Reputation Algorithms

Content Contributor Reputation Model

Metrics:

1. Total Number of Articles Published (N): The cumulative count of articles a contributor has published in each Olas Information Market.

2. Sum of all Outcome Scores given to Articles Written (AS): The total of all outcome scores received for all the articles written by the contributor across all Information Markets.

Reputation Algorithm:

The Reputation Score (RS) for a Content Contributor is calculated using the following formula:

$\boxed{ RS \, = \, \,\dfrac{ AS }{ N } }$

Fact-Checker Reputation Model

Metrics:

1. Total Number of Valid Issues Raised (V): Cumulative count of valid issues identified by a fact-checker.

2. Number of Valid Issues Submitted in Each Category of Severity (S): Number of valid issues submitted, categorized by severity.

   • Low Severity (LS): Minor Issues that have less impact or are easily rectifiable.

     • Multiplier Factor (M1): 1.2

   • Medium Severity (MS): Issues that have a moderate impact or may require a moderate effort to rectify.

     • Multiplier Factor (M2): 1.5

   • High Severity (HS): Significant Issues that have a high impact or may require substantial effort to rectify.

     • Multiplier Factor (M3): 2

3. Number of Unique & Valid Issues Submitted (UV): Number of issues submitted that are both unique & valid.

   • Unique & Valid (UV): Uniqueness is defined as an issue that was identified and submitted by only one Fact-Checker.

     • Multiplier Factor (M4): 2

Reputation Algorithm:

The Reputation Score (RS) for a Fact-Checker is derived using a weighted sum formula. Each metric is multiplied by a corresponding weight, and the sum of these products forms the RS. The formula is articulated as follows:

$\boxed{ Preliminary \, Score \, = \, V \, + \, ( \,LS \, \times \, M1 \,) \, + \, ( \, MS \, \times \, M2 \, ) \, + \, ( \, HS \, \times \, M3 \, ) \,+ \, ( \, UV \, \times \, M4 \, ) }$

$\boxed{ Reputation \, Score \, = \, \,\dfrac{ Preliminary \, Score }{ Total \, Number \, Of \, Issues \, Submitted } }$

where:

• $M1​, M2​, M3​,$ and $M4$ are the weighting factors assigned to each metric based on its relative importance.

Judge Reputation Model

Metrics:

1. Total Number of Judging Panels Completed (JP): The cumulative count of judging panels a judge has completed.

2. Sum of all Voting Accuracy Scores Across All Votes Submitted in Judging Panels (VA): The total of accuracy scores received for all votes submitted by the judge across various judging panels.

Reputation Algorithm:

The Reputation Score (RS) for a Judge is computed using the following formula:

$\boxed{ RS \, = \, \,\dfrac{ VA }{ JP } }$