Sad news from Cambridge this week- Karen Spärck Jones has died, aged 71. I took her classes on NL systems in my M.Phil,, and she was the advisor on my thesis (a text generation program to describe images). Karen was active in developing real systems since the early days of A.I., and it's hard to find an NL or IE research paper these days that doesn't credit at least one of her papers as background.
I will never forget her reviewing a draft of my thesis, a couple of days before it was due. I had worked very hard on it, and when I emailed it to her, I was proud of its depth and clarity. Next morning, I walked into Karen's office. "Sit down," she said severely, "you've got a lot of work to do." For the next two hours she took every paragraph of every chapter apart. Above all, I had too much discussion on the theoretical challenges of discourse generation, and not nearly enough on the practical merits and drawbacks of my program. Her feedback was incisive and profound (and bluntly delivered...). In the 48 hours remaining, I turned my paper from a ramble in the text generation landscape into a clean, focused description of my solution.
I still have that draft copy somewhere, the pages sliced and scarred by her pencil, every margin and white space battered by now-illegible exclamations. Genius.