I just finished up the doc on a complete parser and object model for the for the U.S. National Library of Medicine’s XML-based Medical Subject Headings (MeSH) distribution. MeSH contains over 2… more →
LingPipe Bloglingpipe wrote 3 days ago: I just finished up the doc on a complete parser and object model for the for the U.S. National Libra … more →
lingpipe wrote 1 week ago: I’ve decided, at least in the first pass, to calculate forward/backward for CRFs on a log scal … more →
lingpipe wrote 2 weeks ago: I’m starting to implement conditional random fields (CRF) for LingPipe. (See, I really don … more →
lingpipe wrote 1 month ago: I’ve just implemented information gain for feature selection, but I’m unhappy with how p … more →
lingpipe wrote 1 month ago: We just released LingPipe 3.8.1, which patches LingPipe 3.8.0. As usual, it’s available from … more →
lingpipe wrote 1 month ago: I’d pretty much stopped worrying about short-lived object allocations in Java. GC’s got … more →
lingpipe wrote 2 months ago: I’m designing a conditional random field (CRF) package for LingPipe, so I’ve been readin … more →
lingpipe wrote 2 months ago: LingPipe 3.8.0 is now available at: LingPipe Home Page This is a big release — the details a … more →
lingpipe wrote 2 months ago: Time to ‘fess up. Sometimes when you look at code you wrote in the past, it looks very ugly. … more →
lingpipe wrote 2 months ago: We’re rolling out a suite of feature extraction utilities and tokenizer factories for the upco … more →
lingpipe wrote 3 months ago: The 7th version of the Java Developer’s Kit (aka JDK 7) delivers quite a speed boost over JDK … more →
lingpipe wrote 3 months ago: The method written up in this paper comes independently evaluated and recommended by Bruce Carneal a … more →
lingpipe wrote 3 months ago: In working on the next release of LingPipe, I’m finding many opportunities to generalize the g … more →
lingpipe wrote 3 months ago: Update, 30 April 2009: OK, the good method mentioned in the links below, and in the final comment of … more →
lingpipe wrote 3 months ago: Softmax is another name for the generalization of the logistic sigmoid function to n outcomes. It … more →
lingpipe wrote 3 months ago: This blog entry’s going to show you how to implement K-means clustering a bit more efficiently … more →
lingpipe wrote 3 months ago: A skill developers typically don’t get in school is how to frame problems in terms of the mess … more →
lingpipe wrote 4 months ago: I thought it’d be convenient to upgrade com.aliasi.util’s BoundedPriorityQueue<E> … more →
lingpipe wrote 4 months ago: We’re busy preparing to launch LingPipe 4.0. We’re leaning toward switching from our … more →