After a tough selection process, the NGI Search 1st Open Call results have been announced. Learn more about the first selected projects:
Search-a-licious
Developed by the Open Food Facts association, the project is an open and pluggable search platform that can be deployed for open data projects. It will have core search features and advanced features like interactive charts and personalised ranking, with future iterations, including NLP. The platform will include the infrastructure, an OpenAPI-compatible API, SDKs in TypeScript and Dart, Widgets in Flutter.
The first application will be for Open Food Facts to make their database more easily explorable. The project will be open source under the AGPL license with a public Git repository for community development.
🔎 Read Open Food Facts Alexis GAREL interview for NGI Search
📺 Replay Alexis Garel (Open Food Facts) Search-a-Licious presentation provided during the first open source webinar by OW2
TalentLayer
TalentLayer is a developer toolkit and open data ecosystem that aims to increase the interoperability of hiring and work platforms by pooling job posts and user profile data. This project seeks to address the fragmentation of the labour market caused by information silos, which makes it harder for users to find the best match for their needs. TalentLayer aims to create an openly searchable labour market where work is accessible and users' reputations are uncensorable.
🔎 Read TalentLayer Kirsten Pomales and Romain Martin interview
Listen to the podcast presentation:
Loading the video player...
ChatDKG
ChatDKG is a platform that aims to provide privacy & trust in search & discovery of crucial knowledge assets aiming to improve environmental sustainability of the built environment in the EU. It improves semantic analysis techniques based on key industrial ontologies and helps users by bringing new ways of discovering/accessing information on a variety of built environment topics while preserving their privacy. The platform uses a decentralized knowledge graph (DKG) to organize & index the information from various internet resources, both public and private, such as scientific, industrial and environmental data or IoT devices. The DKG ensures that information is structured, accurate, trusted, discoverable & and it incentivizes crowdsourced data generation and curation. By combining privacy and trust with cutting-edge search and discovery tools, we aim to create a comprehensive resource for sustainable built environment stakeholders (researchers, civil engineers, architects, real-estate planners…).
Listen to the podcast presentation:
Loading the video player...
HeReFaNMi
The Health-Related Fake News Mitigation project (HeReFaNMi) aims to develop an AI-based automatic system to detect fake news related to healthcare. During the COVID-19 pandemic, there was an increase in phoney information about prevention and infection control. The proposed system will use Continual Learning to discriminate between fake and real news. Current NLP techniques for fake news detection have limitations, with most relying on supervised learning. The proposed system aims to overcome these limitations by continuously updating the knowledge base with emerging events.
Listen to the podcast presentation:
Loading the video player...
LabDiscoveryEngine
The LDE project aims to build an open-source remote lab management system that enables institutions to publish, share, discover, and securely access their physical and educational labs. This system aims to democratise access to equipment and resources, regardless of time and space, and save costs. The project addresses the challenges faced by resource-constrained institutions and promotes equality in education, particularly in STEM fields. The LabDiscoveryEngine builds upon WebLab-Deusto RLMS, which is currently used by many institutions worldwide. The proposed system will support the federation and provide efficient discovery tools, facilitating the development and cross-institutional sharing of new remote labs.
Listen to the podcast presentation:
Loading the video player...
MADADA
The Ma Dada project uses digital commons and language models to improve how citizens search for and request information from public authorities in France. The platform guides users to file Freedom of Information requests and acts as an open data archive, publishing the request history, correspondences, and obtained data. Launched in 2019, the site helped over 500 users make over 2,000 FOI requests. The project aims to simplify the FOI process further and make more information discoverable by leveraging open-source software and commons such as Wikidata and NPL language models.
Listen to the podcast presentation:
Loading the video player...
IoT twinning for digital product passports
The project, supported by the Polytechnic University of Catalonia and the eReuse circular economy community, aims to provide a search and discovery service to access public information about digital devices based on digital twins linked to physical devices through unique decentralised identifiers. The digital twins provide identifiers, details, and verifiability information as a digital product passport to enable circular management of ICT devices. The service collects and discovers points that accumulate over the lifespan of the devices, contributed by various stakeholders, to increase confidence in the second-hand device market and calculate social and environmental impacts.
Listen to the podcast presentation:
Loading the video player...
TrustSearch
TrustSearch is an AI tool that aims to offer a search experience based on critical thinking and present different media approaches to a controversial issue. With a growing mistrust in all information sources, the project aims to create a new generation of search engines that promote critical thinking. By analysing the stance of different media, distinguishing sources in terms of audience, and leading political ideology, users will have access to a plural media landscape to make informed decisions.
Listen to the podcast presentation:
Loading the video player...
SVP64 Power ISA Vector Optimisation for Search
The project aims to improve the energy efficiency of search algorithms by optimising a Vector ISA at the hardware level. Rather than optimising search algorithms for existing ISAs such as ARM, x86, and MIPS, the project will work on identifying the features that can be added to the ISA to make search algorithms more power-efficient.
🔎 Read Luke Leighton (LibreSOC project) interview
Listen to the podcast presentation:
Loading the video player...
On My Disk: search integration
On My Disk offers personal cloud storage on private devices with secure access to third parties. PeARS is a local search engine that indexes and searches web pages while preserving privacy. The two platforms are partnering to offer a decentralised search system over personal cloud data, allowing users to search documents across networked devices. The joint project aims to create a new ecosystem that combines private hosting and distributed web search. More...
🔎 Read Aurelie Herbelot and Alexey Volkov (On My Disk: Search Integration) interview
Listen to the podcast presentation:
Loading the video player...