This project compares the effectiveness of three language detection models: a statistical model based on n-grams, a basic model using bag-of-words with trigrams, and an LSTM-based model. Developed in Python, the project leverages scikit-learn, TensorFlow, pandas, and Keras for implementation.
Models can detect following languages: German (deu), English (eng), Spanish (spa), French (fra), Polish (pol), Portuguese (por), Dutch (nld)
scikit-learn: Provides functions for model creation, training, and vectorization.Tensorflow: Enables deep learning and neural network operations.DataFrame: Organizes and flexibly uses data.pandas: Assists in data manipulation.keras: Facilitates sequential processing.matplotlib: Aids in creating performance charts.