uahmad235 / ner-deep-learning

Implementation of CNN + BiLSTM + CRF architecture in Keras for NER Extraction using CONLL dataset.

Hi there 👋

• I am Machine Learning Engineer with more than 4 years of professional experience in the domain of Data Science and NLP.
• I've worked mostly on NLP and Tabular data analysis by drawing insights from both, structured and unstructured, datasets.
• I also have sound knowledge of CV & Image processing techniques.
• I have sound familarity with SOTA NLP techniques and models like the Attention-based Transformers and Sequence Modelling for solving real-world problems.
• My experience spans over a diverse set of tasks like automation through NLP, Document Similarity, NER, Text Classification etc.
• I enjoy exploring different areas of Machine Learning and its applications in the real-world.
• The future plan is to diversify my knowledge and experience by exploring other areas (like Speech, Computer Vision) and applying those expertise to make a difference in the world.

ner-deep-learning's Issues

Code_Review

• I particularly like the usage of the contect.std_err, Prior to checking your code, I wasn't aware there existed a dedicated function for errors
• Also, in task3.py->DecLogsHandler, there was an unclosed file. You can adopt usage of with statement as it will help make sure that the file resources are always managed well
• A refactoring suggestion is to merge the DecLogsHandler into the task3.py inspect as this is a core functionality of the class.

Apart from that, I really liked the overall approach taken in writing the code.