#1_148
2) Model saving
In order to facilitate the use of the model, the training results need to be saved using Joblib.
See the blog for details: https://blog.csdn.net/qq_31136513/article/details/133151049#2_187
4. Model application
The first is to make the layout of the page, obtain and check the input data; the second is to match the obtained data with the previously saved model to achieve the application effect.
1) Make a page
The relevant operations are as follows:
(1) Use code to draw the basic layout of the page and create Recommandationclasses.
class Recommandation(QWidget):
#初始化
def __init__(self):
super().__init__()
self.initUI()
#初始化布局
def initUI(self):
#设置界面的初始位置和大小
self.setGeometry(600,200,450,550)
#窗口名
self.setWindowTitle('steam游戏推荐')
#设置组件，以下为标签
self.lb1 = QLabel('请输入游戏名：',self)
#这是所在位置
self.lb1.move(20,20)
self.lb2 = QLabel('请输入游戏名：',self)
self.lb2.move(20,80)
self.lb3 = QLabel('请输入游戏名：',self)
self.lb3.move(20,140)
self.lb4 = QLabel('请输入游戏名：',self)
self.lb4.move(20,200)
self.lb5 = QLabel('请输入游戏名：',self)
self.lb5.move(20,260)
#以下为下拉输入框的创建
self.combobox1 = QComboBox(self, minimumWidth=200)
self.combobox1.move(100,20)
self.combobox1.setEditable(True)
self.combobox2 = QComboBox(self, minimumWidth=200)
self.combobox2.move(100,80)
self.combobox2.setEditable(True)
self.combobox3 = QComboBox(self, minimumWidth=200)
self.combobox3.move(100,140)
self.combobox3.setEditable(True)
self.combobox4 = QComboBox(self, minimumWidth=200)
self.combobox4.move(100,200)
self.combobox4.setEditable(True)
self.combobox5 = QComboBox(self, minimumWidth=200)
self.combobox5.move(100,260)
self.combobox5.setEditable(True)
#以下为输入的按键设置
self.bt1 = QPushButton('请输入游戏时间',self)
self.bt1.move(330,20)
self.bt2 = QPushButton('请输入游戏时间',self)
self.bt2.move(330,80)
self.bt3 = QPushButton('请输入游戏时间',self)
self.bt3.move(330,140)
self.bt4 = QPushButton('请输入游戏时间',self)
self.bt4.move(330,200)
self.bt5 = QPushButton('请输入游戏时间',self)
self.bt5.move(330,260)
#推荐按钮
self.bt=QPushButton('推荐开始',self)
self.bt.move(20,400)
#初始化下拉输入框
self.init_combobox()
#连接按键与槽
self.bt1.clicked.connect(self.timeDialog)
self.bt2.clicked.connect(self.timeDialog)
self.bt3.clicked.connect(self.timeDialog)
self.bt4.clicked.connect(self.timeDialog)
self.bt5.clicked.connect(self.timeDialog)
#连接推荐
self.bt.clicked.connect(self.recommand)
connect()It is Qt's unique signal and slot mechanism. The slot receives the signal and processes it. Clicked is used as a signal here, and clicking a button will send out a signal.
(2) Initialize the drop-down input box, enter the gamelist into the menu of the drop-down box, and add the auto-complete function.
#初始化下拉输入框
def init_combobox(self):
#增加选项元素
for i in range(len(gamelist)):
self.combobox1.addItem(gamelist[i])
self.combobox2.addItem(gamelist[i])
self.combobox3.addItem(gamelist[i])
self.combobox4.addItem(gamelist[i])
self.combobox5.addItem(gamelist[i])
self.combobox1.setCurrentIndex(-1)
self.combobox2.setCurrentIndex(-1)
self.combobox3.setCurrentIndex(-1)
self.combobox4.setCurrentIndex(-1)
self.combobox5.setCurrentIndex(-1)
#增加自动补全
self.completer = QCompleter(gamelist)
#补全方式
self.completer.setFilterMode(Qt.MatchStartsWith)
self.completer.setCompletionMode(QCompleter.PopupCompletion)
self.combobox1.setCompleter(self.completer)
self.combobox2.setCompleter(self.completer)
self.combobox3.setCompleter(self.completer)
self.combobox4.setCompleter(self.completer)
self.combobox5.setCompleter(self.completer)
(3) Set up slots and store data at the same time
The relevant operations are as follows:
def timeDialog(self):
#获取信号
sender = self.sender()
if sender == self.bt1:
#获取下拉输入框1输入的游戏名
gamename = self.combobox1.currentText()
#通过字典game2idx查询获得的游戏名所对应的序列号
gameid = game2idx.get(gamename)
#没有序列号的情况，可以理解为未输入正确的游戏名，或者输入为空
if gameid == None:
#这种情况下生成一个MessageBox报错
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
#输入正确的情况，将游戏名字、ID，分别记录到一个字典里，方便保存与更改
gamedict[1] = gamename
idxdict[1] = gameid
#弹出一个文本输入框，要求输入对应游戏时长
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
#如果输入正确，将时长记录到一个字典中，方便保存与更改
if ok:
timedict[1] = text
elif sender == self.bt2:
gamename = self.combobox2.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[2] = gamename
idxdict[2] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[2] = text
elif sender == self.bt3:
gamename = self.combobox3.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[3] = gamename
idxdict[3] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[3] = text
elif sender == self.bt4:
gamename = self.combobox4.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[4] = gamename
idxdict[4] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[4] = text
elif sender == self.bt5:
gamename = self.combobox5.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[5] = gamename
idxdict[5] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[5] = text
(4) Verify whether the data has been entered and prepare to call the model
def recommand(self):
#验证是否存在没有写入的数据
c = 0
for i in range(1,6):
if gamedict[i] == "NULL":
c+=1
if idxdict[i] == "NULL":
c+=1
if timedict[i] == "NULL":
c+=1
#全部写完的情况
if c == 0:
#将字典转化为列表
usertime = list(timedict.values())
useridx = list(idxdict.values())
#调用模型
allrecidx = UserSimilarity(useridx,usertime)
#降序排列数据
rr = np.argsort(-allrecidx)
#获取排行前五的游戏ID
top_k = rr[:5]
#将ID对应的游戏名字输入数组
for i in top_k:
recgame.append(idx2game[i])
#将数组转化为字符串并输出
reclist = ','.join(recgame)
reply = QMessageBox.information(self,'推荐的游戏','给您推荐的游戏是'+reclist, QMessageBox.Close)
#存在没有写完的数据，要求重新写入
else:
reply = QMessageBox.information(self,'Error','请输入全部数据!', QMessageBox.Close)
2) Model import and call
The relevant operations are as follows:
(1) Load the Save_data model in the current folder
game2idx = joblib.load('./Save_data/game2idx.pkl')
idx2game = joblib.load('./Save_data/idx2game.pkl')
rec = joblib.load('./Save_data/rec.pkl')
hours = joblib.load('./Save_data/hours.pkl')
buy = joblib.load('./Save_data/buy.pkl')
users = joblib.load('./Save_data/buyers.pkl')
(2) Create a user similarity function to describe the data with the highest similarity between the data collected in Qt and the trained users as the output
def UserSimilarity(games, game_hours):
similarity = np.zeros(len(users)) # 用户相似度矩阵
for i in range(len(users)):
#计算用户输入的游戏与数据集中每个用户购买游戏的重合度
coincidence = 0 #重合度，每重合一个游戏加1
positions = [] #重合游戏在games中的位置
#获取数据集中的第i个玩家与用户输入的重合情况
for ii in range(len(games)):
if games[ii] in np.where(buy[users[i], :] == 1)[0]:
coincidence += 1
positions.append(ii)
#如果没有重合，则相似度为0，跳过
if coincidence == 0:
continue
simi = []
#将重合的游戏，根据时长和相同游戏的时长差取绝对值，根据e^-x计算出相似度
for position in positions:
game = games[position]
hour = abs(game_hours[position] - hours[users[i], game])
simi.append(math.exp(-hour))
#对所有相似度取均值，得到用户与数据集中第i个玩家的相似度similarity[i]
similarity[i] = sum(simi) / coincidence
#相似度与玩家—游戏矩阵每一行相乘
for i in range(len(users)):
user = users[i]
rec[user] = rec[user] * similarity[i]
new_rec = np.zeros(len(rec[0])) # 1*n_games矩阵
#将玩家—游戏矩阵按列相加，得到用户对每个游戏的喜好程度，即new_rec矩阵
for i in range(len(new_rec)):
for user in users:
new_rec[i] += rec[user][int(i)]
return new_rec
3) Model application code
The relevant code is as follows:
import joblib
import numpy as np
import pandas as pd
import math
import sys
from PyQt5.QtWidgets import *
from PyQt5.QtGui import *
from PyQt5.QtCore import *
#读取数据
game2idx = joblib.load('./Save_data/game2idx.pkl')
idx2game = joblib.load('./Save_data/idx2game.pkl')
rec = joblib.load('./Save_data/rec.pkl')
hours = joblib.load('./Save_data/hours.pkl')
buy = joblib.load('./Save_data/buy.pkl')
users = joblib.load('./Save_data/buyers.pkl')
#游戏名称列表
gamelist = list(game2idx)
#游戏数
n_game = len(gamelist)
#传入字典
gamedict = {
1:"NULL",2:"NULL",3:"NULL",4:"NULL",5:"NULL"}
timedict = {
1:"NULL",2:"NULL",3:"NULL",4:"NULL",5:"NULL"}
idxdict = {
1:"NULL",2:"NULL",3:"NULL",4:"NULL",5:"NULL"}
#下面两个是要传递的
usertime=[]
useridx=[]
#下面是返回的推荐游戏
recgame=[]
#相似度推荐
def UserSimilarity(games, game_hours):
similarity = np.zeros(len(users)) #用户相似度矩阵
for i in range(len(users)):
#计算用户输入的游戏与数据集中每个用户购买游戏的重合度
coincidence = 0 #重合度
positions = [] #重合游戏在games中的位置
for ii in range(len(games)):
if games[ii] in np.where(buy[users[i], :] == 1)[0]:
coincidence += 1
positions.append(ii)
if coincidence == 0:
continue
simi = []
for position in positions:
game = games[position]
hour = abs(game_hours[position] - hours[users[i], game])
simi.append(math.exp(-hour))
similarity[i] = sum(simi) / coincidence
#相似度与玩家—游戏矩阵每一行相乘
for i in range(len(users)):
user = users[i]
rec[user] = rec[user] * similarity[i]
new_rec = np.zeros(len(rec[0])) #1*n_games矩阵
for i in range(len(new_rec)):
for user in users:
new_rec[i] += rec[user][int(i)]
return new_rec
class Recommandation(QWidget):
#初始化
def __init__(self):
super().__init__()
self.initUI()
#初始化布局
def initUI(self):
#设置界面的初始位置和大小
self.setGeometry(600,200,450,550)
#窗口名
self.setWindowTitle('steam游戏推荐')
#设置组件，以下为标签
self.lb1 = QLabel('请输入游戏名：',self)
#这是所在位置
self.lb1.move(20,20)
self.lb2 = QLabel('请输入游戏名：',self)
self.lb2.move(20,80)
self.lb3 = QLabel('请输入游戏名：',self)
self.lb3.move(20,140)
self.lb4 = QLabel('请输入游戏名：',self)
self.lb4.move(20,200)
self.lb5 = QLabel('请输入游戏名：',self)
self.lb5.move(20,260)
#以下为下拉输入框的创建
self.combobox1 = QComboBox(self, minimumWidth=200)
self.combobox1.move(100,20)
self.combobox1.setEditable(True)
self.combobox2 = QComboBox(self, minimumWidth=200)
self.combobox2.move(100,80)
self.combobox2.setEditable(True)
self.combobox3 = QComboBox(self, minimumWidth=200)
self.combobox3.move(100,140)
self.combobox3.setEditable(True)
self.combobox4 = QComboBox(self, minimumWidth=200)
self.combobox4.move(100,200)
self.combobox4.setEditable(True)
self.combobox5 = QComboBox(self, minimumWidth=200)
self.combobox5.move(100,260)
self.combobox5.setEditable(True)
#以下为输入的按键设置
self.bt1 = QPushButton('请输入游戏时间',self)
self.bt1.move(330,20)
self.bt2 = QPushButton('请输入游戏时间',self)
self.bt2.move(330,80)
self.bt3 = QPushButton('请输入游戏时间',self)
self.bt3.move(330,140)
self.bt4 = QPushButton('请输入游戏时间',self)
self.bt4.move(330,200)
self.bt5 = QPushButton('请输入游戏时间',self)
self.bt5.move(330,260)
#推荐按钮
self.bt=QPushButton('推荐开始',self)
self.bt.move(20,400)
#初始化下拉输入框
self.init_combobox()
#连接按键与槽
self.bt1.clicked.connect(self.timeDialog)
self.bt2.clicked.connect(self.timeDialog)
self.bt3.clicked.connect(self.timeDialog)
self.bt4.clicked.connect(self.timeDialog)
self.bt5.clicked.connect(self.timeDialog)
#连接推荐
self.bt.clicked.connect(self.recommand)
#初始化下拉输入框
def init_combobox(self):
#增加选项元素
for i in range(len(gamelist)):
self.combobox1.addItem(gamelist[i])
self.combobox2.addItem(gamelist[i])
self.combobox3.addItem(gamelist[i])
self.combobox4.addItem(gamelist[i])
self.combobox5.addItem(gamelist[i])
self.combobox1.setCurrentIndex(-1)
self.combobox2.setCurrentIndex(-1)
self.combobox3.setCurrentIndex(-1)
self.combobox4.setCurrentIndex(-1)
self.combobox5.setCurrentIndex(-1)
#增加自动补全
self.completer = QCompleter(gamelist)
#补全方式
self.completer.setFilterMode(Qt.MatchStartsWith)
self.completer.setCompletionMode(QCompleter.PopupCompletion)
self.combobox1.setCompleter(self.completer)
self.combobox2.setCompleter(self.completer)
self.combobox3.setCompleter(self.completer)
self.combobox4.setCompleter(self.completer)
self.combobox5.setCompleter(self.completer)
def timeDialog(self):
#获取信号
sender = self.sender()
if sender == self.bt1:
#获取下拉输入框1输入的游戏名
gamename = self.combobox1.currentText()
#通过字典game2idx查询获得的游戏名所对应的序列号
gameid = game2idx.get(gamename)
#没有序列号的情况，可以理解为未输入正确的游戏名，或者输入为空
if gameid == None:
#这种情况下生成一个MessageBox报错
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
#输入正确的情况，将游戏名字、ID，分别记录到一个字典里，方便保存与更改
gamedict[1] = gamename
idxdict[1] = gameid
#弹出一个文本输入框，要求输入对应的游戏时长
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
#如果输入正确，将时长记录到一个字典中，方便保存与更改
if ok:
timedict[1] = text
elif sender == self.bt2:
gamename = self.combobox2.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[2] = gamename
idxdict[2] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[2] = text
elif sender == self.bt3:
gamename = self.combobox3.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[3] = gamename
idxdict[3] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[3] = text
elif sender == self.bt4:
gamename = self.combobox4.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[4] = gamename
idxdict[4] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[4] = text
elif sender == self.bt5:
gamename = self.combobox5.currentText()
gameid = game2idx.get(gamename)
if gameid == None:
reply = QMessageBox.information(self,'Error','请输入正确的游戏名!', QMessageBox.Close)
else:
gamedict[5] = gamename
idxdict[5] = gameid
text, ok = QInputDialog.getDouble(self, '游戏时间', '请输入游戏时间：', min = 0.1)
if ok:
timedict[5] = text
def recommand(self):
#验证是否存在没有写入的数据
c = 0
for i in range(1,6):
if gamedict[i] == "NULL":
c+=1
if idxdict[i] == "NULL":
c+=1
if timedict[i] == "NULL":
c+=1
#全部写完的情况
if c == 0:
#将字典转化为列表
usertime = list(timedict.values())
useridx = list(idxdict.values())
#调用模型
allrecidx = UserSimilarity(useridx,usertime)
#降序排列数据
rr = np.argsort(-allrecidx)
#获取排行前五的游戏ID
top_k = rr[:5]
#将ID对应的游戏名字输入数组
for i in top_k:
recgame.append(idx2game[i])
#将数组转化为字符串并输出
reclist = ','.join(recgame)
reply = QMessageBox.information(self,'推荐的游戏','给您推荐的游戏是'+reclist, QMessageBox.Close)
#存在没有写完的数据，要求重新写入
else:
reply = QMessageBox.information(self,'Error','请输入全部数据!', QMessageBox.Close)
#主函数
if __name__ == "__main__":
app = QApplication(sys.argv)
w = Recommandation()
w.show()
sys.exit(app.exec_())
Related other blogs
Project source code download
For details, please see my blog resource download page
Download other information
If you want to continue to understand the learning routes and knowledge systems related to artificial intelligence, you are welcome to read my other blog " Heavyweight | Complete Artificial Intelligence AI Learning - Basic Knowledge Learning Route, all information can be downloaded directly from the network disk without following any routines.》
This blog refers to Github’s well-known open source platform, AI technology platform and experts in related fields: Datawhale, ApacheCN, AI Youdao and Dr. Huang Haiguang, etc., which has nearly 100G of related information. I hope it can help all my friends.