# atpaino/deep-text-corrector

Deep learning models trained to correct input errors in short, message-like text

Repository: https://github.com/atpaino/deep-text-corrector
Canonical: https://ross.abutalabs.com/products/deep-text-corrector
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2019-05-01T18:01:25+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3586, "days_push": 2681, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1235, forks 258 (observed 2026-08-28T04:04:04.835692+00:00)

## What it is
A TensorFlow research project that trains sequence-to-sequence deep learning models to correct small grammatical errors in short conversational English text, such as SMS messages. It builds training pairs by randomly injecting errors into grammatically correct corpora (e.g., the Cornell Movie-Dialogs Corpus) so the model learns to restore the original text.

## Use cases
- correct grammar mistakes in short text messages
- fix missing articles in learner English writing
- train a sequence-to-sequence model for text correction
- generate synthetic training data for grammatical error correction
- build an autocorrect that catches grammatical errors spell-check misses
- experiment with seq2seq models for text normalization

## When to choose
- You want a reference implementation of seq2seq grammar correction with TensorFlow
- You are studying synthetic error-injection techniques for generating NLP training data
- You are reproducing or extending the accompanying blog post / CoNLL-style grammatical error correction experiments

## When to avoid
- You need a maintained, production-ready grammar checker — the project has had no releases since 2019
- You need to correct long-form or formal text — models were trained only on short conversational sentences with a limited set of error types
- You want modern transformer- or LLM-based correction — this uses older RNN-based seq2seq architectures
- You need general-purpose spell-checking or context-sensitive autocorrection

## Facets
- artifact type: library
- maturity: abandoned
- function: machine-learning, deep-learning, nlp, data-generation
- domain: machine-learning, deep-learning
- platform: python
- tags: grammar-correction, sequence-to-sequence, seq2seq, tensorflow, text-correction, synthetic-data, conversational-text, research-code, natural-language-processing, gpu

## Member repositories
- atpaino/deep-text-corrector (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:04.835692+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T08:22:03.836022+00:00, confidence not recorded.
  - readme: https://github.com/atpaino/deep-text-corrector (fetched 2026-08-28T04:04:04.835692+00:00, sha 033c4a8e6045)
- Data as of 2026-08-30T08:39:29.467469+00:00.