i4View: Visually-oriented Search and Content Modeling
As organizations evolve to working with larger and larger data sets, the amount of time wasted searching and exploring there data rises dramatically. Being able to quickly and efficiently discover what is hidden within your content (text, html, pdf, word, etc) is critical. While most content is textually oriented, the fastest means of comprehend it is visual. The average person can only read 3.5 words per second, yet the human eye is capable of transmitting 32 mb per second of pictorial information. i4View lets your content to talk to you visually, quickly and pictorially surfacing patterns, outliers, inter-dependencies, and more - all in a visually oriented search experience.
Semantic Modeling takes a probabilistic approach to automatically uncover the semantics of a domain - a critical tool in analyzing terminology. Our implementation offers the ability to not only discover synonyms and antonyms, but enables using concepts as mathematical units to visually discover related concepts in a high dimension semantic space. The resulting semantic model can be leveraged to quickly develop precise entity extraction or an empirical taxonomy and strongly complements Thematic Modeling.
Topic Modeling automatically uncovers thematic structure (topics), a critical tool in analyzing large volumes of diverse unlabeled and unstructured text. We offer an assortment of statistically-based modeling algorithms, from the classical to the recently published, each married to a set of rich interactive visualizations for discovering the abstract "topics" or thematic clusters within content as well as their inter-dependencies. We also offer this a primer technique for domain-specific language development.