Release Notes

LIWC-22 Release Notes

v1.11.0 (released on 2025-01-15)

  • Major
    • Several new features to enhance accessibility. This includes options in the settings menu related to in-application contrast (i.e., high-contrast mode), as well as several major under-the-hood modifcations to make the application more accessible to screen readers and OS "reading hints" for Windows. These efforts are partially completed for MacOS as well and development will continue to ensure increased accessibility across platforms.
    • Major reduction in application launch time, as well as the time that it takes to load various tabs (e.g., Color Coding, Case Studies).
    • Fixed a bug related to application launch on Windows that could cause the application to crash.
    • Fixed several bugs with the Contextualizer that could cause issues for users who want to load an external dictionary for analysis.
  • Minor
    • Added default names to "Compare Frequencies" output if user does not provide default table names.
    • Several improvements to license-checking behaviors.
    • Removed the ability to move columns around in the results tables to enhance performance and reduce the potential for buggy behaviors.
    • Fixed a small bug where, when analyzing data via the CLI, internal stop lists were not removing punctuation marks. The CLI behavior is now aligned with the GUI application behavior in this regard.
    • Fixed a bug with LSM output where, under some conditions, the exported columns were not a complete export of what is shown in the results table in the application.
    • Fixed a bug where the CLI would throw an exception when an invalid directory was passed as the path.
    • Fixed several bugs related to CLI reading/handling of XLSX files.
    • Fixed a bug where exceptions can be thrown when user halts analysis via the CLI.
    • Fixed a bug in the dictionary workbench where sorting columns could inadvertently copy values to other columns.
    • Updated the internal LIWC-22 dictionary file to remap a particular racial term to match other categories. For more information, please see the footnote of this blog post.

v1.10.0 (released on 2024-10-14)

  • Major
    • We have overhauled the license validation process/options for users who want to run analyses via the CLI. Now, instead of always having to have the GUI application open, a companion license validator is included. This will make LIWC-22 much easier to run on various environments, especially high-powered computing clusters.
    • Related to the above, we have built a multi-platform license activation option for users interested in purchasing network/site licenses.
    • Added lemmatization settings to Word Frequencies feature.
    • Fixed several issues that would cause the Compare Frequencies analysis to fail under many, many different conditions.
    • Several major additions/overhauls to the Dictionary Workbench: processing time has been improved by approximately 25%, dictionary evaluation now includes several additional metrics and statistics (e.g., item-rest correlations, alpha if removed), etc.
    • Previously, various languages (e.g., Chinese) could not be processed correctly via the CLI on some platforms due to console encoding issues. This has been resolved with a major overhaul of how texts are passed to the console.
    • Fixed a bug that was causing several users to have to wipe out their preferences file to successfully activate the application. This issue seems to have been specific to Mac users after a system/OS update.
    • Fixed a bug where summary measure results would be impacted by which dictionary categories were selected for analysis. Special thanks to Satish Krishnan for bringing this to our attention.
  • Minor
    • Fixed a bug that would cause the Word Frequency analysis to fail when run from the CLI.
    • Fixed an issue where category selection for built-in dictionaries in the LIWC Analysis page would carry over to analyses conducted in the Cast Studies page.
    • Majorly simplified the file extension list presented in the Open File(s) dialog on Windows.
    • Fixed an issue with stop lists that would, for some words, fail to omit words ending with a possessive/contraction.
    • Added the ability to include asterisks in stop lists to omit words. For example, the stop list entry ham* will capture and omit words like hamburger and hamster. Asterisks themselves can be captured literally with a backslash escape character (e.g., ham\*.
    • Fixed an issue where stop lists were being applied slightly different depending on whether you were using the GUI or the CLI. The CLI output now matches that received from the GUI application.
    • In the previous version of the application, rounding was applied to non-LIWC scores in output when analyzing a CSV file. Rounding is now only applied to the LIWC scores, and other numeric values in users' input files are not modified.
    • Fixed an exception that was thrown when the application would try to render a word cloud that contains zero words.
    • Fixed an issue that would cause an application error when attempting to write an output file with no content.
    • Fixed an issue where user-selected output file format was not being uniformly enacted across all files when saving multiple tabs.
    • Fixed a bug that would cause problems when trying to segment files during LSM analysis.
    • Fixed a bug where CLI would throw an exception when console output was selected for the Meaning Extraction Method. Now, a friendly, informational message is provided instead. Who ever said that software lacks heart?
    • Added normative lines to the Arc of Narrative graphs. They were previously present only in the case studies tab, but not for the batch-analysis of texts. Major thanks to Dr. Michelangelo Conoscenti for this suggestion!
    • CLI now handles multiple additional exceptions when user requests console output.
    • Fixed a bug where, in some circumstances, color-coded text that was exported from the Case Studies page would be incorrectly italicized.
    • Fixed a bug specific to Ubuntu that would prevent the application from running when distributed via an archive.
    • Changed the style/layout of results from the Evaluate Dictionary Words feature to improve interpretability and exportability.
    • Added a "no groups" option to the LSM analysis when performed via the CLI.
    • Fixed a bug that would throw an exception when the CLI was passed an invalid column index, such as a negative number.
    • Improved column sorting for Dictionary Workbench page.
    • Improved the warnings/error messages provided to users for several pop-up messages that result from problematic data or analysis settings.
    • Removed several instances of unnecessary logging to the application logs.
    • Resolved an issue where text identifiers were not always being saved when running Language Style Matching analyses.

v1.9.0 (released on 2024-04-15)

  • Major
    • Added the ability to perform an LSM analysis on a CSV file with only one group contained in the entire CSV.
    • Several overhauls to the Meaning Extraction analyses to prevent potential missing output when users analyzed a file that included already-generated LIWC analysis results.
    • Several additional overhauls to the Meaning Extraction analysis tab to ensure that results are always generated when possible. Several minor fixes were implemented to correct bugs that could cause users to receive no output under certain circumstances. To our knowledge, these bugs have not been encountered by any users to date.
  • Minor
    • Some UI "quality of life" changes were made to the Dictionary Workbench.
    • Fixed an issue with the Contextualizer that would cause analysis to fail in rare circumstances.
    • Updated the text of some error messages to make issues more clear for users who encountered them.

v1.8.0 (released on 2024-02-28)

  • Major
    • Added the new "Compare Frequencies" functionality, which allows for a statistical comparison of two different corpora.
    • Added network licenses for classroom and computer lab settings.
    • Added new options to the "Prepare Transcripts" feature, which now allows the aggregation of texts for users within any given transcript. Previously, if a speaker tag was used in multiple transcripts, all texts would be aggregated across all transcripts.
    • Added the Spanish LIWC2007 translation to the list of selectable, built-in dictionaries.
    • Added the Japanese LIWC2015 translation to the list of selectable, built-in dictionaries. Additionally, a Japanese-language tokenizer is included in the application so that users do not have to preprocess Japanese-language texts prior to analysis with LIWC-22.
  • Minor
    • Fixed bug that was causing MEM output to be appear broken for non-default output types.
    • Added multiple internal dictionaries to CLI options for wc mode.
    • Fixed UI issue where data grids were not always resizing properly.
    • Fixed UI to prevent error when attempting to view dictionary posters for internal dictionaries with no associated poster.
    • Fixed a bug that would, under certain conditions, keep the LIWC-22 application process running in the background after user closed application.
    • Added a separate batch script to the academic CLI that allows Windows users to analyze console text, bypassing native OS issues that might break the Windows Terminal/PowerShell encoding and prevent the text from being analyzed correctly.
    • Fixed an issue where files would not be analyzed on Linux systems if their file extension contained upper-case letters.
    • Fixed a bug where the Contextualizer was not appropriately accounting for asterisks (*) in user-supplied term lists.

v1.7.1 (released on 2023-12-01)

  • Major
    • Patched a bug that was causing the LIWC Analysis feature to be overly conservative in suppressing summary measure output when there is not enough information to calculate meaningful metrics.

v1.7.0 (released on 2023-11-29)

  • Major
    • Massively improved error-checking performance in the Dictionary Workbench. For extremely large dictionaries, this could often take hours — now, this process is completed in just a few seconds.
    • Fixed an issue where the Word Cloud feature would show visualize fewer words when a small file was analyzed prior to a larger one.
    • Built a framework that allows for site licensing.
    • Improvements to how emoji are detected/parsed.
    • Updated LIWC Analysis output so that the four summary measures (Analytic, Authentic, Clout, Tone) will produce empty values when there is not enough information in a text to generate a score.
    • Added Marathi translation of LIWC2015 dictionary to available, built-in dictionaries.
    • Fixed Meaning Extraction performance issues that were introduced in previous version where URLs are detected.
  • Minor
    • Updated "Positive Emotion" label in Arc of Narrative report to "Positive Tone" for consistency/clarity.
    • Added exception handling for specific cases where user tries to delete a file that is loaded for analysis.
    • Changed how preference file was saved to ensure that it is updated more regularly in case of application error.
    • Updated some UI elements to make "Detect Speakers" options more clear.
    • Removal of SLF4J logs from CLI log file.
    • Fixed issue where "Download Now" button would not successfully log users in when updated version of application was detected.
    • When preprocessing transcript files, leading and trailing whitespace is now removed.

v1.6.1 (released on 2023-10-05)

  • Minor
    • Fixed a bug that was preventing app functionality for some users following license validation.

v1.6.0 (released on 2023-08-08)

  • Major
    • Added a "URL Capture" feature for specialized handling of URLs in text. This feature uses regular expressions to detect URLs and avoid splitting them into multiple words.
    • Meaning Extraction: When analyzing text in a spreadsheet, LIWC-22 now includes all input data alongside the output.
    • Chinese language tokenization: LIWC-22 is now able to automatically tokenize texts in the Chinese language using the Jieba tokenizer.
    • Chinese-translated LIWC dictionaries are now included in the software itself.
    • Better handling of error handling and recognition of problems when users load a potentially broken custom dictionary file.
    • Fixed a bug that would prevent custom dictionary entries from capturing intended text when the dictionary entry started with a number.
  • Minor
    • Fixed an issue related to how user preferences were stored that was causing application to fail launching for a small number of users.
    • Fixed several small bugs that would prevent word clouds from rendering properly.
    • Reordering of LSM "Compare 2+ files" output.
    • Fixed small bug with Meaning Extraction that would prevent analysis when minimum word count was set to 1 under certain conditions.
    • Fixed color-coding bug that was preventing numbers from being handled correctly.
    • Fixed small bug related to LSM menu option display.

v1.5.0 (released on 2023-04-07)

  • Major
    • Case Studies: Added ability in the "color code text" feature of the to color-code all categories of the LIWC dictionaries. This can make it easy to see not just what words are being captured by the dictionary, but what words are not being captured.
    • Language Style Matching: Added the ability to compare 2+ files, allowing for the easy analysis of LSM for complete texts. Additionally, output from this analysis can be generated in a "matrix-style" comparison.
    • Prepare Transcripts: Added a feature where it is now possible to aggregate all text, by speaker, when preparing transcript files. Previously, all transcripts were converted to a line-by-line spreadsheet — now, users can select to combine texts for each speaker across an entire transcript.
    • Fixed a bug where some dictionaries in the LIWC2007 format would fail to load properly.
    • Language Style Matching: Fixed a bug where a group with only a single speaker could cause an error.
  • Minor
    • Case Studies: Fixed an issue where Chinese and Hindi texts were not being displayed properly in the "color code text" feature.
    • Prepare Transcripts: Fixed an issue where speaker tags would sometimes be included in the speaker text output. This issue was dependent on the formatting of the transcripts themselves and was unlikely to impact most users.
    • Fixed the "new release" notification to include both major and minor versions. Clicking the link for the new release in the application also now takes users directly to the download page.
    • Several minor UI tweaks and fixes.
    • Fixed a bug that would cause an error if users tried to save file to a non-existent location.
    • Language Style Matching: Fixed issue where output would always show absolute file paths even if not selected in the settings menu by user.
    • Prepare Transcripts: Leading and trailing whitespace is now trimmed from prepared texts.
    • Prepare Transcripts: Fixed issue that could arise in special circumstances when analyzing a Zoom-generated transcript.

v1.4.1 (released on 2023-01-30)

  • Minor
    • Improved emoji detection engine
    • Fixed a couple of small issues related to new installations
    • Fixed a small issue with scrollbars in the user interface

v1.4.0 (released on 2023-01-06)

  • Major
    • Added ability to analyze/convert transcripts formatted as .vtt files that are exported from the Zoom platform
    • Emojis now shown in Case Studies → Color Coding output
    • Added possibility to use wildcards (*) in stop lists
    • Added the ability for users to create custom dictionaries in Hindi/Marathi and analyze texts in those languages/scripts
  • Minor
    • Added check/notification about when new versions are released
    • Fixed Case Studies “export to docx” feature
    • Fixed bug that was causing tables to be sorted incorrectly within the application when header rows were clicked
    • Fixed bug where outdated information in preferences files would prevent app from launching
    • Changes to Word Cloud settings options to enhance usability
    • Changes to Word Frequencies options to enhance usability
    • Allow users to change dictionary in Case Studies from any tab
    • Improved Dictionary Workbench performance
    • Fixed minor activation bug
    • Fixed several minor bugs related to punctuation detection
    • Added ability to make dictionary entries that are case-sensitive regular expressions. These entries should be prefixed by %%% in the dictionary. Case-insensitive regexes remain the same and should be prefixed with ~~~

v1.3.1 (released on 2022-11-10)

  • Major
    • Added DE-LIWC2015 dictionary to Contextualizer
    • Added DE-LIWC2015 dictionary to available CLI options
    • Added DE-LIWC2015 dictionary to available Case Studies
  • Minor
    • Fixed several small QoL issues related to the Windows installer

v1.3.0 (released on 2022-09-09)

  • Major
    • Added Debian/Ubuntu builds
    • Added ability to color code texts with any dictionary, including user-provided
    • Significant size reduction in application/installer
  • Minor
    • Fixed minor bug related to “help” button
    • Several performance enhancements

v1.2.0 (released on 2022-05-06)

  • Major
    • Fixed several bugs related to the application launching / failing to launch under specific conditions
  • Minor
    • Fixed segmentation bug that would cause texts to be segmented in a non-balanced manner under rare conditions
    • Fixed bug related to CSV files occasionally breaking during export
    • Fixed several bugs related to loading/displaying options for external dictionary files
    • Several performance enhancements

v1.1.0 (released on 2022-04-15)

  • Major
    • Added the ability to calculate Language Style Matching (LSM) within LIWC-22
    • Added the ability to prepare transcript-like texts into analyzable data at the person-level
  • Minor
    • Fix LIWC-22 poster file(s)
    • Fix Meaning Extraction ignoring apostrophes in some cases
    • Boost the amount of system memory accessible to LIWC-22
    • Fixed standard deviation reporting in “Evaluate Dictionary Words” feature
    • Fixed Word Frequency calculation hanging in CLI under specific circumstances
    • Fixed “find dictionary words” feature - new terms with spaces can be captured
    • Fixed default conversions list for English language
    • Added ability to specify custom decimal precision per output type

v1.0.0 (released on 2022-02-19)

  • Major
    • Official launch of LIWC-22. Planet earth suddenly becomes a slightly more fun place to be.
—