explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-02-03 21:24:11 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Upd sales copy

Browse Source
This commit is contained in:
Matthew Honnibal 2014-11-04 17:01:54 +11:00
parent 954c970415
commit 2420d944cb
1 changed files with 18 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
docs/source/index.rst
View File

@ -6,20 +6,19 @@
spaCy NLP Tokenizer and Lexicon spaCy NLP Tokenizer and Lexicon
================================ ================================
spaCy is a library for industrial strength NLP in Python and Cython. Its core spaCy is a library for industrial strength NLP in Python. Its core
values are efficiency, accuracy and minimalism. values are:
* Efficiency: spaCy is TODOx faster than the Stanford tools, and TODOx faster * **Efficiency**: You won't find faster NLP tools. For shallow analysis, it's 10x
than NLTK. You won't find faster NLP tools. Using spaCy will save you faster than Stanford Core NLP, and over 200x faster than NLTK. Its parser is
thousands in server costs, and will force you to make fewer compromises. over 100x faster than Stanford's.
* Accuracy: All spaCy tools are within 0.5% of the current published * **Accuracy**: All spaCy tools are within 0.5% of the current published
state-of-the-art, on both news and web text. NLP moves fast, so always check state-of-the-art, on both news and web text. NLP moves fast, so always check
the numbers --- and don't settle for tools that aren't backed by the numbers --- and don't settle for tools that aren't backed by
rigorous recent evaluation. An algorithm that was "close enough to state-of-the-art" rigorous recent evaluation.
5 years ago is probably crap by today's standards.
* Minimalism: This isn't a library that covers 43 known algorithms to do X. You * **Minimalism**: This isn't a library that covers 43 known algorithms to do X. You
get 1 --- the best one --- with a simple, low-level interface. This keeps the get 1 --- the best one --- with a simple, low-level interface. This keeps the
code-base small and concrete. Our Python APIs use lists and code-base small and concrete. Our Python APIs use lists and
dictionaries, and our C/Cython APIs use arrays and simple structs. dictionaries, and our C/Cython APIs use arrays and simple structs.
@ -27,15 +26,16 @@ values are efficiency, accuracy and minimalism.
Comparison Comparison
---------- ----------
+-------------+-------------+---+-----------+--------------+
| POS taggers | Speed (w/s) | % Acc. (news) | % Acc. (web) | +----------------+-------------+--------+---------------+--------------+
+-------------+-------------+---------------+--------------+ | Tokenize & Tag | Speed (w/s) | Memory | % Acc. (news) | % Acc. (web) |
| spaCy | | | | +----------------+-------------+--------+---------------+--------------+
+-------------+-------------+---------------+--------------+ | spaCy | 107,000 | 1.3gb | 96.7 | |
| Stanford | 16,000 | | | +----------------+-------------+--------+---------------+--------------+
+-------------+-------------+---------------+--------------+ | Stanford | 8,000 | 1.5gb | 96.7 | |
| NLTK | | | | +----------------+-------------+--------+---------------+--------------+
+-------------+-------------+---------------+--------------+ | NLTK | 543 | 61mb | 94.0 | |
+----------------+-------------+--------+---------------+--------------+
.. toctree:: .. toctree::