till-tietz / rbow Goto Github PK

integrate stem_text and clean_text

in dfm_analysis parts of speech filtering with ps and filter_ps currently only support english. Add support for multiple languages.

in dfm_analysis parts of speech filtering with ps and filter_ps currently only support english. Add support for multiple languages.

ideally bow_analysis and dfm_analysis should support proximity (i.e. word window) and syntactic (i.e. word dependency tree with depth = n) associations between phenomena and descriptors.

in order to construct accurate CIs for your bow_analysis with bow_ci you currently need to replicate window and per_occurrence values you used for bow_analysis in bow_ci. Consider letting bow_ci inheret these values from bow_analysis object or merging bow_ci into bow_analysis (perhaps with ci option).