My Computer Science Masters at Northwestern and my natural language processing (NLP) focus.

I've been obsessed with language communication between people and computers, ever since my Commodore 64 had a speech command card plug-in. It didn't work well, but I imagined the Star Trek computer the whole time I played around with it.  For my masters’ project I continued my obsession by creating what I called broadly a "Natural Language Processing Interface for Decision Support System".   Very generic title ;-)  The basic idea is a natural language interface between a database of summary information made for the executive VP level and above management to easily access information.

In 1987 this ended up very impractical due to even mainframe limitations in memory and hard drive space at the time. This didn't stop me of course, and I created a project related which focused more on parallel processing of sentences using a grammar. The basic idea was to have different processes parse at the same time the sentence tree to if there were various outcomes. Parallel processing was another area of fascination.  It seems like a bad idea now especially because of the overkill using parallel processing for something simple.  But at the time I thought it was cool and unique. My implementation “worked” but would often bring down the VAX mainframe or get me in trouble. I would end up creating too many processes, and the system administrator would chastise me. As a kid I thought was funny and cool. It's still kind of is :-)

Since those days I have followed along with the entire natural language processing world.  Having co-founded InstallShield that took a lot of my time and attention.  I have waited for the last mile to be finished so that real natural language processing could be a reality. Transcription now becoming very accurate is part of the last mile. It is only a matter of time and effort from us all to solve the human-computer interaction thru language.  With the gobs of RAM available now, plus true parallel processing at a cloud-scale (e.g. Lambda functions) it's time for semantic brains and to really accomplish almost anything.

I’m excited to be doing my part to bring computers with real language ability to greatly benefit people!!