This is the final project of the course - Mining Web Data for Business Insights. Our chosen topic is utilizing data analytics to flag fake news. The analysis involves descriptive, LDA, stance detection and sentimental analysis. Compared with null accuracy of 50.7%, we achieved accuracy of 93% by ensemble methods and AUC of 0.98.
flagging_fake_news.ipynb
report06_final.docx
Flagging Fake News.pptx
Deivapriya RAJENDIRAN, KOH Jia Xiang, LIAW Sze Wai Sylvia, NG Hui Yi Nicole, XU Yuwen, YAU Chung Yin Jack