code

Looking for a proven and innovative developer for your next web project? I am currently available for hire.

I am also available for speaking engagements (English, Japanese, or both) on the multilingual web, practical natural language processing and interfaces (Mozilla Labs' Ubiquity), the future of the web (and WordPress), and design and research in the open source era.

Technical direction and prior work

Three largely independent factors in the past decade have made the applications and serious considerations of natural language crucial to the future of the web. The expansion of the internet has removed the barriers of geography and economics between people, but in so doing has made language the primary divider of modern humanity. More and more human knowledge is being encoded digitally and online, almost all in our natural languages—a format proven time and time again to be challenging for semantically accurate machine analysis. All the while, only iterative improvements have been made to the usability of computers and the web, largely ignoring our most natural form of communication. These factors have conspired to make practical yet theoretically-inspired multilingual natural language processing a crucial field for the 21st century.

Much of my recent work and writings have been on this area of the multilingual web and natural interfaces.

I also have previous professional experience in database-driven website design and development, from dictionaries and school information systems to worldwide gaming infrastructure. Here is some of my recent technical work.

• Mozilla Labs: Ubiquity research
  JavaScript, NLP, community development
  Contributed to the development of Ubiquity, a natural language interface for the browser. All of my Ubiquity research can be found on my blog.
• Gameloft Japan: Online Game Programmer
  PHP, MySQL
  Managed and refined Japanese market web services infrastructure for the world's largest mobile game producer.
• East Meets Left: local events plugin
  PHP, MySQL, WordPress, XHTML/CSS
  As a contractor for East Meets Left, I developed a custom WordPress plugin for their concert events site. The plugin lets users choose their target city and see only event times related to their location. The city preference is stored in a cookie or, if the user has logged in, WordPress's own user metadata system.
  

  
• GeoPro: total site coding
  XHTML/CSS, JavaScript
  GeoPro is a Taipei-based business location intelligence company. I implemented their website redesign, taking a layout created by their design team and translating it into standards-compliant XHTML and CSS. JavaScript with the mootools library was used to add animations and interactive features.
  

  
• The Academic Approach: TeX automation for the web
  TeX, perl
  The Academic Approach is a tutoring and test preparation company where I worked on various TeX-related projects. I led the development of curriculum typesetting applications for the online SAT math course product. My application parses the LaTeX source and automatically exports lessons and exercises into transparent PNG images, resulting in a beautiful, interactive course. The same LaTeX source can also be typeset for print.
  

  
• University of Chicago Humanities: automated site templating
  perl, XHTML
  The new website of the University of Chicago's Division of the Humanities. The design was developed by the talented folk at Chicago's own Studio Blue. The content was written by a team of people in the Division on a Confluence wiki. I then wrote scripts in perl which automatically extracted the Confluence content and plugged them into the templates, generating the entire website in two seconds.
  

  
• Concordia Language Villages: student information system
  PHP, MySQL, XHTML/CSS
  viper, an online student information system for Concordia Language Villages, specializing in the management and creation of student session summaries. Written in PHP with a MySQL database, the system features a robust permissions system, and PDF output for final mailings, along with email distribution of summaries.
  

  
• Kanji alive: bilingual website code, AJAX
  XHTML/CSS, PHP, MySQL, JavaScript (AJAX), QuickTime, Flash
  Kanji alive, "a searchable, web-based tool to help beginning and intermediate level Japanese language learners read and write kanji. It is freely available, cross-platform, and does not require any Japanese fonts." The system is currently in an interactive QuickTime format, served by a PHP/MySQL backend. I also led the development of a new AJAX front-end, utilitizing a Flash media layer with advanced ActionScript.

Just for fun

If you actively use these plugins or applications, please consider donating to mitcho (Michael Erlewine).

• HookPress
  PHP, WordPress
  Webhooks are a simple paradigm for developing instant notifications and mashups based on simple HTTP requests. With HookPress you can set up webhooks so that a specified URL (a public service or something you set up) is requested when certain WordPress actions occur. Possible uses include generating push notifications or using non-PHP web technology to extend WordPress.
  For more information, visit my HookPress page.

  HookPress 0.1.7
  9 kb - zip
• Naomi Ann Erlewine
  personal branding, design, HTML/CSS
  Designed business cards and website for my sister, Naomi Erlewine, currently a student at the International Christian University in Tokyo.
• select Command for Ubiquity
  Ubiquity, JavaScript, PHP, MySQL
  The select command for Ubiquity lets you execute SQL select statements from Ubiquity. select features:

  • smart suggestions, offering up available table and column names,
  • some simple parsing to block multi-statement queries so you can't injection-attack yourself, and
  • smart insert, so you can immediately use your results as a table or as a single value.

  
  For more information visit my select page.
• Yet Another Related Posts Plugin
  PHP, MySQL, WordPress
  Yet Another Related Posts Plugin (YARPP) automatically returns a list of related entries based on keyword matches, tags, and categories, limited by a certain relatedness threshold. Now with RSS support!
  For more information, visit my YARPP page.

  
• Markdown for WordPress and bbPress
  PHP, WordPress, bbPress
  Markdown syntax allows you to write using an easy-to-read, easy-to-write plain text format. Markdown for WordPress and bbPress (Markdown-WPBB) is based on the famed PHP Markdown Extra by Michel Fortin, in turn based on the original Perl version by John Gruber. All I did was make the same package work with bbPress too. ^^

  Markdown for WordPress and bbPress 1.0
  32 kb - zip
• WordPress Plugin Downloads Pepper for Mint
  PHP, Mint
  This "Pepper" displays WordPress plugin download stats (from wordpress.org/extend) in Mint, the wonderful web site analytics application.
  For more information and to download, visit WordPress Downloads in the Peppermill.

  
  
• WP-Smartdate
  PHP, WordPress
  WP-Smartdate is a WordPress plugin which looks for relative date expressions in your blog posts, such as “tomorrow,” “this coming Monday,” “last Friday,” and adds the date reference (like “2007-11-26”) as a machine-readable microformat. For more information, please read Introducting Smartdate on my blog.
  Extended description on Wordpress.org.

  WP-Smartdate 0.1
  4 kb - zip
• Mailplane Japanese localization
  Localization
  Mailplane is an upcoming standalone interface to Gmail, bringing better Mac OS X integration to the web-based application. I am in charge of its Japanese localization.

Academic

• Extensible Authorship Identification
  Concept, PHP, Perl
  Extensible Authorship Identification (EAI) is a modular system for authorship identification: given two authors and training corpora of their writing, identifying which of the authors most likely wrote a third text. The unique algorithm first identifies in what ways the two known authors diverge and stresses those tests in computing the new text's authorship index. This was my final project for CMSC 35100, Natural Language Processing, written together with Jesse Farmer. It is written in Python and Perl and requires Numerical Python and NLTK Lite.

  Extensible Authorship Identification, the technical report
  122 kb - pdf

  Extensible Authorship Identification source code
  1.31 mb - zip
• Autolexical Resources page
  XHTML/CSS, PHP, BibTeX
  I am also currently working on a new resource page for Autolexical Syntax, a theory of grammar introduced by Jerry Sadock, to be hosted at the Syntax, Semantics, and Pragmatics at the University of Chicago page. The page will host an extensive bibliography of papers written in the framework with downloadable PDFs, maintained through BibTeX.

Whether it's work or play, I'm always on my computer (Apple MacBook Pro). If you're interested in seeing what applications and tools I use, check out my iusethis page.