Back to Projects
NLP

Named Entity Recognition

Application of RNNs for NLP tasks: entity prediction and next character prediction.

Course: Neural NetworksCo-authors: Diego Quezada
Named Entity Recognition

Objectives

  • 1Recognize named entities in text using recurrent neural networks.
  • 2Compare unidirectional versus bidirectional RNN architectures for entity recognition.
  • 3Generate text sequences using character-level language models.

Conclusions

  • The RNN achieved 89% F1-score for named entity recognition on the CoNLL dataset.
  • Bidirectional RNNs did not significantly improve performance over unidirectional architectures for this task.
  • Character-level language models learn syntactic patterns and generate coherent text after sufficient training.

Technologies

  • spaCy
  • FastAPI
  • Keras
  • TensorFlow
  • Matplotlib
  • NumPy
  • Pandas
  • Scikit-learn