Closed kbastani closed 10 years ago
All similarity measures are now TF-IDF and using a vector space model to classify inputs. Caches for the VSM are invalidated every time the model is trained. The VSM cache refreshed when using “classify” or “similar/{name}”
All similarity measures are now TF-IDF and using a vector space model to classify inputs. Caches for the VSM are invalidated every time the model is trained. The VSM cache refreshed when using “classify” or “similar/{name}”