Topic modeling with text networks
Topic Network - topic modeling with text networks
This package introduces a novel method for topic modeling using community detection in complex networks. The algorithms included in the package first create a network of collocated terms, filter out unimportant words based on centrality measures, and then use community detection to reveal the topic groups in the network.
The methods used are language-agnostic, meaning that you can perform the topic modeling on any text in any language.
It is an early version, hence there might be performance issues when modeling big corpora. These are to be resolved in upcoming updates.
To install the current version of the package, use:
pip install topicnetwork
To find the topics, simply use:
import topicnetwork topics = topicnetwork.find_topics(list_of_strings)
For best results, use a text without punctuation and stopwords, and words converted to lowercase. You can perform the cleaning on your English texts with:
text = topicnetwork.clean(list_of_strings)
Package written and maintained by Michal Pikusa (firstname.lastname@example.org)
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size topicnetwork-0.1.4-py3-none-any.whl (2.8 kB)||File type Wheel||Python version py3||Upload date||Hashes View|
|Filename, size topicnetwork-0.1.4.tar.gz (2.7 kB)||File type Source||Python version None||Upload date||Hashes View|
Hashes for topicnetwork-0.1.4-py3-none-any.whl