ClinicalMetaScience
commited on
Commit
•
a54b0d4
1
Parent(s):
b9a9981
Update README.md
Browse files
README.md
CHANGED
@@ -13,7 +13,9 @@ SciBERT text classification model for positive and negative results prediction i
|
|
13 |
|
14 |
## Data
|
15 |
We annotated over 1,900 clinical psychology abstracts into two categories: 'positive results only' and 'mixed and negative results', and trained models using SciBERT.
|
16 |
-
The SciBERT model was validated against one in-domain (clinical psychology) and two out-of-domain data sets comprising psychotherapy abstracts. We compared model performance with Random Forest and three further benchmarks: natural language indicators of result types, *p*-values, and abstract length.
|
|
|
|
|
17 |
|
18 |
## Using the Model on Huggingface
|
19 |
The model can be used on Hugginface utilizing the "Hosted inference API" in the window on the right.
|
|
|
13 |
|
14 |
## Data
|
15 |
We annotated over 1,900 clinical psychology abstracts into two categories: 'positive results only' and 'mixed and negative results', and trained models using SciBERT.
|
16 |
+
The SciBERT model was validated against one in-domain (clinical psychology) and two out-of-domain data sets comprising psychotherapy abstracts. We compared model performance with Random Forest and three further benchmarks: natural language indicators of result types, *p*-values, and abstract length.
|
17 |
+
SciBERT outperformed all benchmarks and random forest in in-domain (accuracy: 0.86) and out-of-domain data (accuracy: 0.85-0.88).
|
18 |
+
Further information on documentation, code and data for the project "Publication Bias Research in Clincial Psychology Using Natural Language Processing" can be found on the Github repository [PubBiasDetect](https://github.com/PsyCapsLock/PubBiasDetect).
|
19 |
|
20 |
## Using the Model on Huggingface
|
21 |
The model can be used on Hugginface utilizing the "Hosted inference API" in the window on the right.
|