V-Spark Online Help

Create a Folder

To create a new Folder, select the blue + button under the Monitored Folder section. Afterward, enter the requested information in the Create New Folder screen:

vspark-creating-folders-org-name.png

The following is a description of each field in this screen:

1

Organization: Drop-down menu of which organization and company the folder belongs to. (This option cannot be modified after the folder is created.)

2

Folder Name: The name of the folder. Folder names are case sensitive. (This option cannot be modified after the folder is created.)

3

Servers: From the drop-down menu, select the name(s) of the V‑Spark servers that you want to use for ASR. You must specify at least one active server.

Note

You may see an x button next to a language model name in the Folders table with the message, "One or more servers does not have this model anymore." This scenario can occur if you have more than one server selected, and those servers do not share the same language models. The scenario will only impact ASR if the model has been removed from all servers.

4

# of speakers: The number of speakers in the audio files that are going to be placed into the folder. All language models work with all supported audio formats. (This option cannot be modified after the folder is created.)

4

Audio Type: For single-channel audio, select Mono. For two-channel audio, select Stereo. If you don’t know what type of audio you have, use the Audio Evaluator tool described in The Audio Evaluator. (This option cannot be modified after the folder is created.)

Note

If your single-channel audio contains two speakers, V‑Spark will automatically diarize, or split, the audio according to speaker. Speakers in the resultant transcript will then be classified as either Agent or Client using a process called side classification.

Note

To prevent diarization of calls that contain more than two speakers, create your folder with “1 speaker” and “mono” settings.

5

Use redaction Text: Cleans the text transcript of any numbers for Payment Card Information (PCI) or other sensitive numbers that are in the audio source. (This option cannot be modified after the folder is created.)

6

Use redaction Audio: Cleans the audio MP3 that gets generated of any locations where numbers exist for Payment Card Information (PCI) or other sensitive numbers that are in the audio source so that these numbers cannot be heard. Removing PCI information from audio and text protects a company from exposure in case of a data security breach. (This option cannot be modified after the folder is created.)

Choose values for these options carefully, as many of them cannot be changed after the folder is created.

Note

When the Use redaction Text and Use redaction Audio options are used with the 1-channel, 2-speaker settings, redaction accuracy is somewhat reduced. For maximum redaction accuracy, select 1-channel, 1-speaker in conjunction with the Use redaction Text and Use redaction Audio options. Note that this selection will prevent diarization, meaning that the 2 speakers will not be separated and distinctly labeled as different speakers on the File Details page.

Once the form has been filled out, select the Next button. The Create New Folder dialog displays additional Folder settings options:

vspark-creating-folders-adv-settings.png

This dialog enables you to set basic and advanced folder options. Select the Next button to continue through the pages of the dialog. After completing the settings, select the Create button to create the folder.

Basic Folder Settings

These settings should be configured for all folders:

1

Link to Applications: Use the drop-down menu to select applications to use for analysis. See Linking/Unlinking Folders and Applications.

2

Model: Select which language models to use for the folder. Folders cannot be created without selecting a model for each channel. Models are domain specific, and affect the accuracy of the transcription process. A "domain" is a type of audio such as voicemail, survey, call center, earnings calls, medical transcription, etc. Each domain has its own jargon, so selecting the proper one for each speaker is important for capturing each word correctly. You may only have one model installed. If you have more than one model available, and you are not sure which model is best for your audio, ask your Voci service representative.

3

Which channel is the agent on?: For stereo files only. Select either 0 or 1, where 0 refers to the left audio channel and 1 refers to the right audio channel. Correct analysis of your audio depends on this value, and it cannot be changed after the folder is created. If later on you realize that the assigned Agent channel is incorrect, create the folder again using the other channel as the agent.

Advanced Folder Settings

These settings should only be configured for folders if the basic settings do not meet your needs.

4

Configure callback delivery method: Select this option to enter a location where JSON files (and, optionally, MP3 or Text) will be sent automatically. See Callback Delivery Method.

5

Add/remove custom metadata fields: Select this option to add or remove custom metadata fields before creating the folder. See Custom Metadata Fields.

6

Add/remove ASR options: Select this option to add and remove ASR tags and view existing ASR options before creating the folder. See ASR Options.

If you selected one or more advanced settings for the folder, after completing all settings, select the Next button to continue. When you have reached the end of the folder configuration, there will be a Create button instead of a Next button. Select the Create button to create the folder.