Abstract:Classification over data streams is a crucial task of explosive social stream mining and computing. Efficient learning techniques provide high-quality services in the aspect of content distribution and event browsing. Due to the concept drift and concept evolution in data streams, the classification performance degrades drastically over time. Many existing methods utilize supervised and unsupervised learning strategies. However, supervised strategies require labeled emerging records to update the classifiers, which is unfeasible to work in the practical social stream applications. Although unsupervised strategies are popularly applied to detect concept evolution, it takes tremendous run-time computation cost to run online clustering. To this end, in this paper, we address these major challenges of social stream classification by proposing an efficient incremental semi-supervised classification method named CODES (Classification Over Drifting and Evolving Stream). The proposed CODES method consists of an efficient incremental semi-supervised learning module and a dynamic novelty threshold update module. Thus, in the drifting and evolving social streams, CODES is able to provide: 1) semi-supervised learning ability to eliminate dependency on the labels of emerging records; 2) fast incremental learning with real-time update ability to tackle concept drift; 3) efficient novel class detection ability to tackle concept evolution. Extensive experiments are conducted on several real-world datasets. The results indicate a higher performance than several state-of-the-art methods. CODES achieves efficient learning performance over drifting and evolving social streams, which improves practical significance in the real-world social stream applications.

Social Stream Classification with Emerging New Labels.

Emotion Detection in Online Social Networks: A Multilabel Learning Approach

Semi-Supervised Streaming Learning with Emerging New Labels

Classification under Streaming Emerging New Classes: A Solution Using Completely-Random Trees

A New Ensemble Method for Multi-label Data Stream Classification in Non-stationary Environment

CODES: Efficient Incremental Semi-Supervised Classification over Drifting and Evolving Social Streams

Mining Textual Stream with Partial Labeled Instances Using Ensemble Framework

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

Streaming Classification with Emerging New Class by Class Matrix Sketching.

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

A Novel Semi-Supervised Classification Approach for Evolving Data Streams

A Novel Online Real-Time Classifier for Multi-Label Data Streams

A Feature Weighted Ensemble Classifier on Stream Data

Clustering-based Active Learning Classification towards Data Stream

Active Learning With Drifted Data Streams Using Time-related Weight Classifier Ensemble Framework

Learning from Short Text Streams with Topic Drifts.

Toward online node classification on streaming networks

Classification with Streaming Features: an Emerging-Pattern Mining Approach

A New Semi-Supervised Learning Based Ensemble Classifier For Recurring Data Stream

Multistream Classification with Heterogeneous Feature Space

Online Active Learning with Drifted Data Streams Using Paired Ensemble Framework