Currently many words in the inputs are of questionable similarity to the topic.
The packing densities are now high enough that we can afford to trade off longer-word-lengths for more topical words
Take a closer look at the inputs to understand where to better place the cutoffs, and if the other meta-data included in the json can be used to make more intelligent decisions
Currently many words in the inputs are of questionable similarity to the topic.
The packing densities are now high enough that we can afford to trade off longer-word-lengths for more topical words
Take a closer look at the inputs to understand where to better place the cutoffs, and if the other meta-data included in the json can be used to make more intelligent decisions