Learning syntactic structure
Yoav Seginer · UvA-DARE (University of Amsterdam) · 2007
It was Dick de Jongh who suggested to me the problem of language learning as the subject of this dissertation.At the time, that was not exactly what I had in mind, but Dick managed to convince me in a very simple way: he left me Makoto Kanazawa's work to read, attaching a short note: "another interesting subject".This was the beginning of a long journey together, which Dick supervised with patience and dedication.From beginning to end, his influence was exerted by short comments and questions which had a significant influence on the outcome.I would like to thank him for all this, for his attention to detail and for much more.This work would not have existed without him.Khalil Sima'an was the second major influence on the content of this work.I first met Khalil when I followed his course on statistical methods for natural language processing.This was my first encounter with natural language processing in general and with statistical methods in natural language processing in particular.Though it did not happen in a day, it was Khalil's influence that had made me turn from theoretical mathematical research to empirical work based on corpora.Khalil taught me most of what I know about statistical computational linguistics and much more that I probably should know, I could have learned from him.When it came to discussing my work or anything even remotely related, it was always interesting to talk with Jelle Zuidema.Whatever the subject, Jelle always reminded me what the bigger question was.Jelle was also the organizer and the driving force behind our weekly reading group and seminar.This weekly meeting was a great place for discussion, learning new things and exchanging ideas and the discussion was always lively.It was also great fun.I would like to thank Jelle for organizing these meetings and all the participants for making them such a success: