ProHealth Summer REU – Day 4 – May 26, 2016

We started the morning with another round of LaTeX to settle any unanswered questions, then confirmed documentation procedures for some future projects.  (I’ll need to update my post from Wednesday to reflect some changes.)

We followed with Ben giving a quick talk on elevator pitches, and some practice and criticism with our classmates.  His talk also included some recommendations to work on our personal websites and LinkedIn Profiles.

We spent most of the afternoon with Professor Connelly, starting with her research area assisting at-risk populations and some followup discussion.  Finally we worked for a few hours on app development, our group created a basic trivia app with radio buttons that we could expand on later if we wanted.

At the end of the day I received an email confirming myself and Vanessa Pereira would be working on the machine learning project with Dr. Natarajan and Devendra Dhami.

Task: Develop a method for automatically extracting adverse drug events from text data. The text considered could be medical abstracts and/or social blogs. The goal is, given a drug and its potentially adverse event, the system should infer whether the event is a side-effect of the drug. To do this, the system should mine literature and other textual sources and obtain the necessary confidence scores. The students will work with a team of graduate students in designing and developing machine learning models.


  • A learning algorithm that can potentially infer the drug event associations from text data
  • Ability to provide evidence for the result of the inference
  • A paper that can be submitted to a workshop or a conference focused on medical data mining.