This section presents a comprehensive list of deliverables produced by ReTV to present the results of the project.
This deliverable outlines the identified media vectors to be supported by the ReTV Trans Vector Platform (TVP) as well as the implementation plan for acquiring data from each vector based on existing APIs and requirements.
This deliverable is an update of D1.1. It covers the topics of content collection across vectors, annotation and knowledge graph alignment, concept-based video abstractions, and brand detection.
This deliverable is an update of D1.2. It covers the topics of annotation and knowledge graph alignment, concept-based video abstractions, and brand detection. Concerning annotation and knowledge graph alignment, it updates on the implementation of an accurate NLP & NEL pipeline for annotating the collected data with respect to keywords and Named Entities and aligning annotations to our Semantic Knowledge Base (SKB) as well as external knowledge sources.
This deliverable presents the plans for implementing the ReTV event extraction capabilities and the temporal annotation of content items, identifies the content-based success and audience metrics to be measured and describes the intended approach to extract metrics across published vectors.
This deliverable presents the first version of the success factors and predictive analytics based on the annotations and metrics derived from TV content which are stored and analysed by the Trans Vector Platform
This deliverable is an update of D2.2. It covers the topics of event collection and modelling,audience and success metrics, and predictive analytics for the success of future online con-tent publication. Specifically, it updates on our collection and modelling of events, audiencemetrics and success metrics for the purpose of using them as inputs to predictive analytical models. It presents the results of using these inputs both individually and in combination toenable content-based success predictions for content owners
This document summarizes the current state of the WP3 tasks in the first phase of ReTV project (month 10).
This deliverable reports on our work to define viewer profiles and ensure data interoperability from varying models and vocabularies by defining mappings of properties and values across metadata specifications. Our overall aim is to deliver components for content adaptation, re-purposing and scheduling and to make recommendation based on annotations, metrics and predictions research.
This deliverable is an update of D3.2 and the final deliverable of WP3. It reports on our improved results in content adaptation, summarization, recommendation and scheduling that are employed in ReTV’s Content Wizard and 4u2 use case applications. We present a new architecture for unsupervised video summarization which works together with the previously developed, non-learning method of video summarization. We present the state-of-the art results of our Text to Video matching service and how we built a search API on top of it. This search API is used in the Content Wizard for improved scheduling of content. We also use Text to Video matching in our extended approach to recommendation in the 4u2 use case applications.
This deliverable presents the technology roadmap of the TVP.
The final version of the TVP combines the technical results from the project partners into a comprehensive modular platform, enabling broadcasters, media archives and other digital content owners to tap into the next generation of digital content publication. The provision of prediction, personalisation and content transformation capabilities will help to maximise the impact of media organisations’ communication activities and the engagement with their audience.
This deliverable presents the process of the definition of the requirements for the Content Owner Use Case. It describes the creation of the relevant ReTV Content Owner target groups,persona types and personas. It also describes the four scenarios from the Content Owner perspective and defines their technical requirements. To validate the initial requirements the partners conducted interviews and sent out a survey.
This deliverable presents the process of the definition of the requirements for the Content Owner Use Case. It describes the creation of the relevant ReTV Content Owner target groups,persona types and personas. It also describes the four scenarios from the Content Owner perspective and defines their technical requirements.
This deliverable presents the final versions of the two professional user applications: Content Wizard for the semi-automated workflow of audiovisual content publication across social media channels, and Topics Compass for media listening and predictive analyses. Both applications were validated by professional users and the gathered results demonstrate the potential impact of the ReTV solutions in operational contexts and point to promising areas for future research.
This deliverable presents the process of the definition of the requirements for the Consumer Use Case. It describes the creation of the relevant ReTV consumer target group, persona types and personas. It also describes the four scenarios from the consumer perspective and defines their technical requirements.
This deliverable focuses on the preparation of the consumer use case scenarios – Content sWitch and 4u2 chatbot., We developed the first prototypes for the two scenarios and executed their evaluation with consumers. The deliverable presents detailed descriptions of the two user scenarios and reports on the results of the user tests.
This deliverable describes how the initial prototypes within the project were further developed to create the 4u2 Messenger application, the 4u2 Smart Speaker Skill and the Video Summarization Web Service.
This document provides an overview of the project website and social media channels that are used in the ReTV project.
This deliverable is a quantitative and qualitative report on planned as well as completed communication and dissemination activities, including measures of success against the stated goals in the ReTV proposal document.