Note: This page is a work in progress. Use cases will be created or updated as the information is written.

We have defined some common use cases—some simpler, some more complex—to help you get started with FACTS. Over time, these use cases will form libraries distributed with FACTS, to be reused and adapted as required. The current common use cases are:

Prioritizing Contexts by Using Context Preclusion

You can prioritize which context or document fact should be returned if more than one is found.

This is useful in the following scenarios:

1. Prioritizing which version of the same fact, extracted by several different contexts, should be returned.
   This is typically called a fact fall-back strategy. For example, you may have fall-back contexts where you prefer the strongest context to be used, as it will most reliably find the correct fact. However, sometimes the content does not allow for such certain extraction, and you must fall back on less certain contexts. In such cases, you want to preclude the less certain contexts with the more certain ones.

2. Prioritizing which of several different facts should be returned if more than one is found.
   Preclusion can also be used to prioritize one fact over another.

Stopping Fact and Anchor False Positives by Using Precluding Label Evidence

• Stopping false positive evidence being used for facts and anchors using forced preclusion of labels

Stopping Facts from Being Returned by Blacklisting Overlapping Facts

• Barring a fact from being returned if it overlaps another fact
• Barring a fact's context from being returned if it overlaps another context

Returning Only Contexts Dependent on Other Contexts Firing

Extracting Entire Contexts Using Context Facts

• Extracting an entire sentence's or paragraph's text as a fact

Selecting One Fact from Many Using Fact and Context Indexing

• Extracting multiple similar facts from a single context
• Selecting which fact to return if more than one is found

Extracting Zones and Entities Using Facts

• Stopping a fact being returned if it overlaps another in the content
• Extracting entities
• Anchoring and extracting text that matches parts of speech