Web Detection in Image Analysis: New Insights

Web Detection in Image Analysis: New Insights

New paper out 🎉

I’m pleased to share a new paper published in AoIR Selected Papers: “Web Detection Methods for Contextual Interpretation of Image Collections”, co-authored with Ángeles Briones, Scott Rodgers, Eduardo Vasconcelos and Simon Ceh.

We make two main arguments about the value of web entities and visual matching for image research:

1. Web entities integrate web semantic knowledge into image analysis as they “include strongly related subjects not visually depicted in the image itself” (Leetaru, 2019). 
2. Web entities & Visual matching put us in touch with “the native, nutritive (technical) environment of images” (see Joyeux-Prunel & Zerbib, 2024)

Results show the validity and possible research scenarios for analysing image collections contextually. 𝐄𝐦𝐩𝐢𝐫𝐢𝐜𝐚𝐥 𝐟𝐢𝐧𝐝𝐢𝐧𝐠𝐬 𝐞𝐬𝐭𝐚𝐛𝐥𝐢𝐬𝐡 𝐚 𝐭𝐞𝐜𝐡𝐧𝐢𝐜𝐚𝐥 𝐟𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐰𝐞𝐛 𝐝𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐦𝐞𝐭𝐡𝐨𝐝𝐬, 𝐬𝐮𝐩𝐩𝐨𝐫𝐭𝐢𝐧𝐠 𝐭𝐡𝐞𝐢𝐫 𝐫𝐞𝐩𝐫𝐨𝐝𝐮𝐜𝐢𝐛𝐥𝐞 𝐮𝐬𝐞 𝐢𝐧 𝐜𝐨𝐧𝐭𝐞𝐱𝐭𝐮𝐚𝐥 𝐢𝐦𝐚𝐠𝐞 𝐢𝐧𝐭𝐞𝐫𝐩𝐫𝐞𝐭𝐚𝐭𝐢𝐨𝐧:

𝐌𝐞𝐭𝐡𝐨𝐝 𝐈𝐧𝐬𝐢𝐠𝐡𝐭: web detection privileges authoritative sources and dominant top-level domains, shaping what counts as “context.”

𝐌𝐞𝐭𝐡𝐨𝐝 𝐕𝐚𝐥𝐢𝐝𝐢𝐭𝐲: entity + visual-match signals are time-sensitive, most informative when captured early.

𝐌𝐞𝐭𝐡𝐨𝐝 𝐄𝐯𝐢𝐝𝐞𝐧𝐜𝐞: returned entities mirror contemporary social perceptions and, with longitudinal capture, can surface historical/cultural framings.

𝐌𝐞𝐭𝐡𝐨𝐝 𝐎𝐮𝐭𝐜𝐨𝐦𝐞: web detection maps cross-platform (incl. GenAI/ChatGPT) images and related web content into distinct issue spaces.

If you work with visual collections, platform research, or AI-generated images, I’d love to hear how you’re handling “context” methodologically.

References

Joyeux-Prunel, B., & Zerbib, D. (2024, July 15). The milieu is the message – Digital Materialities and the New Theatre of Visual Operations. ISSUE. https://www.hesge.ch/head/issue/en/publications/milieu-message-beatrice-joyeux-prunel-david-zerbib

Leetaru,K(2019,June1).UsingGoogleVisionAI’sReverseImageSearchtoRichlyCatalogTelevisionNews.Forbes.https://www.forbes.com/sites/kalevleetaru/2019/06/01/using-google-vision-ais-reverse-image-search-to-richly-catalog-television-news/

Omena, . J. J., Briones, Ángeles, Rodgers, S., Liete, E., & Ceh, S. (2026). WEB DETECTION METHODS FOR CONTEXTUAL INTERPRETATION OF IMAGE COLLECTIONS. AoIR Selected Papers of Internet Research. https://doi.org/10.5210/spir.v2024i0.15269 

Omena, J. J., Briones, Á., Rodgers, S., Leite, E., & Ceh, S. (2024). Web detection methods for contextual interpretation of image collections [slides]. Selected Papers of Internet Research. DOI: 10.13140/RG.2.2.35986.31687