Entity resolution, the task of automatically determining which mentions refer to the same real-world entity, is a crucial aspect of knowledge base construction and management. However, performing this task for a large data could be very challenging and could lead to problems of rule-based mention extraction system. The system models has to handle the mention extraction through predefined rules. We used a deep learning model for entity identification problem in order to recognize all the entities present in a text. Moreover another model is been used to classify the entities and generate the features correspondingly. These features are combined to form the vector representation of the mentions in order to solve coreference resolution problem. We would use concepts of deep learning and Natural Language Processing to build a model mostly using recurrent neural network and LSTM to provide a better working and scalable solution to entity linking.
For complete instructions to setup and run the project please refer to ReadMe.txt
./code/- contains all required programs and also stores the saved graphs and log files./dataset/- contains the Rich ERE corpora./ere_dict/and./ere_dict_new/- contains saved dictionaries created fron parsing the dataset./weights/- contain saved weights of different models./PIC.png- showing needed file./ReadMe.txt- contains all the essential instructions./manual.pdf- contains solar setup instructions and glove indexing instructions
Bhiman Kumar Baghel - 17CS60R74
Shah Smit Ketankumar - 17CS60R72
Hussain Jagirdar - 17CS60R83
Nikhil Agarwal - 17CS60R70
Lal Sridhar Vaishnava - 17CS60R39
For any clarification contact us @[email protected]