The Science Behind Our
Language Detective
We combined modern web technologies with classical machine learning to give you lightning-fast, highly accurate language detection directly in your browser.
Naive Bayes Algorithm
At the core sits a Multinomial Naive Bayes classifier implemented in Python using Scikit-Learn. It excels at text classification by mathematically calculating the probability of a language given the frequency of its characters and words.
Comprehensive Dataset
The model was trained on a strictly curated dataset comprising thousands of sample texts natively written in 22 distinct languages. The model breaks text down using a CountVectorizer to learn unique vocabulary distributions.
FastAPI Microservice
The Python machine learning brain operates inside an incredibly fast, minimalist web framework called FastAPI, which evaluates and returns predictions with microsecond latency.
Next.js Edge Delivery
Everything the user interacts with is served statically at the literal edge of the network using modern Next.js React, styled meticulously with Tailwind CSS for beauty and speed.