Akademik Veri Yönetim Sistemi
International Journal on Technical and Physical Problems of Engineering, cilt.16, sa.61, ss.322-335, 2024 (Scopus)
Automating Search Engine Optimization (SEO) for HTML images represents a significant innovation by streamlining the generation of alt text descriptions for image elements, a critical component of improving search engine visibility and website traffic. This study leverages the technical capabilities of BLIPProcessor models, specifically the "base" and "large" versions of the "Salesforce/blip-image-captioning" dataset, within the Python and Hugging Face platforms. By automating the "Image-to-Text" process, the system efficiently generates SEO-compatible annotations and file names for HTML images, significantly reducing the time and effort required for manual description creation in image-intensive domains such as e-commerce and news platforms. The findings highlight the technical trade-offs between the models: the "base" model provides rapid and sufficiently descriptive outputs. In contrast, albeit slower, the "large" model generates more nuanced and detailed annotations. Beyond enhancing SEO practices, the automation of image descriptions has practical implications for digital marketing efficiency and web accessibility. By enabling visually impaired users to understand image content better and simplifying the content management process for businesses, this approach enhances inclusivity while driving efficiency in digital content management.