Research talk:Automated classification of article importance/Work log/2017-03-13

Monday, March 13, 2017

Today I will continue working on improving our classifier performance, following up on last Friday's investigation of how various view-based calculations affect performance. I ended up writing up all the findings in Friday's work log, since they all go together.

List of (potential) sources of signal

It is useful to maintain a list of (potential) sources of signal as that will allow us to document and discuss reasons for why certain sources should or should not be included in our model. I've created Research:Automated classification of article importance/Potential sources of signal and will start documenting things there.