CasualMallet

CasualMallet is a test application to use the topic modeling features of Mallet (MAchine Learning for LanguagE Toolkit). This is still early in development, so your feedback (bug reports, feature requests) is crucial.

The development of CasualMallet is suspended. All the functionalities are now included in CasualConc. If you would like to continue using CasualMallet, please let me know.

This is still under development, so some features/options may not work at all and there is no guarantee the results are accurate. Your feedback is crucial for further development, so if you try this app, please send bug reports/feature requests, though I cannot guarantee I will incorporate your suggestions/requests (mostly due to my limited programming skills [and limited time]).

Download: CasualMallet (on Google Drive)

Last updated: 2023/08/09
System Requirement: macOS 12.3 Monterey or later
You need to have Mallet and R separately installed on your Mac (the instruction is on the manual)

CasualMallet Manual (on Google Drive)

Last updated: 2023/8/9

Recent Changes

2023/8/9 version 1.1 release

• Fixed: splitting files to create a corpus

• Added: tagging text when creating a corpus

• Added: experimental universal Binary build; if the app crashes frequently, run it in Rosetta

• Added: TreeTagger installer

2023/2/13 a bug fix

• Mallet Process: Token Regex is not applied properly

2023/2/11 bug fixes and some feature enhancements

Bug fixes

• Corpus: group labels were not able to be applied from the context menu

• Corpus Creation: splitting files were not functional

• Topic-Phrase table did not expand with the window

Feature enhancement

• Copying data on tables: you can copy data on the tables from the context menu

• Corpus Creation: create a new corpus only with words that appear on the specified percentage of files in the files

2022/12/31 some feature enhancements

• Loading Data: loading data on a large corpus consumes much less memory

  • Loading Topic-Word-Weight data: you can now select "None" (not loading the data on to table); this can reduce memory consumption significantly

• Show File on Doc-Topic table: you can select different statistics for coloring text

2022/12/12 multiple bug fixes (and some feature enhancement)

• Word Cloud: (bug) the selected statistic was not reflected

• Word Cloud: (feature) seed can be specified to stabilize the result

• Scatter Plot: (bug) the selected statistic was not reflected for some

• Scatter Plot: (bug) Circle size adjust was grayed out

• Heat Map / R Package: gplots package will be automatically installed at start up (if it is not on your system)

• Mallet Process: you can specify memory allocation for Mallet

2022/12/11 added an option to use a stop word list of your own

2022/12/09 fixed the process of selecting the Mallet path when Mallet is not installed via Homebrew (or not installed at all)

2022/12/07 released the first beta version