A go-to-solution for name-entity recognition and disambiguation
Molecular Connections’ in-house platform MC IDENTIFY provides tailored, methodical, and a multicomponent solution for named entity recognition, disambiguation, and standardization. This proprietary author and affiliation disambiguation system facilitates identification of co-author profiles, referee searches, etc.
EDGE
An integrated, cross-talking, multi-component solution including
- A parser to ingest various forms of inputs (structured and unstructured)
- A module to Identify different categories of named entities, ontologies/CVs of named entities (authors, topics, institutions, places, etc.)
- An AI-based clustering algorithm for automated clustering of similar named entities
- ML rule-based engine to influence the said clustering via plug and play user inputs
- Optional component allowing for manual intervention to disambiguate semi-automated data points
- APIs to access the linked data store consisting of the disambiguated and standardized entities from all/any of the above workflow components
Key product highlights
- Semi-automated system
- Systematic and inclusive disambiguation approach
- Threshold criteria for the initial machine aided clustering based on user inputs
- Outlets to allow normalizations and standardizations with external standards
- Visual interface to manage and review the linked store
