Topic Analysis of Scientific Sentences

An API to apply BERTopic to scientific papers (https://maartengr.github.io/BERTopic/index.html).

source

SentenceClusterAnalysis

 SentenceClusterAnalysis ()

Topic-based analysis functions for a corpus made up of sentences. The goal is to use discourse models (BACKGROUND, OBJECTIVES, METHODS, RESULTS, CONCLUSIONS) to filter each set and then look for semantic patterns / clusters in each subtype to drive analysis.


source

DiscourseType

 DiscourseType (value, names=None, module=None, qualname=None, type=None,
                start=1)

Simple enumeration of different discourse types

  • BACKGROUND = 0
  • OBJECTIVE = 1
  • METHODS = 2
  • RESULTS = 2
  • CONCLUSIONS = 2

source

load_from_file

 load_from_file (fpath)

Simple pickle function to load SentenceClusterAnalysis objects.


source

save_to_file

 save_to_file (obj, fpath)

Simple pickle function to save SentenceClusterAnalysis objects.