Improve GenAI accuracy by 30% with Lettria’s Knowledge Studio. Download our free white paper.

Lettria Participates in SemWeb.Pro 2024

Lettria showcased its Ontology Toolkit at SemWeb.Pro 2024, presenting a streamlined approach to semi-automated ontology creation and its real-world applications.

Increase your rag accuracy by 30% with Lettria

Lettria is proud to have participated in SemWeb.Pro 2024, a leading conference dedicated to advancements in the semantic web. Bringing together academics, industry leaders, and technology enthusiasts, SemWeb.Pro provides a platform for exploring innovative tools and methodologies that push the boundaries of data representation and knowledge management.

During this year’s event, Lettria presented its Ontology Toolkit, a solution designed to simplify and streamline ontology generation. This toolkit reflects Lettria’s commitment to empowering organizations with practical, scalable solutions for managing complex information systems.

Julien Plu at the SemWeb Pro Conference

A Focus on Semi-Automated Ontology Generation

The presentation showcased Lettria’s innovative three-step pipeline for semi-automated ontology creation:

  1. Entity Type Generation:
  2. Leveraging input such as text, domain expertise, and use cases, the toolkit generates a list of entity types. Users can modify these outputs to align with specific requirements.
  3. Competency Question Generation:
  4. The system creates competency questions based on the generated types, guiding the ontology structure and properties. These questions help ensure that the ontology serves its intended purpose effectively.
  5. Ontology Generation:
  6. Using the previous outputs, the toolkit produces a complete ontology, including classes, properties, and axioms tailored to the user’s specific context.

The process integrates user feedback at critical stages, ensuring that the outputs are relevant and precise for real-world applications.

Want to see how easy it is to implement GraphRAG?

Extending the Work to the Research Community

Building on the concepts presented at SemWeb.Pro, Lettria also published a paper at the International Semantic Web Conference (ISWC 2024), delving deeper into the methods and evaluations behind the Ontology Toolkit. The paper provides a comprehensive analysis of the semi-automated approach, including insights into quantitative and qualitative assessments. By presenting to both practical and academic audiences, Lettria bridges the gap between research innovation and industry application, driving progress across the semantic web ecosystem.

Outcomes and Feedback

The session at SemWeb.Pro was well-received, sparking meaningful discussions with professionals across various industries. Key outcomes included:

  • Validation of Interest: Attendees engaged enthusiastically, asking numerous questions and expressing interest in Lettria’s approach.
  • Constructive Feedback: Conversations provided actionable insights, highlighting areas for improvement, such as class hierarchies and ontology depth.
  • Real-World Applications: The toolkit’s potential for practical use was evident, reinforcing its value for projects like knowledge graph generation and other semantic web applications.

This feedback will guide the continued refinement of the Ontology Toolkit, with enhancements focusing on prompt optimization, hierarchical structuring, and incremental ontology development.

Advancing Semantic Web Solutions

SemWeb.Pro 2024 reaffirmed the importance of collaboration and idea exchange in driving innovation. Lettria’s participation not only demonstrated its expertise in semantic technologies but also showcased its dedication to solving real-world challenges with accessible and effective tools.

As Lettria continues to evolve its solutions, opportunities like SemWeb.Pro and ISWC 2024 play a vital role in fostering connections and aligning its efforts with the needs of the broader community. The company remains committed to advancing the semantic web through impactful, user-focused innovations.

Ready to revolutionize your RAG?

Callout

Get started with GraphRAG in 2 minutes
Talk to an expert ->