The other day, Daniel Lemire posted a comment extolling Peter Turney as someone who does a great job blogging about his research. His blog, Apperceptual, is one of the highest-quality blogs I’ve seen in the information retrieval community.
Turney is a Research Officer at Canada’s National Research Council (NRC) Institute for Information Technology. His two decades of research cover a broad spectrum of topics in machine learning, information retrieval, and computational linguistics. Moreover, the practial orientation of the NRC helps ensure that Peter’s scholarly work is grounded in the real-world.
The best way to get a feeling for Turney’s blog is to read it. Here are a few posts I’d suggest:
- A Uniform Approach to Analogies, Synonyms, Antonyms, and Associations
- SVD, Variance, and Sparsity
- Beyond Proportional Analogy
This last post, published today, offers a promising approach towards establishing analogies as the central problem in a theory of semantics. Or, as Turney quotes Douglas Hofstadter, that “all meaning comes from analogies”.
Turney doesn’t crank out blog posts daily or even weekly–he sometimes goes for over a month between posts. But what he does write is well worth reading.