An Automata Based Authorship Identification System

AN AUTOMATA BASED AUTHORSHIP IDENTIFICATION SYSTEM by Shangxuan Zhang This thesis gives a design and implementation for an authorship identification system based on automata modeling. The writing samples of an author were collected to build a tree and use the ALERGIA algorithm to merge all the compatible states of the tree in order to get a stochastic finite automaton. This automaton represents the writing style of the author. We can use this automaton to test whether an anonymous writing piece belongs to this author. ACKNOWLEDGMENTS Foremost, I would like to thank my advisor Tsau Young Lin for his invaluable insight and inspiring guidance, without which I would have lost my direction and never come to the end of my research. Moreover, I offer my deepest appreciation to Dr. Robert Chun and Dr. Howard Ho for participating in my thesis committee. My special thanks go to Dr. Cay Horstmann for his help in the past two years. Without his email, I would not know my application material wasn't delivered correctly and who I should talk with about this issue. I have been very lucky to have many supportive and loving family members. The person who deserves my gratitude the most is my husband, Baosen Wu. I thank my parents, Chengji Zhang and Tingting Hong, for giving me the freedom to explore my interests. Finally, I would like to express my sincere thanks to the Department of Computer Science at San Jose State University. Without two years study here, I might not be able to be admitted to