Siren released Siren 12, introducing features which remove the need for using many disparate products in investigative intelligence scenarios, which can range from Law Enforcement to Intelligence to Cyber Security. Siren 12 continues to extend the platform’s ability to perform enterprise-wide search, providing sets of links which enables the clear visualization and exploration of relationships between records.
Siren’s latest release makes it much easier for users to organize and join data in a way that suits their requirements, with intuitive UI driven schema editing and ETL. It allows organizations to forensically analyze device data and link it to other available data sources.
Siren 12 enables investigators to not only browse existing information, but also to create new records and edit graphs freely, for the first time merging the “analysis”, the “data entry” and “hypothesis and presentation” phases in investigation in a single intuitive interface.
Lastly Siren doubles down on Investigative AI capabilities by introducing Siren Vision, a deep learning based toolkit for automatic image annotation and classification, integrating with Elastic’s anomaly and outlier detection in a way that is consistent with Siren Investigative use cases.
John Randles, CEO of Siren, comments: “Siren is on a clear course to become the global reference investigative intelligence platform. Version 12 is a huge step forward. It delivers integration of enterprise-wide investigative Search with analytics, while also having the ability to create graphs and tell a story. To perform this function customers previously used expensive legacy software just as a graph drawing board.”
Dr. Giovanni Tummarello, Founder and Chief Product Officer, at Siren, said: “This major new version is a core step in the Siren vision for all-integrated – yet flexible – intelligence. Our customers have started saying “Siren shapes the data you put in”, and this is now more true than ever with enterprise-wide search, in addition we have extended this functionality as users can now create new data per investigation to describe hypothesis and fill missing gaps”.
“We are also proud”, continued Dr. Tummarello, ”to extend the deep integration in the Elasticsearch ecosystem both technically and commercially, with features which leverage Elastic Platinum subscriptions and provide early support for the forthcoming Elasticsearch v8”.
Key features of 12.0
Global Enterprise Search with associative navigation: The new Global Search interface complements the previous specific dashboard searches with instant access to all the data within Siren’s reach in a single unified interface. Once a record has been located, it is now possible to perform “associative browsing” quickly exploring related records.
Data editing and graph editing – per investigation: It is now possible to create records, such as a person or a car, specifically per investigation and also directly work on the graph. This is key to creating hypotheses and summary graph visualizations. Furthermore, it is possible to modify existing records in a way so that the modifications are seen only as part of a specific investigation
New simplified yet extended data import capabilities: Data imports from CSV, Excel or external data sources (e.g. databases and others) has been completely redesigned with a new interface that allows transformation operations using simple drag and drop as well as advanced operations using an imperative language syntax. These transformations can be saved and later reused for new data files which have similar structure.
Data model editing capabilities: System administrators and investigators alike can now create new tables to contain new kinds of entities (e.g. data that was not previously expected) or modify the existing ones adding/removing/deleting fields as required.
New graph capabilities: The graph browser now has a much improved ability to on-demand transform “nodes to edges”, allowing an unprecedented flexibility of viewing certain types of records as nodes or links when required. This greatly improves graph clarity while retaining flexibility for special use cases. New operations added to the graph include node sorting as well as new performance improvements.
Taxonomy and advanced data browsing: A new taxonomy browser and improved taxonomy annotation capabilities in Siren built in NLP work together to allow the exploration of advanced technical and domain specific datasets based on possible multiple, concurrent, multi root, taxonomy terms. Typical use cases include precision drilldowns based on technical terms for patents, technical documentation and scientific literature based on concepts of chemical or technical taxonomies.
Siren Vision preview: For eligible customers, Siren can now provide its beta version of Siren Vision, a machine learning infrastructure to annotate pieces of media with features like classification, extraction of specific objects, OCR and matching based on visual vector projection in search spaces.
Outlier detection and transforms in Siren: Siren now provides the means for customers to leverage Elasticsearch outlier detection capabilities, which enables it to immediately detect outliers based on indexes that describe the behavior of an entity in terms of transactions. Typical examples are immediately spotting customers with anomalous transaction values or quantities based on the automatic backend “rollups” of values collected from all the indexes related to customers.