Keep in mind that NeuroSpell is not a magic crystal ball.
It does beautiful things, but it is still experimental.
There may be delays or blockages during developments.
Do not hesitate to come back later if you encounter a problem.
NeuroSpell tries to solve the following problem:
① knowing all the possible ways to misspell something... ② ... what is the more probable way to rewrite all words of the sentence ?
The first part of this problem (misspelling something) is defined in a set of typical cases to take into account.
This list is enriched/improved in time.
If you find a bad corrected error, it's often because we need to enrich the list of typical errors to take into account.
Thanks to you all to provide us with the opportunity to fill some lacks. Don't hesitate to come later, NeuroSpell will certainly be improved.
The second part of this problem (more probable way to rewrite) is computed by the neural network (Deep Learning)
based on the ModernMT technology.
It was trained on 60 million sentences. It takes several days to re-train.
It's possible to enrich training sentences with specific texts of a domain, using special vocabulary and phrasing.
This process is completed (before, during, and after) by an analysis done with LanguageTool
which offer a lexical search, a POS-tagger, a detection of the named entities, grammar rules, and a statistical model.
The result of this analyse is shown below the correction proposed by NeuroSpell.
NeuroSepll doesn't simply search for possible errors, it rewrites everything. If a word is kept unchanged, even an unknown word,
this means that NeuroSpell consider it to be the more probable way to write it, taking into account the whole sentence.
NeuroSpell works globally : by completing the sentence, one can obtain a different correction than what was proposed for what
was entered at the beginning. Writing a small part of a sentence provides with a less efficient result than writing a completed sentence.
Trying to write randomly, outside of a real writing situation, often provides with poor results.
On one side, this creates errors not taken into account in the typical error list. On an other side, it often creates phrasing for which
the neural network was not trained.
By using NeuroSpell, you contribute to its improvement.
You agree that your texts may be stored some time on our servers with your IP address, to be analyzed by our developers.
This site uses some cookies to recognize you on your next visit,
and among others things, no longer display this message anymore.
No data will be provided to third parties.
① write your text as well as you can in the box above. Example : I dont nid to wori, I ave beter tings to do tan corect my erors...
② NeuroSpell suggests you some corrections to improve your sentences. Example : I don'tneed to worry, I havebetterthings to do thancorrect my errors...
• Bold words: correction propositions from NeuroSpell.
• Underlined words: words to which particular attention should be paid, because NeuroSpell has a doubt.
• Italicised words: words not found by NeuroSpell in its dictionary.
• : locations of words or punctuations that NeuroSpell suggests you to delete.
⇰ If you wish to have a document corrected or translated by a professional,
at our email address firstname.lastname@example.org ⇰ If you have a project which requires our Artificial Intelligence and/or Linguistic skills
NeuroSpell uses LanguageTool
and spaCy to finalize the neuronal correction
with a statistical model, a lexical analisis, named-entity recognition, and grammatical rules.
The neural network (Deep Learning) is based on the ModernMT technology.
NeuroSpell is partially open-source via the OpenNeuroSpell project.
Exemple : I remein at yur dysposal for ani aditionel informatin. Has yu kow I an aveilable to mit u at your convenianse. NeuroSpell : I remain at yourdisposal for anyadditionalinformating. Hasyoukow I anavailable to meetyou at your convenience.
informating : Possible spelling mistake found → information, informing, informative, informatics, infuriating, informations, informatise, informatising, informatize, informatizing Has : The pronoun you must be used with Have. → Have kow : Possible spelling mistake found → chow, chough, now, how, low, know, row, rough, bow, tough an : Did you mean am (first person of 'to be')? → am
Exemple : te nattural curs off tese creasis weel emd sun. It wil come souner or latter. te later is te best. NeuroSpell : Thenaturalcursofthesecrisiswillendsoon. It will come sooner or later. telater is the best.
these crisis : Did you mean this crisis or these crises? → this crisis, these crises te : Possible spelling mistake found → the, Te, to, he, we, be, me, ten, tea, tie