Week 6

Over the weekend, I read the paper on Bidirectional Inference with the Easiest-First Strategy for Tagging Sequence Data, by Yoshimasa Tsuruoka and Jun’ichi Tsujii, since our POSTagger was written by them. Had some difficulty understanding the paper though.

Found another good resource for studying NLP from Stanford University.

Presentation to the visitors from Swedish Institute of Technology on Monday (15th Sept) did not take place in the end, though, due to a lack of time. Some other project presentations went ahead as planned, but ours and a few others were dropped as it was nearing lunch time.

During the week, I received an email from Ken, asking me to work with him on the memory leak problems on the TextDisplay portion of the code. LTA was asking for the visuals. Looked through Makino's code and did some self-studying online for that purpose. However, by Friday, another email was received from Makino, indicating that most of the memory leak issues have already been solved and more updates will come over the weekend.

A Skype meeting was proposed for Tuesday. I plan to visit the lab on Monday to discuss with Ken about his study on natural language processing and seek some advice.