In the out-of-the-box i2k Platform, users can select from a predetermined, hierarchical list of classes and taxonomies. Each taxonomy indicates a broad category, perspective, or “view” that is available to refine a search. In practice, the taxonomies and displays for users are often customized to match the specific needs of a community. The basic format is refined by working with customers, analyzing user behavior, and collecting experimental data on the effects of changes.
Much of the effort in delivering AI within a commercial system is devoted to non-AI issues such as integration into the existing corporate infrastructure, information security architecture, and workflows that reveal useful results. Because corporate repositories frequently contain tens of millions of documents, processing speed is an essential consideration in deployment.
Other out-of-the-box AI toolkits can immediately find existing tags and taxonomies. The i2k AI Platform goes beyond with machine learning, intensive tagging of content and data within documents, and the application of existing, incisive domain expertise.