leobeeson / gensim-core-concepts

Gensim Core Concepts - Exploration

As per here though with different sample data and a few tweaks.

Work Breakdown Structure:

High Priority:

• Add computer science/programming controlled vocabularies. #DONE
• Parse software engineering sample_corpus with gensim utilities. #DONE
• Corpus Streaming – Build dictionary streaming one document at a time (data/build_corpus_dictionary.py) #DONE
• Explore Corpus Formats #DONE
• Topics and Transformations #FOCUS
• Create bespoke MyCorpus class in .classes/. #TODO

Mid Priority:

• Design ontology model for software engineering domain. #TODO
  • Include marker for single/few-letter domain words (i.e. "c", "R", etc.)
  • Identify relevance and utility of tags present in stackexchange_tags.tag_description (surrounded by square brackets).
    • Can a reliable relationship be established between them and the parent tag?
• Build modelled controlled vocabulary for software engineering domain. #TODO

Lower Priority:

• Design mechanism to disambiguate single/few-letter domain words from non-domain instances. #TODO

<style> todo { background-color: Yellow; color: SteelBlue } recurrent { background-color: Gold; color: SteelBlue } refactor { background-color: SpringGreen; color: DarkGreen } done { background-color: Green; color: PaleGreen } test { background-color: Coral; color: DarkRed } fixme { background-color: Crimson; color: white } focus { background-color: DeepSkyBlue; color: MediumBlue } reqspec { background-color: MediumPurple; color: Indigo } doubt { background-color: #FF00FF; color: Yellow } wait { background-color: Pink; color: Crimson } blocked { background-color: Crimson; color: Yellow } </style>