Causality is the Infrastructure of Thought
Humans use causality to reason about the past and predict the future from daily life to scientific exploration. However, it is a difficult and time-consuming process to attain causal understanding about how the world ticks
Finding evidence of how biological systems work requires reading tens of thousands of documents, connecting all the dots; all while the amount of scientific publications is increasing at an exponential pace
What we do
Build causal knowledge graphs
Our semantic AI-platform machine reads large corpora of scientific articles and extracts causal associations through the use of our own linguistic and statistical models. Customers can provide their own corpora and combine them with our inhouse knowledge graph
Users search for cause & effect relationships and accumulate evidence on how entities interact
- What are the factors for liver cancer in the context of diabetes?
- What pharmacological substances prevent depression?
- What hypothetical evidence establishes a relationship between rhinosinusitis and asthma?
Advanced causal analytics
Extract causal intelligence that goes beyond semantic search
- What are the comorbidities of a disease in the context of other diseases?
- How does a protein interact with a disease over multiple hops?
- What are potential confounders influencing entity (A) and (B)?
Hypothesis generation & simulation
Run simulations on the knowledge graph
- Disambiguate diseases by computing common and unique causes of a disease
- Predict missing links between proteins, genes, cellular mechanisms or diseases
Natural Language Understanding
Our hybrid AI deconstructs sentences deeply and extracts directional entity relationships using linguistic and machine-derived rules that can be understood and calibrated by humans
Deep Linguistic processing
We understand language at all levels including syntax, semantics and discourse by applying our own linguistic models using grammars, ontologies and dictionaries
Statistical approaches are applied on our own linguistic data sets to further improve the outcomes of semantic relational extraction from our linguistic models
Computational Knowledge graph
Relational Machine Learning approaches are applied on the causal knowledge graph to derive new knowledge such as missing edges, community detection and entity disambiguation
Send us a message
Get in touch with us
20 Ropemaker St, London, EC2Y 9AR
Tzaferi 16, Athens, 11854