# speciallurain/CNKI\_Patent\_SVM

<https://repo.masters-masters.com/en/repos/speciallurain/CNKI_Patent_SVM>

[Home](https://repo.masters-masters.com/en) › [speciallurain](https://repo.masters-masters.com/en/owners/speciallurain) › [CNKI\_Patent\_SVM](https://repo.masters-masters.com/en/repos/speciallurain/CNKI_Patent_SVM)

speciallurain/CNKI\_Patent\_SVM has 107 stars on GitHub. Its RepoMasters card: Prototype tier. 文本分类是指在给定分类体系下 , 根据文本的内容自动确定文本类别的过程。首先我们根据scrapy爬虫根据中国知网URL的规律，爬取7…

| Field | Value |
| --- | --- |
| Owner | [speciallurain](https://repo.masters-masters.com/en/owners/speciallurain) |
| Stars | 107 |
| Forks | 32 |
| Language | [Python](https://repo.masters-masters.com/en/languages/python) |
| License | None |
| Created | 2018 |
| Tier | [Prototype (100 to 249 ★)](https://repo.masters-masters.com/en/ranks/rush) |
| View on GitHub | [speciallurain/CNKI\_Patent\_SVM](https://github.com/speciallurain/CNKI_Patent_SVM) |

## Repository

文本分类是指在给定分类体系下 , 根据文本的内容自动确定文本类别的过程。首先我们根据scrapy爬虫根据中国知网URL的规律，爬取70多万条2014年公开的发明专利，然后通过数据清洗筛选出了60多万条含标签数据。通过TF-IDF对60多万条本文进行词频提取，依照词频排序提取前3000个词语形成语义词典，然后根据观察设置停用词。然后再用TF-IDF的方式对每个摘要进行词频选取，通过布尔模型，对比语义词典生成文本向量。然后对标签进行数字化转换。取90%的文本为训练集，10%的文本为测试集。用有监督学习的SVM算法对文本进行分类，（人类生活必需品、作业运输、化学冶金、纺织造纸、固定建筑物、机械工程…

The tier for repos with 100 to 249 ★. Published odds per card pulled: 70%. Rarity: Common.

## Nearby in Python

- [narratarr/narratarr · ★ 107 · Python · Prototype](https://repo.masters-masters.com/en/repos/narratarr/narratarr)
- [lingquan11/ai-chip-explorer · ★ 107 · Python · Prototype](https://repo.masters-masters.com/en/repos/lingquan11/ai-chip-explorer)
- [4yk5/Ghostman · ★ 107 · Python · Prototype](https://repo.masters-masters.com/en/repos/4yk5/Ghostman)
- [chrisdrackett/django-typogrify · ★ 107 · Python · Prototype](https://repo.masters-masters.com/en/repos/chrisdrackett/django-typogrify)
- [laanwj/decuda · ★ 107 · Python · Prototype](https://repo.masters-masters.com/en/repos/laanwj/decuda)
- [akaihola/pgtricks · ★ 107 · Python · Prototype](https://repo.masters-masters.com/en/repos/akaihola/pgtricks)

---
Data: GitHub · Illustration drawn by RepoMasters from the repo’s ID and its language color.