Joshua S. Hou
Software Engineer, Computational *****************@*****.************
Building great products by writing elegant software with state-of-the-art natural language processing technology.ExperienceSubstantial
Software Engineer, Aug 2012 - PresentIndependent
Software Engineer, July 2011 - Present
Designed, implemented, and deployed a Ruby on Rails app for managing product information and creating product order quotations for Ziggity Systems, Inc, replacing an existing desktop application that ran only on Windows XP
Designed and implemented a Google Chrome extension called Sunglasses for tinting bright browser windows, available for download in the Chrome Web Store, with over 1000 users
Created and built SurvivalPhrases, an Android app for playing audio of survival phrases in different languages for travelers, available for download from the Google Play storeLivestar
Software Engineer, May 2011 - July 2012
Designed, implemented, and maintained a distributed data processing system that crawled the web, identified relevant sources and automatically extracted structured data from semi-structured documents using Ruby, Rails, and Sinatra
Designed and implemented a web system for data entry workers to validate automatically extracted information
Managed deployment and operations of aforementioned systems with Chef
Contributed to Android client codebase, implementing several UI screensKiha Software
Software Engineer, September 2010 - January 2011
Built and maintained statistical named entity recognition models using OpenNLP and Minorthird for a semantic services platform for mobile devices
Developed part-of-speech tagger, tokenizer, and phrase chunker for improving recognition
Feature engineering for improving named entity recognition models
Wrote Maven plugin for automating training and testing of statistical models and continuous integration with Hudson, streamlining the named entity recognition experimentation processGraduate Institute of Applied Linguistics
Research Intern, June 2009 - September 2009As part of the Open Language Archives Community project, developed rule-based and machine learning classifiers in MALLET and NLTK for identifying web-crawled library records as language-related and identifying the subject language of language-related recordsLogos Bible Software
Software Development Engineer Intern - Design and Editorial, July 2008 - September 2008Improved the Libronix Controlled Vocabulary, the flagship product’s controlled vocabulary, with data mining and processing from multiple sources using Python and DjangoEducationUniversity of Washington
M.A. - Computational Linguistics, September 2008 - August 2010
GPA: 3.88
Thesis: Automatic Tone-System Acquisition for Language Documentation in the Field
Relevant Coursework:Shallow Processing for Natural Language ProcessingDeep Processing for Natural Language ProcessingAdvanced Statistical Natural Language ProcessingIntroduction to Speech TechnologyNetwork Security and Cryptography
Funding: National Science Foundation Graduate Research FellowshipUniversity of Washington
B.A. - Linguistics, September 2006 - June 2008
GPA: 3.91
Relevant Coursework:Introduction to Syntax for Computational LinguisticsKnowledge Engineering for Deep Natural Language ProcessingIntroduction to Computational LinguisticsResearch
Automatic Tone-System Acquisition for Language Documentation in the Field
Unsupervised learning of tone systems by clustering vectors of audio features extracted from Mandarin syllables using x-means and other clustering algorithms from Pycluster
3rd i2b2 Shared-Task and Workshop Challenges in NLP for Clinical Data, Medical Extraction Challenge
Participant in the University of Washington’s co-organizing team for the i2b2 Medical Extraction Challenge. Responsibilities included design and implementation of a system for automatic extraction of medications, dosages and other metadata from hospital discharge summaries, to be used as a reference system for challenge competitors
Lexical acquisition system for Lillooet grammar
Implemented English-Lillooet MT system within LOGON infrastructure, Implemented HPSG grammar based on LinGO Grammar Matrix of LillooetSkills
LanguagesIntermediate knowledge of JapaneseBasic familiarity with Spanish and American Sign Language
Links
Projects