Siren: introducing modern investigative intelligence

Siren is a modern investigative intelligence platform – it uses the data schema (ontology) to coherently coordinate BI style dashboards with best in class, full text search, knowledge graph link exploration, domain specific visualizations and more.

Probably the best way to feel its difference is to watch Siren Investigate (the Siren Platform UI component) in action on media and investment data in this quick video.

Siren indexes relation

Unified data model, the key to investigative intelligence

What is it and how can it help ?

An unified data model (or ontology) is a formal representation of the entities and relationships that exist in a domain – to a level of detail that is useful to answer core domain questions.

In a law enforcement example, a business intelligence drill down on a car make and color can be a start but would likely always produce way too many results to find a specific suspect vehicle.

With the capabilities of Siren and the unified data model, the investigator can drill down leveraging other datasets e.g. asking for “cars owned by males in this age group whose mobile was in this approximate location on that date”.

Siren Platform

At scale (on your big data infrastructure)

While previous ontology based tools worked on very limited datasets and were confined to niche applications, Siren Platform is very easy to deploy as it works at scale on top of the data you already have in your existing infrastructure.

Visit our technology page for more details.

Siren Managed

Siren Investigate – investigative intelligence frontend and alerting

Siren Investigate provides investigative intelligence functionality to the end user and is capable of generating alerts automatically when conditions happen in streaming data:

  • Full-text search with support for misspellings, phonetics, relevance ranking, highlighting and more;
  • Business intelligence and scientific visualizations;
  • Relational “set to set” navigation and drilldowns: pivot operation across connected big datasets;
  • Graph/temporal relational link analysis.

Relational drilldowns – across indexes, across systems

Possibly the biggest innovation of Siren Investigate is the unique “relational drill down” capabilities. Backed by Siren Federate/relational join technology, Siren Investigate can provide real-time relational analysis across indices and across external, non-elasticsearch systems (SQL, REST APIs, etc).

Furthermore, at each step of the relational navigation (also called “set to set” navigation), Siren Investigate can visualize and drill down using a variety of tools ranging from fulltext search to BI and link analysis.

Siren Investigate cross index

Time series analytics and temporal components

Time might be just another variable in your data, but it deserves top of the line visualizations and tools. Siren Investigate temporal analytics and components shows connected events coming from one or multiple indices in a single zoomable view.

Knowledge Graph and Link Analysis

Built-in knowledge graph/link analysis

  • Link data cross backends: e-mails in elasticsearch, customer records on Oracle and transactions on hadoop? not a problem (without ETL)
  • Use dashboard filters to enable focused expansions: expand to just the records you want out of billions
  • Graph database queries, suspicious graph pattern search and alerts, shortest paths, graph grouping, graph metrics
  • Clustering and aggregates within the graph: Obtain clarity with big data aggregates on the graph (node clustering into aggregate links), instantaneously via native pushdown to backend systems
  • Beautiful: smoothly animated, powered by best in class in browser link analysis technology

Spatial and Temporal relations on the graph

See the network evolution over time. Explore relations on maps.

Siren Investigate cross index

Statistical graph inference: auto-discovering the missing links

Siren Investigate leverages in full the big data aggregation capabilities of the underlying backend (Elasticsearch and others) to summarize huge amounts of connections into meaningful summary links at interactive speed. Sophisticated algorithms like the weighted significance auto discovery automatically ensure just the most significant connections emerge.

Siren Investigate cross index

Scientific analytics

Siren Investigate comes with a set of components that are highly suited to scientific data visualization. Among this are big data/interactive speed scatterplots, heatmaps, probability density functions and several other visualizations which are key in advanced quantitative analysis.

Siren Alert and Siren Investigate integration

Alerting and reporting

Receive alerts and reports (PDFs in your email) with logic ranging from simple queries to advanced complex event processing (CEP) scripts. Also includes Siren Investigate UI integration.

In the following screenshot, an operator wants to receive an email when a new NASA article appears.

Load balancer problem

Automated statistical anomaly detection

Sometimes setting rules and thresholds for alerts is problematic or not desirable.

Siren Alert scripting capabilities are powerful enough to implement the ebay statistical anomaly detection methodology.

Follow our documented recipes in Siren Alert(Siren Investigate alerting components) to have these alerts on any metric that might require it.

Siren asset protection

Enterprise grade access control, auditing, encryption

Siren Investigate comes with built in enterprise access control – at index, record or field level, auditing of user activity, SSL and encryption at rest, protect data and communications.

Siren Federate

The backend: Siren Federate for Elasticsearch

Siren Federate is a proprietary Siren plug-in for Elasticsearch, which delivers two core capabilities:

Federation, the ability to connect to remote JDBC sources and virtualize them making them respond as local Elasticsearch tables.

High performance, in cluster cross table join capabilities.

Whereas with Elasticsearch alone one can do simple “drilldowns” (restricting a result set by filtering on some of it’s attributes), Federate enables “relational drilldown” (filtering by attributes of relationally connected records) on local and remote sources, a game changing enhancement for investigators.

Featuring sophisticated patent pending join optimizations specifically designed for invested indexes like Elasticsearch, Federate can be also installed on existing Elasticsearch clusters, delivering non plus ultra realtime cross stream investigation capabilities.

Featured Posts

Siren 10.4 now available - efficient “shortest path” in Elasticsearch, ...

Siren "Dashboard 360": associative multi-index dashboards for Elasticsearch ...

Siren Secures $10 Million Series A Investment as Global Demand for ...

Unstructured data discovery: visual large-scale topic clustering is now in ...

Let's talk

Ready to kick-start your project?

Our experts can show you exactly how to leverage your data to uncover powerful insights!

Download PDF

Siren Platform - Product overview white paper

Get a detailed and thorough understanding of the Siren Platform