By leveraging its Co-Innovation Framework and advanced digital technologies, Encora created a solution that significantly accelerated the process of transforming unstructured news articles into structured semantic representations within a knowledge graph. Additionally, it streamlined information retrieval and discovery through a conversational interface. The services provided by Encora include: 

• Innovation Ideation:
  • Utilized Encora’s Co-Innovation Framework
  • Encora’s Data Scientists were responsible to evaluate public research, architectures and technologies providing solution alternatives to address the problem space
  • Client’s domain knowledge was key to identify key use cases to be demonstrated
    
    
• Product Development: Designed and developed a solution that extracts Knowledge Graphs from unstructured news posts, further enriching them with external data sourced from the WikiData ontology
  
  
• Machine Learning & AI Models: Utilized Semantic cluster aggregation combined with Large Language Models (LLM) to resolve object duplication and to infer a multi-layer taxonomy hierarchy from extracted data, providing detailed classification for identified entities
  
  
• LLM-Powered Virtual Assistants: Leveraged multi-step LLM powered approach to identify entities and relationships inside the text and suggest classifications. LLM was also used to enable a chat interface with the knowledge graph, allowing users to extract insights from the data using natural language