Skip to content

Predicting scientific success with an LSTM Network after mining timeseries citation data from AMiner.org

License

Notifications You must be signed in to change notification settings

eloukas/scientific-success-prediction

Repository files navigation

scientific-success-prediction

About

This repo has been developed as a term project for the Neuro-Fuzzy Computing course in the Univ. Of Thessaly, Greece.

Its purpose is to predict the future citations of a research paper, considering past influences, the impactness of the writers or the conference/journal level, etc.

Outline:

  • Gathered citation data for multiple papers, authors, conferences and journals from AMiner.org
  • Processed & transformed the text data into an understandable time-series dataframe
  • Designed, developed and tuned an LSTM Neural Network
  • Presented results inside the Jupyter Notebook file

Technologies:

  • Python 3.6.x
  • Keras wrapped on TensorFlow
  • sci-kit, pandas, numpy toolkits

Authors:

About

Predicting scientific success with an LSTM Network after mining timeseries citation data from AMiner.org

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published