Description of the challenges faced by the Tech Project
The challenges have been defined based on 2 different concepts from the scenarios discussed earlier. The first one is the chatbot concept for radio, and a second one is to create a lively environment for the radio editorial team. Both concepts are described in somewhat more detail below: 1) Current chatbot technology uses mostly textual interaction with the user, based on the patterns known from instant messaging services. The challenge is to imagine how future and possibly more natural bot interactions could look like, possibly integrating other types of media. 2) The radio editorial team is quite isolated from the events they are talking about, having text and voice messages and maybe a few images as their window to the outside world. In order to better get a feeling what is happening at events the challenge is to make the studio environment lively, visualising the topics the audience responded to as well as their emotions.
Brief description of technology
We develop a service-driven software platform to easily manage listener interaction with the radio editorial team, automation for incoming text, audio, picture or video content by using AI technologies and novel multimedia content analysis technology. More specifically the technical components include machine learning / AI for: - Text analysis, named entity recognition, topic detection, sentiment analysis; - Chatbots and automatic conversational agents; - Image and video analysis enabling e.g. face detection, face recognition, visual descriptor extraction and indexing, visual classification; - Social media analysis. Further services are: - Gathering listener input from different channels, e.g. radio app, social media - Extracting location information from pictures - Search and filtering of the conversations and content Pilots carried out by radio stations under real conditions (in operational environments) play a key role in the project.
What the project is looking to gain from the collaboration and what kind of artist would be suitable
Through the collaboration with an artist we aim to explore new forms of interaction and visual ways for editorial teams to deal with the interaction. Even use-cases, user interfaces and user experiences that are designed in a user-centered way, following an iterative process of consultation with users and technical development can sometimes miss key issues and opportunities. This is especially true for projects like MARCONI, where new forms of interaction are sought. We believe that by involving an artist in the process we could gain an additional perspective on the solutions that we build. We encourage artists to consider addressing one or both of the concepts described above (i.e. chatbot concept for radio and lively environment for the radio editorial team). This does not exclude other applicants who are interested in exploring other aspects of the MARCONI platform and scenarios (as defined in deliverable D1.2 that can be downloaded from the project website).
Resources available to the artist
The project has developed a platform running a number of services to process user contributed content and interactions. This platform, including analysis and chatbot services, as well as content from pilots will be made available to the artist. The artist will have the technical staff of the partners available to provide support if needed. Moreover, it would be possible for the artist to interact with professional radio makers (through the radio stations involved in the piloting activities). We also offer the possibility to host the artist in one of the media partner’s locations.