YaleParser Resumes Classification

YaleParser is a python tool for NLP classification Task and generate databases with this classification. This model is a fineting on named-entity-recognition and zero-shot sequence classifiers. The method works by posing the sequence to be classified as the NLI and Bayesian weigths to construct hypothesis from each candidate label, and stepwise with regex, build a Database.

Design

predict_single('''08/1992-05/1996 BA, Biology, West Virginia University, Morgantown, WV''')

# 'Education'

                                 precision    recall  f1-score   support

        Administrative Position      0.73      0.73      0.73        49
        Appointments                 0.73      0.84      0.79       115
        Bibliography                 0.94      0.83      0.88        87
        Board Certification          0.94      0.77      0.85        44
        Education                    0.86      0.86      0.86       100
        Grants/Clinical Trials       0.94      0.85      0.89        40
        Other                        0.69      0.77      0.73       156
        Patents                      0.98      0.98      0.98        43
        Professional Honors          0.80      0.85      0.82       170
        Professional Service         0.85      0.61      0.71        85

           accuracy                           0.81       889
          macro avg       0.85      0.81      0.82       889
       weighted avg       0.82      0.81      0.81       889