DEV Community

Natural Language Processing

NLP is Natural Language Processing the technology behind Home assistants and search engines.

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
Where context decides the reading: a Japanese homograph tagger (0.723 vs the dictionary's 0.557)

Where context decides the reading: a Japanese homograph tagger (0.723 vs the dictionary's 0.557)

Comments
4 min read
Beyond Closed-World Jev: Training-Free Choice, Yes/No, and Open-Word Extraction in One Tree Pass

Beyond Closed-World Jev: Training-Free Choice, Yes/No, and Open-Word Extraction in One Tree Pass

Comments
32 min read
Freezing a model that caught 2% of real errors: OCR error detection for prewar Japanese print

Freezing a model that caught 2% of real errors: OCR error detection for prewar Japanese print

Comments 1
5 min read
"No wait, I meant...": when my self-correction feature ate people's sentences

"No wait, I meant...": when my self-correction feature ate people's sentences

Comments
4 min read
A Japanese error detector that points at the suspicious span and stops there (F1 0.867)

A Japanese error detector that points at the suspicious span and stops there (F1 0.867)

Comments
4 min read
A document router for Japanese business text that is allowed to say "I don't know"

A document router for Japanese business text that is allowed to say "I don't know"

Comments
4 min read
A Japanese toxicity detector whose headline metric is not flagging legitimate criticism (Macro F1 0.975)

A Japanese toxicity detector whose headline metric is not flagging legitimate criticism (Macro F1 0.975)

Comments
4 min read
Putting the punctuation back into Japanese ASR transcripts (F1 0.89 vs Whisper's own 0.61)

Putting the punctuation back into Japanese ASR transcripts (F1 0.89 vs Whisper's own 0.61)

Comments
4 min read
Extracting Japanese addresses and refusing to guess the municipality code (span F1 0.998)

Extracting Japanese addresses and refusing to guess the municipality code (span F1 0.998)

Comments
4 min read
You're Learning LLMs in the Wrong Order

You're Learning LLMs in the Wrong Order

Comments
11 min read
MLB Data in Python: Keep Current Teams out of Historical RAG Answers

MLB Data in Python: Keep Current Teams out of Historical RAG Answers

Comments
6 min read
Testing Summarization Prompts: Numbers, Hedges, Negations and Omissions

Testing Summarization Prompts: Numbers, Hedges, Negations and Omissions

Comments 1
8 min read
Following one German word through my app, from a textbook page to static HTML

Following one German word through my app, from a textbook page to static HTML

2
Comments
8 min read
Build a Reproducible Portuguese NLP Pipeline in Colab with sri

Build a Reproducible Portuguese NLP Pipeline in Colab with sri

Comments
7 min read
Um modelo de 1.7B consegue igualar um LLM 4,7x maior? Destilando o Qwen3-8B com QLoRA em avaliações reais em PT-BR

Um modelo de 1.7B consegue igualar um LLM 4,7x maior? Destilando o Qwen3-8B com QLoRA em avaliações reais em PT-BR

1
Comments
9 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.