字符串比较算法（或字符串相似度）

编码, 文本相关

自由的
无需下载/安装
无需注册

https://asecuritysite.com/forensics/simstring

在编码过程中，如果您想比比较两个字符串是否完全相等（"abc" == "abd" ?）更聪明......比如第一个字符串与第二个字符串的差别有多大（比如用户输入的内容有错别字），您可以在几种算法中进行选择，每种算法都有优缺点或针对特殊情况。

这个网站正是为此而设：提供两个字符串样本，它会应用多种算法并告诉你它们之间的差异。根据你要识别的字符串，经过几次测试，你就能知道最适合你情况的算法和阈值。它会一次性测试所有这些算法，并返回字符串相似度百分比：

莱文斯坦
针线人温奇
史密斯-沃特曼
史密斯-沃特曼·戈托
Smith-Waterman Gotoh 窗口仿射
年
温克勒
QGrams距离
街区距离
余弦相似度
欧氏距离
查普曼长度偏差
重叠系数

（有些可以直接集成到你喜欢的语言中，有些需要外部库或由你自行编写代码）

注意：该网站设有类似 Cloudflare 的机器人拦截墙，很烦人，但也没什么特别的（除了我们看到的难看截图）。

学科： 计算机科学、数据科学、生物信息学、语言学、软件工程。

Free alternative to Diffchecker, Beyond Compare, Python difflib.

Strings Comparison Algorithms compared with Diffchecker, Beyond Compare and Python difflib
Strings Comparison Algorithms	Diffchecker	Beyond Compare	Python difflib
What it measures: The page takes two sample strings and runs 13 algorithms at once, returning a similarity percentage for each. You can then pick the algorithm and the threshold that fit your data.	What it measures: The vendor describes core side-by-side comparison of two texts or files, plus Redline and folder or batch comparison on paid plans. The pricing page does not mention selectable similarity metrics.	What it measures: The vendor describes comparison of folders, text files, images and tables, with syntax highlighting and comparison rules for documents and source code.	What it measures: SequenceMatcher.ratio() returns a float between 0 and 1 computed as 2.0*M/T, where M is the number of matches and T the total number of elements in both sequences.
Algorithm basis: The list includes Levenshtein, Needleman-Wunsch, Smith-Waterman (with Gotoh variants), Jaro, Jaro-Winkler, q-grams, block distance, cosine similarity, Euclidean distance, Chapman length deviation and overlap coefficient. Edit-distance and alignment methods sit side by side.			Algorithm basis: The matching method is described as a little fancier than the Ratcliff-Obershelp gestalt algorithm. The documentation does not mention Levenshtein distance.
Use in code: Some of the listed algorithms ship with common languages, while others need external libraries or your own implementation.	Use in code: Separate API plans allow the comparison engine to be embedded in other applications.	Use in code: A scripting language is offered to automate syncs.	Use in code: It belongs to the Python standard library. get_close_matches(word, possibilities, n=3, cutoff=0.6) returns the best candidates. Worst-case time is quadratic, and strings of 200 or more items trigger an automatic junk heuristic that can be disabled with autojunk=False.
Licence, cost and platform: The page runs in the browser and sits behind a bot-blocking wall. A price or licence is not stated in the material available.	Licence, cost and platform: The Basic plan is free with no file size limit, but many Pro features are capped at about 5 to 10 uses monthly. Pro with desktop app is a paid per-seat plan, and the desktop app runs on Windows, Mac and Linux.	Licence, cost and platform: A per-user licence in Standard and Pro editions is a one-time purchase with no annual renewal fees. It runs on Windows, macOS and Linux and can be evaluated for free.	Licence, cost and platform: No separate cost, since it ships with Python. The documentation is under the Python Software Foundation License Version 2.

涵盖的主题： 字符串比较、字符串相似性、Levenshtein、Needleman-Wunsch、Smith-Waterman、Jaro、Jaro-Winkler、Q-Grams Distance、Block Distance、Cosine Similarity、Euclidean Distance、Chapman Length Deviation、Overlap Coefficient、Algorithm、Threshold、Similarity Percentage、Typo DetectionISO/IEC 30170、ISO/IEC 24765、ISO/IEC 25010、ISO/IEC 25012 和 ISO/IEC 25051。.