Abstract:Data stream learning has been largely studied for extracting knowledge structures from continuous and rapid data records. As data is evolving on a temporal basis, its underlying knowledge is subject to many challenges. Concept drift,1 as one of core challenge from the stream learning community, is described as changes of statistical properties of the data over time, causing most of machine learning models to be less accurate as changes over time are in unforeseen ways. This is particularly problematic as the evolution of data could derive to dramatic change in knowledge. We address this problem by studying the semantic representation of data streams in the Semantic Web, i.e., ontology streams. Such streams are ordered sequences of data annotated with ontological vocabulary. In particular we exploit three levels of knowledge encoded in ontology streams to deal with concept drifts: i) existence of novel knowledge gained from stream dynamics, ii) significance of knowledge change and evolution, and iii) (in)consistency of knowledge evolution. Such knowledge is encoded as knowledge graph embeddings through a combination of novel representations: entailment vectors, entailment weights, and a consistency vector. We illustrate our approach on classification tasks of supervised learning. Key contributions of the study include: (i) an effective knowledge graph embedding approach for stream ontologies, and (ii) a generic consistent prediction framework with integrated knowledge graph embeddings for dealing with concept drifts. The experiments have shown that our approach provides accurate predictions towards air quality in Beijing and bus delay in Dublin with real world ontology streams. (C) 2021 Elsevier B.V. All rights reserved.

Robust Prototype-Based Learning on Data Streams

Prototype-Based Learning On Concept-Drifting Data Streams

A reliable adaptive prototype-based learning for evolving data streams with limited labels

A Prototype-Based Adaptive Concept Drift Classification Method

Selective Prototype-Based Learning on Concept-Drifting Data Streams

A Better Algorithm for Classifying Data Streams with Concept Drifting

Active Broad Learning with Multi-Objective Evolution for Data Stream Classification

Online Active Learning for Drifting Data Streams

FedStream: Prototype-Based Federated Learning on Distributed Concept-Drifting Data Streams

Learning Evolving Prototypes for Imbalanced Data Stream Classification with Limited Labels

Incremental Learning with Concept Drift Detection and Prototype-based Embeddings for Graph Stream Classification

Knowledge Graph Embeddings for Dealing with Concept Drift in Machine Learning

A Probabilistic Framework for Adapting to Changing and Recurring Concepts in Data Streams

Synchronization-based Semi-Supervised Data Streams Classification with Label Evolution and Extreme Verification Delay

Synchronization-based Clustering on Evolving Data Stream.

Classification and Adaptive Novel Class Detection of Feature-Evolving Data Streams

SACCOS: A Semi-Supervised Framework for Emerging Class Detection and Concept Drift Adaption Over Data Streams

Continual Prototype Evolution: Learning Online from Non-Stationary Data Streams

Online Reliable Semi-supervised Learning on Evolving Data Streams

Learning with Feature Evolvable Streams.

TS-DM: A Time Segmentation-Based Data Stream Learning Method for Concept Drift Adaptation