AndrewTrieu/fake-news-detection
Archived
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add preprocessing and evaluation for LIAR dataset using DistilBERT

Browse Source
This commit is contained in:
Andrew Trieu
2025-04-03 13:41:14 +03:00
parent 1df0e66bc8
commit 3cf9c715bc
6 changed files with 264 additions and 67 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
archives/fnc2.log
View File

@@ -5,7 +5,7 @@
🪙 Preprocessing text...
🔍 Training models...
📊 Logistic Regression Test Performance:
📊 Logistic Regression FakeNewsCorpus Performance:
precision recall f1-score support
Reliable 0.84 0.90 0.87 54706
@@ -16,7 +16,7 @@
weighted avg 0.83 0.83 0.83 85290
📊 Naïve Bayes Test Performance:
📊 Naïve Bayes FakeNewsCorpus Performance:
precision recall f1-score support
Reliable 0.79 0.92 0.85 54706
@@ -25,3 +25,28 @@ weighted avg 0.83 0.83 0.83 85290
accuracy 0.79 85290
macro avg 0.79 0.74 0.76 85290
weighted avg 0.79 0.79 0.78 85290
📚 Loading LIAR dataset...
🧮 Grouping into binary classes...
🪙 Preprocessing text...
📊 Logistic Regression LIAR Performance:
precision recall f1-score support
Reliable 0.75 0.79 0.77 926
Fake 0.32 0.26 0.29 338
accuracy 0.65 1264
macro avg 0.53 0.53 0.53 1264
weighted avg 0.63 0.65 0.64 1264
📊 Naïve Bayes LIAR Performance:
precision recall f1-score support
Reliable 0.74 0.98 0.84 926
Fake 0.55 0.06 0.11 338
accuracy 0.74 1264
macro avg 0.65 0.52 0.48 1264
weighted avg 0.69 0.74 0.65 1264