We’re Building A Trustworthy Knowledge Database System
Try this, Gunther
Broadly speaking our approach has the following elements —
- Database starts by defining what constitutes “ideal” content within each topic area. Using pro research librarians, there area numerous criteria that are rated on a scale for importance to each topic, thus the ideal varies for each topic (patent approved).
- Multiple experts within each topic area identified are vetted for any real or perceived conflict of interest.
- Multiple experts within each topic area identify top 5–8 superior providers of content in their area of expertise and rate each provider on the same criteria list used for an ideal without knowing what is ideal.
- Algorithm compares expert ratings for each content provider against ideal ratings to generate numerical percentage of adequacy in reaching the ideal, and then rank lists all the superior providers in that topic.
- Experts provide a curated description of each content provider’s strengths and weaknesses.
- Users/consumers can see all expert biographies and actual ratings for each criteria on each content provider offered and, if desired, submit a different rating for any given criterion. If the disparity from a given number of users is wide enough the expert’s recommendation and ratings are reviewed.
- No expert or user can access or alter public depiction of results
- Experts update their list of superior providers as warranted and queried at a preset time period. Users can recommend superior content providers not listed and are paid a bounty if review determines recommendation is superior.
- Superior content providers can be displayed independently, overlaid adjacent to search results, accessed by virtual assistants and augmented reality systems (patent approved).
In essence it’s a top-down filtering process that reduces knowledge asymmetry, saves time and effort, reducing risk for both services and products without user’s needing to modify their behavior.