NESTOR

NESTOR is a storage and retrieval system for complex analytics on big sequence collections. Data series (a.k.a. sequences, or time series) are present in virtually every domain (e.g., in Internet -of-Things, manufacturing, etc.). Several modern applications involve Terabytes of sequence data. NESTOR enables scientists to transparently use a specialized query processing systems for accessing their sequential data. NESTOR enables non-expert users to easily and efficiently conduct complex data analytics such as similarity search, motif discovery, clustering classification and anomaly detection. NESTOR uses novel summarization, indexing, and analysis techniques for both reducing the size of data series, but also for allowing blazing fast analytics. NESTOR’s storage layer continuously and adaptively reorganizes the underlying data layout to match the current workload, without incurring additional overhead. 

Applications :

• Internet of Things
• Industrial production site monitoring
• Control systems such as SCADA
• Operation monitoring: aeronautics, automobiles, railways
• Smart cities / smart cars / smart buildings
• Health: monitoring physiological parameters

Competitive advantages : 

• Enables very fast complex analytics on very large sequence collections
• No prior knowledge of the domain is necessary
• Anomaly detection: no need of labeled instances (unsupervised method)
• Supports entire spectrum of very fast exact answers to ultra-fast approximate answers
• Approximate answers have deterministic, or probabilistic quality guarantees

Keywords : Data series, Time series management system, Data series, Time series complex analytics and machine learning, Large scale analytics, Pattern matching, Anomalies, Classification