The colours indicate the pos tags of the words. The words are extracted using textrank from a textbook on international law, and are grouped together on the canvas based on their co-occurrence frequency. Here's a visualization of the force-directed algorithm. ![]() Word-mesh uses spacy's pretrained language models to gather textual features, graph based algorithms to extract keywords, Multidimensional Scaling to place these keywords on the canvas and a force-directed algorithm to optimize inter-word spacing. Keyword filtering: Extracted keywords can be filtered based on their pos tags or whether they are named entities.įont colors and font sizes: These can be set based on the following criteria - word frequency, pos-tags, ranking algorithm score. Word clustering: Words can be grouped together on the canvas based on their semantic similarity, co-occurence frequency, and other properties. Keyword extraction: In addition to 'word frequency' based extraction techniques, word-mesh supports graph based methods like textrank, sgrank and bestcoverage. word-mesh strikes a balance between the two and uses the various statistical, semantic and grammatical features of the text to inform visualization parameters. ![]() Most popular open-source wordcloud generators ( word_cloud, d3-cloud, echarts-wordcloud) focus more on the aesthetics of the visualization than on effectively conveying textual features. ![]() ![]() A wordcloud/wordmesh generator that allows users to extract keywords from text, and create a simple and interpretable wordcloud.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |