Knowledge graphs represent collections of interconnected descriptions of entities. Such entities can describe objects, events, concepts, or any other phenomenon. The purpose of a knowledge graph is to link entities with semantic metadata and to provide context. This way, enriched data can be used for further analysis and advanced analytics.
Natural Language Processing
Until a few years ago, both natural language processing (NLP) and computer vision algorithms were struggling to do well on entity recognition from text and object detection from images. Because of recent progress, these algorithms are starting to move beyond the basic recognition tasks to extracting relationships among objects necessitating a representation in which the extracted relations could be stored for further processing and reasoning.
Graph structures are well suited for feature extraction tasks where the newly generated insights are used for machine learning techniques. Graph algorithms can help identify hidden patterns. For example, graph scientists at AstraZeneca use graphs in conjunction with machine learning to uncover patient journey archetypes and patterns. Such techniques can help in performing early interventions and improve patient outcomes for illnesses like kidney disease.
Where to next?
Sign up for the special Memgraph Webinar and learn how AstraZeneca ingests data sources in the Biological Insights Knowledge Graph (BIKG) and distributes it to data scientists and domain experts. You will also find examples of how this knowledge graph assists scientists in therapeutics development.
This text is a summary of one area that fits perfectly with the application of graphs. Therefore, we would like to have you with us when implementing some of these solutions. Share opinions, experiences and problems you encounter when working with Memgraph on our Discord server. We are here for you and we will help you along the way.