FAQs STILINGUE - Eyes

Index:

1. How is sentiment classification handled in the case of a post collected by STILINGUE Eyes (meaning, brought in solely through image extraction)?
2. Is it possible to collect Instagram Stories using STILINGUE Eyes?
3. Can retroactive collection of publications be done with STILINGUE Eyes?
4. Is it possible to perform text search based on the text extracted from the image?
5. Are logos captured through image text?
6. Are the texts extracted from images used in the system for any kind of research enrichment?
7. Do texts extracted via STILINGUE Eyes alter the sentiment classification of publications?
8. Can I use both filters of STILINGUE Eyes (AI Collection and Image Transcription) simultaneously?
9. What is the difference between the Image Transcription and AI Collection filters?

How is sentiment classification handled in the case of a post collected by STILINGUE Eyes (meaning, brought in solely through image extraction)?

In this case, sentiment classification will consider only the original text of the publications, regardless of whether it was collected through image text.

Is it possible to collect Instagram Stories using STILINGUE Eyes?

No. Due to an API limitation, collecting Stories via Listening is not possible, and currently, STILINGUE Eyes is available only for Listening. Apart from this limitation, any other image collected via Listening can be gathered.

Can retroactive collection of publications be done with STILINGUE Eyes?

Yes, if your research has STILINGUE Eyes enabled, it is possible to perform retrospective collection of publications using this feature. 

Is it possible to perform text search based on the text extracted from the image?

No, this action cannot be performed. Textual search can only find the original text of the publication.

Are logos captured through image text?

If the logo contains the brand name written, and it is clearly visible in the image, there is a high chance that this content can be transcribed by STILINGUE Eyes.

Are the texts extracted from images used in the system for any kind of research enrichment?

Yes. The texts extracted in term extraction are used to enhance graphics such as Correlated Terms, Terms Over Time, and others in the Discoveries module. These terms enable the extraction of new narratives that were previously hidden in the image.

Do texts extract via STILINGUE Eyes alter the sentiment classification of publications?

No. The text extracted from an image in a publication will not influence the sentiment of that publication. Thus, sentiment classification remains related to the original text of the publication.

Can I use both filters of STILINGUE Eyes (AI Collection and Image Transcription) simultaneously?

You can use both filters simultaneously without any issues; just leave both options checked in the Publication Filter. 

However, it's important to note that every publication filtered by STILINGUE Eyes had text extracted. In the case of AI Collection, the difference is that the publication's collection was based on text extracted from the image, while in Image Transcription, this collection may have occurred through other means, despite having the text extracted. 

If you use both together, keep in mind that the less comprehensive filter (AI Collection) will prevail. If you desire a more comprehensive search, you can choose only Image Transcription.

What is the difference between the Image Transcription and AI Collection filters?

These filters represent two possibilities for visualizing texts extracted from images via STILINGUE Eyes and can be used complementarity for a more in-depth analysis:

By opting for the Image Transcription filter, it will retrieve posts with images that have had some text extracted via Artificial Intelligence. On the other hand, the AI Collected option will show only posts captured by Artificial Intelligence through the extraction of text from images. 

The main difference is that in the case of Image Transcription, the post may have been collected through the match between the extracted text (or the original text of the post) and the search queries. Meanwhile, the AI Collected filter will display only posts that were collected from the text extracted from the image, meaning the identified text in this image was related to the searches included in the Groups. 

This means that the Image Transcription filter can help understand what is being discussed in the texts extracted from these images, while the AI Collected filter will show what is being additionally collected through this functionality. Therefore, these are different analyses that can be used according to the needs of the research.

For more information, visit the discussion on the subject at our community or videos on our channel. 😃