jbiFusion™ is a Java-based, distributed framework of intelligent agents combined with a virtual knowledge system for singular access to any and all data, information, and knowledge required to satisfy business needs on a daily basis.
The intelligent agents can be deployed to collect and normalize data from a variety of relevant sources, including RSS feeds, e-mail servers, databases, and even Twitter (to name a few). Once data is aggregated and categorized from the multiple sources, correlated results can be visualized via our built-in geo-mapping components. These results can then be shared among users and external systems, presenting reliable and actionable information.
Our intelligent agents are an effective alternative to using people for handling information provided from intelligence communities, FAA, INS, local and federal law enforcement, coalition partners, and others. The jbiFusion™ system processes the data from these sources using customizable ontologies and rules sets, correlating possible critical events and thereby alerting the authorities. Additionally, a foreign language plug-in module provides extraction of common terms in a variety of languages, allowing analysis of foreign information and intelligence.
The end result is to automatically provide the answers to who, what, when, where and why. By answering these key questions, the system tremendously improves the rapid first response by getting the right information to the right people at the right time.
Click to Enlarge
Key Features
Data Acquisition
Our Intelligent Agent framework allows for the rapid deployment of custom data collectors to interface to a variety of data sources:
- E-Mail Accounts
- RSS/CAP Feeds
- SQL Databases
- Web Pages
- And more
Correlation
Intelligent agents correlate similar articles to eliminate redundancy.
Categorization
Articles are categorized based on topics and keywords. Rules can be created by users or generated automatically by the system based on user interests.
Ad Hoc Searches
Users can generate ad-hoc queries which are scored with a bias relevant to the topics defined for the currently selected collection.
Collections
Users create unlimited document collections by subscribing to topics.
User Feedback/Bias
The user can assert favorable and unfavorable biases to effect the scoring of alerts within a collection to prioritize alerts of greater interest.
Visualization
Integrated mapping of Geographically tagged articles provides geo-spatial context of relevant alerts. Temporal Analysis shows alert volume over time aggregated by alert topic.
Case Analysis
Users create cases to manage specific information gathering requirements. Articles can be added from collections or adhoc search results.
Entity Extraction
Entity extraction can be performed on saved articles within a case file in order to automate knowledge formation for answering the 5Ws (Who, What, When, Where, Why/ How).
Collaboration
Users can collaborate securely on cases and share information with selected users. The system supports PKI/CAC access as well as username/password authentication using an LDAP server for account management.
Architecture
The application is deployed as an enterprise web-application to support multiple concurrent users. It also employs a Service- Oriented Architecture (SOA) for interoperability with other systems. Our agent framework provides a publish/ subscribe messaging model which allows agents to be deployed as plug-ins to perform functions based on customer requirements.