数据分析-paper

• 主要做一些论文相关的数据统计ing

---

TIFS2024

• 相对来讲TIFS收录的方向会比较均匀
• 因此虽然只选取了2024年收录的跟Android安全强相关的11篇论文，但从上面的词云中也能反映出一些问题

IEEE S&P

• 这个统计了从2012-2024的96篇Android强相关的论文，确实够priacvy
• 可以看出来近10年围绕权限所展开的移动端安全对抗来是非常多的，什么时候授予权限？什么授予多少权限？什么时候收回权限？

附录

• 词云生成脚本

import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
import matplotlib.pyplot as plt
from wordcloud import WordCloud

# 1. 读取 IEEE 导出的 Excel 文件
df = pd.read_excel("S&P.xlsx")

# 2. 选择论文标题和摘要列
texts = df["论文名称"] + " " + df["Abstract"]

# 3. 自定义停用词
custom_stop_words = set([
    "app","apps","application","attack","mobile","applications","devices","android","research","paper","google","software","analysis",
    "study","security","attacks","used","use","using","new","accuracy"
])

# 4. 计算 TF-IDF 关键词权重
vectorizer = TfidfVectorizer(stop_words="english", max_features=50)
tfidf_matrix = vectorizer.fit_transform(texts)

# 筛选掉自定义的停用词
filtered_keywords = [word for word in vectorizer.get_feature_names_out() if word not in custom_stop_words]

# 生成词云
wordcloud = WordCloud(width=800, height=400, background_color="white").generate(" ".join(filtered_keywords))
plt.figure(figsize=(10, 5))
plt.imshow(wordcloud, interpolation="bilinear")
plt.axis("off")
plt.show()