stephbuon / democracy-lab

Code, manuals, and concepts for Democracy Lab research and affiliate projects.
MIT License
0 stars 0 forks source link

Write Blurb for "Clustering" (github repo) #157

Open stephbuon opened 2 years ago

stephbuon commented 2 years ago

(see syllabus for instructions).

HaileyHazen commented 2 years ago

Clustering Clustering classifies and groups documents, paragraphs, or sentences based on similar characteristics they share. The clustering algorithm does this by analyzing similarities between documents and associating them accordingly. First, this process is useful for scholars because it allows them to effectively organize their evidence according to their own criteria. Clustering also helps scholars manage extensive data by grouping it into smaller categories. In short, the clustering process aids scholars by making their research more organized and manageable. For example, a researcher may need to search a database of parliamentary debates for conversations on climate change. A clustering algorithm can analyze the database to create a group of debates that similarly discuss the phrase ‘climate change,’ narrowing a multitude of evidence down to a smaller quantity. The process can further help scholars by organizing the debates on climate change into smaller groups based on their unique characteristics.