In this video, we demonstrate the powerful features of the IOTA Identity framework applied to the SEDIMARK participant onboarding scenario.  It showcases how offering tokenization and access control are enabled in a DLT-enabled ecosystem exchanging assets stored in an IPFS infrastructure.

Mage.ai helps at orchestrating data pipelines throughout the entire data lifecycle, from preprocessing to model creation and prediction. By utilizing data managed by EGM (including weather and water dam information), it supports the development and maintenance of the four pipelines demonstrated in the video. Leveraging Mage.ai’s capabilities enable the creation of reliable workflows that can be executed on demand.

This scenario demonstrates the establishment and management of the Offering lifecycle within the SEDIMARK ecosystem.
The Offering serves as the means of describing Assets available for exchange and used by the Catalogue. During the lifetime of an Offering, an Offering will undergo changes based on the Providers preferences or circumstances, and therefore need to be considered.
The scenario enables the publication of Offerings from Providers to the Marketplace. It also enables Consumers to search and discover Offerings through a distributed Catalogue. This is done through a defined set of interactions between modules within the SEDIMARK toolbox and the Baseline Infrastructure.
The lifecycle is split into 2 main phases, Registration, and Discovery, and breaks down into six stages: Creation, Registration, Population, Discovery, Update, Withdrawal.

The Ontology that Offerings are modelled from can be found here: https://w3id.org/sedimark/ontology/

Overview and Demonstration of Technologies for Trustworthy Data Sharing in SEDIMARK Marketplace. In this video, we present a detailed technical demonstration of the technologies enabling secure and trustworthy data sharing among participants in the SEDIMARK Marketplace. Using Eclipse DataSpace components and based on MVD from the Eclipse Foundation, we showcase how these innovative solutions facilitate efficient and secure collaboration and information exchange.

In this video, we demonstrate an AI pipeline using an enhanced deep learning model, the improved version of Crossformer, applied to the SEDIMARK use case for multivariate time-series forecasting. The pipeline's process, including data loading, multivariate selection, model training, and inference, is also covered.

Crossformer

Dataset Recommender System

This video presents the recommendation module, a key part of the SEDIMARK platform. The goal of the recommendation module is to help the user navigate the platform and retrieve relevant datasets promptly. The recommendation module enables users to search for datasets using query-based or similarity-based search. In the query-based search, the user submits a query and the recommendation module returns a list of the most relevant datasets. In the similarity-based search, users submit a specific dataset, and the recommendation module returns a list of similar datasets.