"The next wave in classifying and indexing paper based information with minimum user-intervention". CMP’s Imaging & Document Solutions Magazine awards 'Best of AIIM 2000" to SERbrainware.

Content Analysis with Human Like Intelligence

SERbrainware is a powerful software engine that performs content analysis using advanced neural network and associative search techniques. It forms the basis of a new generation of software programs, which read, analyze and act on text, with human-like intelligence. SERbrainware analyses text from any media type, including email, web sites, faxes, paper documents or even from human speech, and routes the information to the relevant user.

SERbrainware uses learning and memory to derive meaning from text. It learns the inherent patterns of all documents of a subject and saves the resulting extract for the future identification of documents, which contain information about the same subject. It automates the process of categorization and linking text, whatever the media type, removing the need for costly manual methods.

SERbrainware also acts as a vital e-commerce tool to analyze and categorize unstructured information. It cost effectively brings paper documents online without the need or associated cost of manually categorizing and subsequently reduces the cost of providing information online.  It analyses the content of information received and routes relevant documents or queries to customer service representatives with suitable skills. Alternatively, it can analyse text and generate an automatic answer, delivering fast response, high customer satisfaction and improved productivity.

The ‘brain’ in SERbrainware

SERbrainware uses both artificial neural network and associative search techniques to automatically classify and search unstructured information. These two complementary approaches are used transparently within the software engine.

The systems method of learning is quite similar to the way humans learn: it is purely sample-based. This learning process is controlled by the human and typically requires not more than a few mouse clicks. The business expert (not the IT specialist) selects the text samples and checks the results of the learning process. After the learning process is completed, SERbrainware contains an electronic version of the expert’s relevant knowledge within the text samples or ‘learn sets’.  On a single PC, thousands of documents can be categorized within minutes.

SERbrainware has a fully integrated associative search capability  - content and context searching  - built into its core to enhance the quality and precision of classification and retrieval. This also has the capability of providing a resilient storage of documents in a format capable of rapid retrieval.  SERbrainware performs associative searches with incredible speed and accuracy, searching several hundreds of Megabytes of text in less than a second – using both exact and fuzzy matching.