StreamStory is a cutting-edge tool designed to analyze multivariate, continuously time-varying data streams. Designed by JSI, StreamStory provides a qualitative and highly interpretable view of complex systems by representing data streams as states and transitions. Whether users upload their datasets or use pre-loaded examples, StreamStory simplifies the analysis process for even the most intricate systems.
How Does StreamStory Work?
- Summarization Through States: The tool organizes data into states, representing different system conditions.
- Modelling Temporal Dynamics: It uses Markov chains to model transitions between states.
- Hierarchical Analysis: States and transitions are aggregated into a hierarchy, enabling exploration across multiple spatial granularities.
- Mapping Back to Domain Concepts: It maps abstract insights back to specific domain knowledge for practical understanding.
StreamStory’s Methodology
The tool follows a structured process:
- Point Cloud Construction: Represents multivariate time series data.
- State Clustering: Partitions the ambient space into conceptual states.
- Transition Modelling: Uses Markov chains to represent dynamics between states.
- State Aggregation: Groups states and transitions into a hierarchical structure.
Monitoring and Real-Time Insights
StreamStory can monitor systems in real time, detecting activities and anomalies. It serves as both an analytical and monitoring tool, making it invaluable for industries that rely on time-sensitive operations.
StreamStory API
StreamStory provides an open API accessible to everyone. The API includes key functions such as:
- Fetching models by UUID.
- Data classification.
- Building or deleting models.
Access the API at StreamStory API.
Service Openness
StreamStory is available as open-source software with a free API. Try it at StreamStory.