Watchman Analytics™ enables organizations to achieve more comprehensive, near-real-time data analysis and intelligence extraction from a variety of disparate structured and unstructured data sources. Leveraging best in class data analysis models and algorithms such as Higher-Order Low-Resource Learning (HO-LRL™), Watchman Analytics™ drives information discovery and insights that are normally not possible due to data silos, varying data formats and schemas, and the fact that many useful data sources are unstructured and fragmented.
The Watchman Analytics™ core platform and the array of optional components are extensible and scalable in virtually any technical infrastructure, complementing an organization’s existing IT investments.
Intuidex Watchman Analytics™, based on its open architecture and modular design, can be installed on virtually any IT infrastructure, and operate invisibly “in the background” or if preferred, leverage the Watchman Analytics™ UI components.
The Watchman Analytics™ platform is comprised of a number of core components and functionality, that are delivered in a service oriented architecture, and can be deployed ‘standalone’, or augmented with a selection of optional components based on client need.
Watchman Analytics™ allows users to drill deeply into enterprise information to locate key content and gain understanding. Data stores are often hierarchies of thousands of attributes and values. The technology facilitates rapid, automated indexing of those attributes and value frequencies to deliver accurate content, even using fields with partial or non-standard text (e.g., names, addresses, incident types, etc.). Watchman Analytics™ drives exploration into enterprise data stores, enabling drill-down and expanded analysis to identify critical information.
RAPID DATA INGESTION AND TRANSFORMATION
Watchman Analytics™ allows for the rapid import of data from multiple, diverse data sources into the high-performance analysis database. Sources can be available over the network or local to the user, as either a traditional DBMS or standard file (e.g., flat text, MS Excel). It connects to major proprietary databases (e.g., Microsoft SQL Server) as well as open source databases (e.g., MySQL). Once configured, it can continuously run in the background, applying established business logic rules, and user defined data refresh. This component can be used to mitigate up-front integration work, or to consolidate multi-source content in a single repository (e.g., a data warehouse – a common data store for analytical tasks).
AUTOMATED DATA CLEANSING AND VALIDATION
Watchman Analytics™ performs interactive data validation to deliver improved data quality, consistency and integrity. Users can apply checks to data based on existing established data structures to ensure that it follows specific syntactic or semantic guidelines, and captures the necessary information. These validation rules can be configured by the user and applied to existing or newly entered data.
This function algorithmically determines the likelihood that two entities are actually the same, even if the entity data has inadvertently or intentionally been mischaracterized. Data pertaining to an entity is often input in a variety of formats or labels, or is intentionally changed to hide associations. An example of the importance of entity resolution is in Law Enforcement or counter-terrorism, where criminal multiple aliases can obfuscate involvement in criminal activity. This component can be used for resolution of many types of key entities (persons, places, things, concepts, etc.) upon data capture, or using existing stored data.
SIMPLIFIED, INTUITIVE DATA VIEWS
Query results may be displayed in a traditional list view, where individual or multiple records may be selected, sorted or viewed in detail. Charting tools render query results in easily understood visual displays. These can be the native tools, or Watchman Analytics™ can export to third party mapping and visualization tools. These info graphs (scatterplots, histograms, pie-charts, etc.) are auto-generated with each query and enable rapid factor assessments, (e.g., trending over time), while eliminating need to copy results into spreadsheets to produce graphs.
The mapping option allows users to view the analytics results on a native map viewer (e.g., as points or ‘hot spots’), or view the information in a third party map viewer (e.g., Google or Bing Maps) or traditional GIS system (e.g., ESRI or Intergraph).
LINK ANALYSIS AND DATA CLUSTERING
Helps reveal hidden relationships and groupings within multiple data sets. For example, when incident records or textual narratives are analyzed by Watchman Analytics™, latent associations, groupings or clusters will be identified and visualized using Intuidex’s powerful HO-LRL™. The nature of these associations can then be further analyzed and interrogated using other analytical capabilities of the platform. For example, “all of these improvised explosive devices have a design that originated from a specific terrorist training facility”. Determining patterns and trends in textual narratives has been particularly useful by capturing latent relationships from text that would otherwise have been missed.
Watchman for Defense™ is an all-source intelligence application providing data fusion, extraction, analysis, monitoring (with alerting) and federated search capabilities that support situational awareness by offering multi-INT fusion, correlation, prediction, forecasting and pattern of life and anomaly detection.
The Natural Language Question Answering system improves decision making and efficiency by taking textual documents (reports, emails, etc.) and generating answers to questions you provide in report format, including source references and summary statements, utilizing advanced HO-LRL™ based analysis.
Watchman for Law Enforcement™ offers much more than just license plate reader technology. HO-LRL™ allows agencies to identify related associates, cohorts, convoys and many more patterns using Intuidex advanced algorithm technology across multiple data sources.
Watchman for License Plate Reader Predict™ is a license plate number and vehicle make, model, body, and color (MMBC) prediction solution. Built from the ground up, W4LPR Predict provides server-side and edge-based ALPR + MMBC with superior performance and accuracy.