I led a team, at Microsoft Research, to design and implement an algorithm to detect the credibility of news authors. My algorithm achieved accuracy higher than 80%. This was novel research and, at the time, the literature only handled the credibility of short text (like twitter).
I built the machine learning model, engineered the model features, managed the dataset storage, managed a team of judges to judge the data, and designed the labeling tool for the data.
Our work was published and I presented it in ASE 2014 Big Data Conference at Stanford University.