The HyLighter Idea
HyLighter ESP combines an advanced social-sensemaking engine with various applications (e.g., data acquisition, text analytics, visual analytics, image pattern matching, audio transcription, and machine translation). Machines help teams turn large volumes of mostly unstructured data into actionable knowledge. Teams send feedback to machines from the social-sensemaking layer to enable machine learning and improve the capacity of machines to support sensemaking activities. The approach is scalable for virtually any number of users and any volume and variety of data. In sum, HyLighter ESP increases the ROI on machine analytics across the organizational ecosystem by making machine intelligence compatible with human sensemaking activities.
While people struggle with the effects of information overload and silo mentality, many organizations have increased their use of machine intelligence to tap into growing volumes of largely unstructured data. In most cases, organizations employ smart machines as autonomous systems that process information with a minimum of human input and require data scientists to turn the results into actionable knowledge.
The lack of simple tools that enable all members of the organizational ecosystem to interact with smart machines is a major barrier to delivering on the anticipated ROI of machine analytics for Big Data and smaller data sets too.
The HyLighter Solution
- Smart Machines: Acquire and process volumes of largely unstructured data using various applications for data acquisition, audio transcription, multilingual text analytics, visual analytics, machine translation, visual search, and image pattern matching.
- Machine-Generated HyLights: HyLighter ingests the processed files and converts machine analytics into Machine HyLights (i.e., a layer of meta-content including color-coded fragments and related comments, replies, tags, and links).
- Human-Generated HyLights: Users add HyLights and and create links between related HyLights.
- Reports, Mashups and Visualizations: Users control a powerful interactive workspace for filtering and assembling HyLights with links back to their exact locations in their sources (e.g., to maintain provenance).
- Feedback Loops: Users send selected HyLights back to smart machines to support social-sensemaking activities and accelerate machine and human learning.
- Preserved Record of Thinking: The system provides secure access to the historical record of thinking across the organizational ecosystem.