The pipeline is ran in the order:
1.Tagging-PreProcessing.ipynb
2.Tagging Model.ipynb
3.FuncLoc Prediction.ipynb
To change to ELMo embeddings, run the Keras_Elmo file instead
2.Tagging Model adapts the model found at https://allennlp.org/tutorials
Keras_Elmo adapts the model found at https://github.com/nxs5899/Named-Entity-Recognition_DeepLearning-keras
Yiyang Gaos' code can be found at https://github.com/gyy150/Thesis_PipeLine_2. The output of the NLP pipeline as well as the raw maintenance work order data will need to be obtained separately.