make_picture ()Ĭheckout the Colab notebook for interactive examples ![]() determines a summary for each cluster and # 3. text embedding pt () # This step puts it all together: # 1. Getting started Installation conda create - name pt python = 3.6 conda install - n pt nb_conda_kernels conda activate pt pip install picture_text A simple exampleĬonsider the default values and their result txt = from picture_text.picture_text import PictureText # initializing just sets the text corpus pt = PictureText ( txt ) # Calling the method does the heavy lifting: # 1. For instance: news headlines, natural language questions and social media posts would be good candidates. The approach is intended for grouping large sets of non-domain specific short texts. It also allows the reader to explore each group in more detail by going deeper into a hierarchy and dynamically pulling out of it when needed. Given a corpus of short documents (think news headlines) it can group them into hierarchical groups, that semantically belong together. It defaults to SBERT for text representation, leverages Hierarchical Agglomerative Clustering (HAC) for grouping and tree maps to visualize text interactively. ![]() ![]() PictureText converts a list of short documents to an interactive tree map with minimal code.
0 Comments
Leave a Reply. |