arxiv论文趋势统计工具

展示

（和 GPT）写了个小组件来查关键词在 arxiv 的论文发布趋势，用 ploty.js 写的前端，bs4 写的后端，感觉还行，等个时间整理了传在 github 上。

粗糙源码

前端

React 版本

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
import React from "react";
import { Card, CardHeader, CardTitle, CardContent } from "@/components/ui/card";
import {
  LineChart,
  Line,
  XAxis,
  YAxis,
  CartesianGrid,
  Tooltip,
  Legend,
  ResponsiveContainer,
  Bar,
  BarChart,
  ComposedChart,
} from "recharts";

const TrendVisualization = () => {
  const data = [
    { month: "2022-09", papers: 1 },
    { month: "2023-02", papers: 1 },
    { month: "2023-03", papers: 1 },
    { month: "2023-07", papers: 1 },
    { month: "2023-08", papers: 5 },
    { month: "2023-09", papers: 2 },
    { month: "2023-10", papers: 2 },
    { month: "2023-11", papers: 2 },
    { month: "2023-12", papers: 1 },
    { month: "2024-02", papers: 2 },
    { month: "2024-03", papers: 1 },
    { month: "2024-04", papers: 2 },
    { month: "2024-05", papers: 3 },
    { month: "2024-06", papers: 3 },
    { month: "2024-07", papers: 2 },
    { month: "2024-08", papers: 5 },
    { month: "2024-09", papers: 2 },
    { month: "2024-10", papers: 8 },
  ];

  // 计算 3 个月移动平均
  const calculateMovingAverage = (data, window = 3) => {
    return data.map((item, index) => {
      const start = Math.max(0, index - window + 1);
      const values = data.slice(start, index + 1).map((d) => d.papers);
      const avg = values.reduce((a, b) => a + b, 0) / values.length;
      return {
        ...item,
        movingAverage: parseFloat(avg.toFixed(2)),
      };
    });
  };

  const dataWithMA = calculateMovingAverage(data);

  return (
    <Card className="w-full max-w-4xl">
      <CardHeader>
        <CardTitle>Graph Prompting Research Trend</CardTitle>
      </CardHeader>
      <CardContent>
        <div className="h-96">
          <ResponsiveContainer width="100%" height="100%">
            <ComposedChart data={dataWithMA}>
              <CartesianGrid strokeDasharray="3 3" />
              <XAxis
                dataKey="month"
                angle={-45}
                textAnchor="end"
                interval={0}
                height={60}
              />
              <YAxis
                label={{
                  value: "Number of Papers",
                  angle: -90,
                  position: "insideLeft",
                  offset: -5,
                }}
              />
              <Tooltip />
              <Legend />
              <Bar
                dataKey="papers"
                fill="#8884d8"
                name="Monthly Papers"
                barSize={20}
              />
              <Line
                type="monotone"
                dataKey="movingAverage"
                stroke="#ff7300"
                name="3-Month Moving Average"
                strokeWidth={2}
                dot={false}
              />
            </ComposedChart>
          </ResponsiveContainer>
        </div>
      </CardContent>
    </Card>
  );
};
export default TrendVisualization;

html 版本

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="UTF-8" />
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <title>Graph Prompting Research Trend</title>
    <script
      src="https://cdn.plot.ly/plotly-2.35.2.min.js"
      charset="utf-8"
    ></script>
    <style>
      body {
        font-family: Arial, sans-serif;
        margin: 0;
        padding: 0;
      }
      .card {
        width: 100%;
        max-width: 800px;
        margin: auto;
        border: 1px solid #ccc;
        border-radius: 8px;
        overflow: hidden;
        box-shadow: 0 2px 8px rgba(0, 0, 0, 0.1);
      }
      .card-header {
        background: #f5f5f5;
        padding: 16px;
        font-size: 1.5em;
        font-weight: bold;
      }
      .card-content {
        padding: 0;
      }
      .chart-container {
        width: 100%;
        height: 400px;
      }
    </style>
  </head>
  <body>
    <div class="card">
      <div class="card-header">Graph Prompting Research Trend</div>
      <div class="card-content">
        <div id="chart" class="chart-container"></div>
      </div>
    </div>

    <script>
      const data = [
        { month: "2022-09", papers: 1 },
        { month: "2023-02", papers: 1 },
        { month: "2023-03", papers: 1 },
        { month: "2023-07", papers: 1 },
        { month: "2023-08", papers: 5 },
        { month: "2023-09", papers: 2 },
        { month: "2023-10", papers: 2 },
        { month: "2023-11", papers: 2 },
        { month: "2023-12", papers: 1 },
        { month: "2024-02", papers: 2 },
        { month: "2024-03", papers: 1 },
        { month: "2024-04", papers: 2 },
        { month: "2024-05", papers: 3 },
        { month: "2024-06", papers: 3 },
        { month: "2024-07", papers: 2 },
        { month: "2024-08", papers: 5 },
        { month: "2024-09", papers: 2 },
        { month: "2024-10", papers: 8 },
      ];

      // Calculate cumulative papers
      const cumulativeData = data.map((item, index) => ({
        month: item.month,
        cumulativePapers: data
          .slice(0, index + 1)
          .reduce((acc, cur) => acc + cur.papers, 0),
        papers: item.papers,
      }));

      const trace1 = {
        x: cumulativeData.map((d) => d.month),
        y: cumulativeData.map((d) => d.cumulativePapers),
        type: "bar",
        name: "Cumulative Papers",
        marker: { color: "#8884d8" },
      };

      const trace2 = {
        x: data.map((d) => d.month),
        y: data.map((d) => d.papers),
        mode: "lines+markers",
        name: "Monthly Papers",
        line: { color: "#ff7300", width: 2 },
      };

      const layout = {
        title: "Graph Prompting Research Trend",
        xaxis: {
          title: "Month",
          tickangle: -45,
        },
        yaxis: {
          title: "Number of Papers",
          titlefont: { size: 16 },
          autorange: true,
        },
        barmode: "overlay",
      };

      Plotly.newPlot("chart", [trace1, trace2], layout);
    </script>
  </body>
</html>

后端

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
import urllib.request as libreq
import xml.etree.ElementTree as ET
import pandas as pd
import time
from datetime import datetime
import csv
from collections import Counter
import re

def create_search_query(search_terms, mode='broad'):
    """
    创建不同模式的搜索查询
    """
    if mode == 'strict':
        return f'%22{search_terms}%22'
    elif mode == 'medium':
        return f'({search_terms})'
    else:
        return f'({search_terms})'

def fetch_papers(search_terms, mode='broad', start=0, max_results=100):
    """
    获取arXiv论文数据
    """
    query = create_search_query(search_terms, mode)
    url = f'http://export.arxiv.org/api/query?search_query=all:{query}&start={start}&max_results={max_results}&sortBy=submittedDate&sortOrder=descending'

    try:
        with libreq.urlopen(url) as response:
            r = response.read()
        return ET.fromstring(r)
    except Exception as e:
        print(f"Error fetching data: {e}")
        return None

def check_relevance(text, search_terms, mode='broad'):
    """
    检查文本是否相关，支持不同的匹配模式
    """
    text = text.lower()

    # 定义不同严格程度的关键词组合
    strict_patterns = [
        rf'{search_terms}'
    ]

    medium_patterns = strict_patterns + [
        rf'{search_terms.replace(" ", ".*")}'
    ]

    broad_patterns = medium_patterns + [
        rf'{search_terms.replace(" ", ".*")}'
    ]

    # 选择对应模式的模式列表
    patterns = (strict_patterns if mode == 'strict' else
               medium_patterns if mode == 'medium' else
               broad_patterns)

    # 检查是否匹配任何模式
    return any(re.search(pattern, text) for pattern in patterns)

def parse_entry(entry, search_terms, search_mode='broad'):
    """
    解析单个论文条目，增加相关性评分
    """
    ns = {'atom': 'http://www.w3.org/2005/Atom'}

    title = entry.find('atom:title', ns).text.strip()
    summary = entry.find('atom:summary', ns).text.strip()
    published = entry.find('atom:published', ns).text

    date = datetime.strptime(published, '%Y-%m-%dT%H:%M:%SZ')
    year_month = f"{date.year}-{date.month:02d}"
    full_date = date.strftime('%Y-%m-%d')

    authors = [author.find('atom:name', ns).text for author in entry.findall('atom:author', ns)]
    categories = [cat.get('term') for cat in entry.findall('atom:category', ns)]

    text = f"{title} {summary}"
    is_relevant = check_relevance(text, search_terms, search_mode)

    relevance_score = 1 if is_relevant else 0
    if 'cs.AI' in categories or 'cs.LG' in categories:
        relevance_score += 0.5

    return {
        'title': title,
        'summary': summary,
        'year_month': year_month,
        'full_date': full_date,
        'authors': '; '.join(authors),
        'categories': '; '.join(categories),
        'is_relevant': is_relevant,
        'relevance_score': relevance_score
    }

def collect_all_papers(search_terms, search_mode='broad'):
    """
    收集所有相关论文数据
    """
    papers = []
    start = 0
    batch_size = 100
    total_checked = 0

    while True:
        print(f"Fetching papers from index {start}")
        root = fetch_papers(search_terms, search_mode, start=start, max_results=batch_size)

        if root is None:
            break

        entries = root.findall('{http://www.w3.org/2005/Atom}entry')

        if not entries:
            break

        for entry in entries:
            paper_data = parse_entry(entry, search_terms, search_mode)
            total_checked += 1

            if paper_data['is_relevant']:
                papers.append(paper_data)

        print(f"Checked {total_checked} papers, found {len(papers)} relevant papers...")

        time.sleep(3)
        start += batch_size

        if len(entries) < batch_size:
            break

    return papers

def save_results(papers, search_terms, search_mode, base_filename='results'):
	"""
	保存结果并生成详细报告
	"""
	if not papers:
		print("No papers found!")
		return None, None

	df = pd.DataFrame(papers)

	detailed_filename = f'{base_filename}_{search_terms[:5]}_{search_mode}_detailed.csv'
	df.to_csv(detailed_filename, index=False, encoding='utf-8')

	stats = {
		'total_papers': len(df),
		'by_category': df['categories'].str.get_dummies(sep='; ').sum().to_dict(),
		'by_month': df.groupby('year_month').size().to_dict(),
		'avg_relevance': df['relevance_score'].mean(),
	}

	report_filename = f'{base_filename}_{search_terms[:5]}_{search_mode}_report.txt'
	with open(report_filename, 'w', encoding='utf-8') as f:
		f.write(f"Research Report for {search_terms} ({search_mode} mode)\n")
		f.write("="* 50 + "\n\n")
		f.write(f"Total Papers Found: {stats['total_papers']}\n\n")

		f.write("Distribution by Category:\n")
		for cat, count in stats['by_category'].items():
			f.write(f"{cat}: {count}\n")

		f.write("\nMonthly Distribution:\n")
		for month, count in sorted(stats['by_month'].items()):
			f.write(f"{month}: {count}\n")

		f.write(f"\nAverage Relevance Score: {stats['avg_relevance']:.2f}\n")

	return df, stats

if __name__ == "__main__":
    search_terms = 'resnet'
    modes = ['strict']
    # modes = ['strict', 'medium', 'broad']
    results = {}

    for mode in modes:
        print(f"\nRunning {mode} search mode...")
        papers = collect_all_papers(search_terms, mode)
        df, stats = save_results(papers, search_terms, mode)
        results[mode] = {
            'papers': papers,
            'stats': stats
        }

        print(f"\n{mode.capitalize()} mode results:")
        print(f"Total papers found: {len(papers)}")
        if stats:
            print(f"Average relevance score: {stats['avg_relevance']:.2f}")

结语

内嵌这个 html 在网站里面还挺缺德的，hexo 不能直接插入 div 和 script 导致我自己是用 html 然后用 if-frame 插入的。

今天先这样，快乐结束！