Whether you are your organizations sole Watson user or one of a larger number of individuals using Watson within a team, sooner or later you will be faced with an ever-growing list of artifacts in your collection (perhaps referred to as Watson knowledge objects). In the most vivid example, organizations that have multiple individuals using Watson may even have a number of duplicate objects saved.
You may at some point find yourself:
- Sorting through a large Watson collection
- Interpreting misleading or conflicting names given to objects saved within a collection (more on this later in this chapter)
- Struggling to just find existing objects
- Recreating objects that have already been created or even shared
To that end, even a minimalist effort to organize and manage your organization's (or your own) collections can provide practical returns.
Organizational efforts include:
- Simple inventory and organization: Time should be taken to inventory and monitor your Watson collection to minimize the reinvention of objects that solve the same thing and promote useful general purpose objects that are created and shared on a global basis, across deployments.
- Standardized normalization and modeling of data: As mentioned earlier in this chapter, data files can (and most likely should) be explored and enriched by various means—internally (in Watson), or using an external tool or technology.
- Standardizing on Version: Organizations serious about using Watson should demand that all Watson users be on the same version—optimally the Professional version (more on this later in this chapter). In addition, within Watson, there is a provided versioning feature for your predictions (more on this later in this chapter as well).
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.