Bio/research statement

I analyze data. Nowadays it's mostly web data, and at scales which would qualify it as "big data". For that I use machine learning (which I was trained in), as well as NLP and IR. I specialize in observing and analyzing human behavior on the web. The resulting insights translate into new features and products, or, alternatively, into academic research.

I have a Ph.D. from CMU, under Andrew Moore, on "efficient implementations of statistical machine learning algorithms". That led to a widely-cited paper on K-Means (with a widely-used software package), and the main application area was astrophysics. After graduation, I joined the machine learning group of the IBM research lab in Haifa. There I worked on machine learning as applied to water management and (separately) to computer hardware virtualization, and was nominated IBM master inventor. Today, I manage a group at the Haifa branch of Yahoo Labs, where I'm a director.

I support open-source software, and in the past I maintained several FreeBSD ports and contributed documentation to the FreeBSD handbook.