HOW DO I FIND THE ERROR DETAILS?

There is a View Log bottom on the bottom right of the screen when the Project Control Panel is displayed. Click on this button and everything Leximancer has done will be displayed. The errors will appear and can be copied and pasted into an email if you need support.

ON WINDOWS, I GET AN ERROR RUNNING GENERATE CONCEPTSEEDS. WHAT COULD CAUSE THIS?

If you are on a Windows computer, this may be because of limitation on the length of file paths for the text extraction utility software.

Try shortening the data file name, and/or the project name (or moving the project to a top-level folder). When you try this, the old error will be shown to you again, and you will need to click Proceed with Retry to try running the project again.

LEXIMANCER REPORTS AND ERROR ON RUN PROJECT. SHOW LOG SAYS “CLUSTER: NO ENTITIES TO CLUSTER.....”

This means Leximancer was unable to create concepts. It is often caused by too little text and default configuration. Leximancer uses a statistical algorithm, so the more text you use the better. If you have a small data set or are using very irregular and sparse data such as Twitter data. there are several settings to consider. These are:

Turning off Prose testing (if the data is known to be clean) Turning off Thesaurus development (concepts are summaries, so providing little text often causes several concepts to merge together during learning, resulting in a sparser map). A small vocabulary in your data does not require a thesaurus abstraction anyway.

PROJECT STOPPED WITH A ‘PREPROCESSOR EXCEPTION’ WITH SOME TEXT SAYING ‘PROCESS EXITED WITH AN ERROR.’

The text may also say ExecuteException near the bottom. Leximancer makes use of text pre-preprocessor programs to extract text from pdf, word, and rtf documents. If running one of these causes an error you may receive the preprocessor exception message.

To find more details about the error, please press on the ‘view log’ button, which can give additional detail about the cause of the problem.

Common causes of errors are:

  • Pdf documents that are password protected or are locked against text extraction

  • Documents that have no text and are just scanned images of text pages

  • Documents with incorrect file extensions (for example, a plain text or rtf document ending with ‘.doc’, which is the extension for Microsoft Word documents.)

  • Documents that are ill-formed for the format specified. (These may have been generated by exporting a document to a different format which is not the native format of the software used.)

Note Windows only: The full pathname of the file that Leximancer is working with exceeds 255 characters. This may happen if you have installed Leximancer in a deep directory hierarchy, the file name itself is very long, and/or you have placed the Leximancer project in multiple enclosing project folders. In this case, you may need to:

  • Move the Leximancer project folder to the top-level instead of nesting it within other project folders.

  • Rename individual files to much shorter names

Try exporting the document to a different format:

If a document seems to be in the correct format, contains text, is not password protected, but still will not load, you may have to open the document in an editor and export the document to a different format. It is usually possible from most text editors/pdf viewers to export to different text file formats, or at the very least to select all the text and copy it. The copied text can then be pasted into the text editor of your choice to save as either plain text (use the extension .txt), or another format such as word (.doc).

If the problem is the Windows file path length, workarounds to try:

  • rename the file to a shorter name

  • create the project at the top-level (not inside project folders)

  • move the Leximancer installation directory to something like c:\Lex4

If there are multiple source documents in a project, Leximancer will only halt processing if all of the documents have no usable text.