View non-compliant words in video and image posts in Engagement Dashboards
Updated
This article shows you how to enable compliance for images and videos by generating transcripts. Once the transcript is generated, you can view non-compliant words in image and video posts in the engagement dashboard. This article covers steps for the same.
View Non-Compliant Words In Video Posts In Engagement Dashboards
Click the New Tab icon
. Under the Sprinklr Social tab, click Engagement Dashboards within Engage.The post in the Engagement Dashboard will have a highlighted message at the top, which will reveal the number of compliance issues found in the post.
Double click the message in the post. Details will open in the third pane. You can view Caption Recognition(non-compliance in visual) as well as Speech Recognition(non-compliance in audio), which will reveal where exactly in the video the non-compliant words are placed.
View Non-Compliant Words In Image Posts In Engagement Dashboards
Click the New Tab icon
. Under the Sprinklr Social tab, click Engagement Dashboards within Engage.The post in the Engagement Dashboard will have a highlighted message at the top, which will reveal the number of compliance issues found in the post.
Double click the message in the post. Details will open in the third pane. In the Overview tab, you can view the particular areas of the image, where the non-compliant words have been used.
Limitations
Note:
Presently the feature supports videos of length - 5 min and 512 MB only. This is not a limitation, however self placed regulation. This can be changed, based on the use case. Please connect with your Success Manager.
To be able to generate the transcription, media detail(url) of the video/image is necessary. When a post is published natively, some channels do not provide media detail in the post. For such channels, we can not support transcript generation if published from native. Eg. - YouTube and TikTok
For Instagram channels, if the post contains copyrighted content, the channel does not provide media details. Such posts will fail on transcript generation.
Feature is tested for Brand Posts at the moment. To run over fan/audience posts, its important to first discuss the use case with the Success Manager. ( To be able to evaluate the volume of posts, and required AI resources )
This is only available in the English Language at the moment.
Accuracy of the transcript immensely depends on the various factors like accent used. Having prior information on the accent could help to set up the correct AI Model.
Video compliance feature is not supported for YouTube and TikTok.
Failed State/Skipped State:
If one of the videos fails to generate a transcript, Post is considered to be in Failed State.
If a video goes beyond the cap limit, it will not be sent for transcript generation. It's called Skipped State.