In Gartner’s Top 10 Strategic Technology Trends for 2018 “The Intelligent Digital Mesh” refers to the “entwining of people, devices, content, and services." For this to happen the AI needs to make sense of everything by relying on natural language’s inherent model of reality.
Semantic AI (or Semantic Web or Symbolic AI) should be used to collaboratively develop workflows that are a good fit to tackle the underlying problem. For example, Machine Learning based Entity Extraction can be combined with Knowledge Graph based Text Mining for the AI to reason with Heterogenous Data Streams on the Intelligent Digital Mesh.
Semantically enriched data serves as a basis for better data quality and provides more options for feature extraction. This results in higher precision of prediction & classification.
Semantic data models bridge the gaps between most used data formats such as XML, relational data, CSV or even unstructured data using NLP and text mining methods. This allows AI to link data across heterogeneous sources and to provide training data sets which are composed of information from structured and unstructured data. The ontology of a domain terminology provides a model of concepts that can be used to form a semantic framework for many data storage, retrieval and analysis tasks.
More work needs to be done on W3C Web Ontology Language (OWL) based data mining query languages, covering a wide spectrum of tasks for data analysis, data classification, data characterization, data discrimination and data clustering. Unlike traditional rule mining where the basic unit of data under consideration is the database record, and the construct unit of a discovered pattern (e.g., an association relationship) is an item that has an atomic value, OWL data mining encounters a large collection of OWL documents, each of which is analogous to a database record, and possesses a graph-like structure.