The Experiment This analysis evaluates many variants of the LSA algorithm against some measurements appropriate for an imaginary document summarization UI. This UI displays the most important two sentences with the important theme words highlighted. The measurements try to match the expectations of the user. Supervised v.s. Unsupervised Learning Machine Learning algorithms are classified as supervised, unsupervised, and semi-supervised. A supervised algorithm creates a model (usually statistical) from training data, then applies test data against the model.
Read full article from Uncle Lance's Ultra Whiz Bang: Document Summarization with LSA #2: Test with newspaper articles
No comments:
Post a Comment