- Home
- What We Do
- e[datascientist] platform
- e[datascience]
- e[valuation]
e[valuation]
Scientific challenge
Data value is context-specific and defined by use. Fundamentally, the value of data is characterized by whether its availability would result in a change of decision. Data valuation is a collaborative and conversational process between multiple stakeholders.
Research organizations struggle to assess the value and reliability of an ever-increasing array of knowledge and data assets, applied to the scientific endeavor. Without a systematic framework for ranking data assets in a particular context, the decision-making process is ad hoc, often influenced by the loudest voice in the room.
Life scientists seeking to exploit and explore data are missing opportunities through the inability to share and reuse previously validated work and knowledge assets, at much cost to the enterprise.
e[valuation] instantiates a game-changing collaboration framework to deliver a range of applications, including cohort selection, target prioritization and automated curation to promote understanding of the validity and importance of data, linking business and scientific stakeholders.
Key features
-
Create data valuation models for selection, prioritization and optimization of data and knowledge entity relationships
-
Establish consistent, explainable and referenceable valuation standards, fit-for-purpose in the evaluation of very large networks of data
-
Calibrate, manage, extend and reuse data valuation models based on research goals and ambitions
-
Apply valuation models to datasets to rank and prioritize data, test hypotheses, visualize and export models and results
-
Integrate with other e[datascientist] applications to orchestrate a range of workflows and analyses
Benefits
-
Simplify and standardize the interpretation of highly dimensional and complex constellations of data
-
Drive cross-enterprise standards and best practices in multi-criteria decision making, achieving concordance across R&D
-
Elicit and calibrate expert-led (bayesian) prior data valuation models
-
Transform and accelerate data-driven experimentation to drive reliable and trusted in silico outcomes
-
Translate the tacit knowledge of individuals and teams into organizational intelligence, knowledge and intellectual property