Introduction

The VIDIZMO Audio Indexer can be configured to generate summaries for transcribed audios and videos on your Portal. It performs abstractive summarization, which means that it analyzes the input text, extracts the most important ideas with context, and presents them in the summary in a coherent manner. You also have the option to exclude specific words during summary generation, which is helpful for moderating content and ensuring that the summaries don't include offensive or inappropriate words. As a substitute for these words, the application uses alternatives that are inoffensive but similar in meaning to ensure consistency of information in the generated summary. 

 

To process audios and videos on your Portal for summarization, you need transcriptions generated on them. If no transcriptions are found prior to processing, then the VIDIZMO Audio Indexer will automatically generate them. For additional information regarding summarization, visit Understanding Summarization in VIDIZMO. 


Pre-requisites

  • You need a Manager+ user role to configure the VIDIZMO Audio Indexer App in your Portal's settings.

Navigating to VIDIZMO Audio Indexer

1. Open the action menu via the button on the top left. 

2. Click the ‘Admin’ dropdown.

3. Select 'Portal Settings' 



4. Click ‘Apps’ on the Portal Settings navigation
5. Select ‘Content Processing’ to view all the processing applications. You will find the VIDIZMO Audio Indexer App here.
6. Click the settings icon to begin its configuration.



Configuring the VIDIZMO Audio Indexer

  1. Media Formats: Select the Media or Evidence Formats you want to perform the summarization on. Select Audio for sound-only files and select Video if your file includes both sound and visuals. If you are on a DEMS Portal, this field appears as ‘Evidence Formats.' 
  2. Insights: Use the dropdown menu to select 'Summarization', it will then be available as an AI Insight that can be processed on your content.  You can also add additional AI Insights in this field to generate them for your content along with the summarization process. 
  3. Forbidden Words: Provide the word(s) that the AI model will avoid generating in the summaries. The model will select the next most probable word(s) (which is similar in meaning) to replace the words in this list. See Forbidden Words for more details.

    Note: This field is case-sensitive, please ensure that you keep capitalization in mind for the words you want to avoid in your summaries. The words in this field also need to be input independent entries. Avoid using sentences or phrases as the application will also count the space between the words as a separate entry.

  4. Automatic Processing: Select the option to turn automatic processing on or off. If the "On" setting is chosen, any uploaded Evidence or Media will undergo automatic processing without requiring manual intervention. 
  5. Use the ‘Reset’ button to restore all the fields to their default values. The default values for the fields are set by your administrator.
  6. Click ‘Save Changes’ to save and finalize your application settings.


  

7. Make sure you have the application enabled so that you can process your content according to your configurations. 



For a guide on how you can perform summarization on your Portal, visit How to Perform Summarization using VIDIZMO Audio Indexer.