Advanced Applications and Tools for an Intelligent Archive

The ADIMEN project emerges as the natural continuation and evolution of GAMES, a project funded under the Hazitek 2018–2020 calls and recently completed. GAMES enabled the development of a solution capable of automating the extraction and generation of new information from multimedia content in text, video, and audio formats within the communication ecosystem of the 66 local media outlets that make up the TOKIKOM network.

ADIMEN aims to advance and enrich this solution by incorporating new capabilities and enhancing existing technological modules. The project includes the integration of new AI-driven analysis systems, such as machine translation, biometric analysis, and image processing, while also expanding the range of sources to include image-based content. It will also deepen the development of technologies already in place, particularly automatic summarization in Basque and speech recognition for dialects with limited linguistic resources.

Once the information is extracted, the project will apply advanced data mining, analytics, and visualization techniques supported by Big Data technologies, enabling a more comprehensive and modern exploitation of the generated knowledge. At the same time, ADIMEN will undertake the homogenization of the more than twenty databases within the TOKIKOM network, which currently vary in structure and storage methods, with the goal of creating a centralized and efficient search and indexing system.

Finally, the project will drive the development of advanced applications that demonstrate the potential of the integrated technology. Examples include a personalized alarm system embedded in conversational assistants, capable of reading local news with synthetic voices adapted to each region or dialect, as well as the automatic generation of mini websites organized using metadata created by the AI modules, such as entities, topics, or summaries.

Overall, ADIMEN represents an innovative and forward-looking solution for the advanced digital archiving of multimedia content in all formats. It combines R&D activities in Artificial Intelligence with state-of-the-art data analytics and visualization techniques, while also integrating centralized search and indexing systems that connect multiple knowledge bases, delivering a significant technological leap for the TOKIKOM network and its information ecosystem.