threshold parameters as relative values #3
Description
Build_network() and extract_tf_idf() offer threshold parameters, respectively for edges and frequency. The input is an absolute number. A great feature would be to input a relative value. For edges, a relative value to the average number of edges in the network, for instance. For the frequency, a relative value to the number of words in the vocabulary.
It is especially important for normalizing the edge thresholds when the corpus and the number of citations tend to increase over time.