Summary

Fully integrated framework for building and deploying service-based networks (SOA / Microservices). Communication includes XML-RPC, REST and Web Services. Other features include service protection, admin scripts, metadata, query processes, autonomous, agent-based, or self-organising.

Special to the system are its AI and text-processing features. In particular, there is an Autonomic Manager framework with a BPEL-style execution script. Sub-packages include text or query processing and a problem solver provides a range of algorithms. These can also be used separately, but advanced programming and configuration is required.

A free All-in-One GUI allows you to manage services and function, and add your own modules. Example Apps are included and the images illustrate the functionality. The open source p2p server requires only the HTTP address and optional passwords. Small jar file sizes and mobile compatible.

Features

• The capability to build distributed networks of (autonomous) service-based components.
• Local or remote communications, including XML-RPC, REST, SOAP or HTTP request.
• Message Bus communication system, with synchronous or asynchronous calls.
• All-in-one GUI for viewing or testing your networks. Not open source, but provided free for personal use, with a set of free services.
• Framework for adding an Autonomic Manager and policy scripts to a service.
• Framework for adding metadata, with default query and script execution engines.
• Permanent and dynamic linking mechanisms, to construct network architectures.
• Service wrapper classes allowing legacy code to be loaded.
• Problem-solving framework, allowing for the addition of more complex heuristic search processes.
• Android compatible.

Current Release

licas, ai_heuristic and licas_ai packages updated.

LearningData renamed Dataset, with new categories list.

LearningData interface requires parameter to indicate if category included or not.

Data rows changed to LinkedHashMap.

NearestNeighbour re-written to not use a KD Tree. Now called KNearestNeighbour.

SOM algorithm updated to use base class of nearest neighbour.

Some Result classes moved to ai-heuristic package. You may have to change the reference. Result, ClusterResult, etc.

New Parsers for each linking type - Dyamic, permanent and associated. Use these to get XML descriptions. No LinksParser now.