metadata
tags:
- spacy
- token-classification
language:
- en
model-index:
- name: en_student_name_detector
results:
- task:
name: NER
type: token-classification
metrics:
- name: NER Precision
type: precision
value: 0.8311688312
- name: NER Recall
type: recall
value: 0.8421052632
- name: NER F Score
type: f_score
value: 0.8366013072
6,293 student writing assignments were submitted as PDF files. All documents were reflection assignments in response to the same prompt in the same online course. Student names were labeled by human raters (one rater per document). A preliminary model was trained and all disagreements between this model and the human annotations were adjudicated by two additional reviewers. The training dataset includes all 6,293 documents, 845 of which include student names. There are 1,155 student name annotations in total.
This model has been packaged using spaCy. It is available as a huggingface model or a pip package. Performance of the model should be evaluated on in-domain data before deployment in production, particularly when confidential information is involved.
