CDMP-RMD practice question 34 of 35

Which of the following is NOT a possible outcome of a probabilistic matching algorithm?

Full set: 100 questions
Answer options
A: Likely Match
B: Non- match
C: None of answers are correct (
D: Underminable match Correct answer
E: Match
Correct Answer:
D. Underminable match
Explanation

Probabilistic matching algorithms calculate a similarity score between two records to determine if they represent the same real-world entity. This score is then compared against two pre-defined thresholds. The outcome falls into one of three categories:
1. Match: The score is above the upper threshold, indicating a high probability of a match.
2. Non-match: The score is below the lower threshold, indicating the records are distinct.
3. Likely Match (or Potential Match): The score falls between the two thresholds, indicating uncertainty that requires manual review by a data steward. The term 'Underminable match' is not a standard or recognized outcome in probabilistic matching theory or practice.

Why Incorrect

A. Likely Match: This is a standard outcome for records with scores between the match and non-match thresholds, which are flagged for manual review.
B. Non-match: This is a standard outcome for records with scores below the lower threshold, indicating they are considered distinct entities.
C. None of answers are correct: This is incorrect because 'Underminable match' is demonstrably not a possible outcome of a probabilistic matching algorithm.
E. Match: This is a standard outcome for records with scores above the upper threshold, indicating a high confidence that they are the same entity.

References

1. DAMA International. (2017). DAMA-DMBOK: Data Management Body of Knowledge (2nd ed.). DAMA International.

Reference: Chapter 11, Section 3.4.2, "Matching," p. 369. The text explicitly states, "The results of matching fall into three categories: matches, non-matches, and potential matches... Potential matches are records that fall between the two thresholds." This confirms the three valid outcomes.

2. Fellegi, I. P., & Sunter, A. B. (1969). A Theory for Record Linkage. Journal of the American Statistical Association, 64(328), 1183–1210.

Reference: Section 4, "The Decision Rule," p. 1188. This foundational paper defines the three decision regions: "designate the pair as a link" (Match), "designate the pair as a non-link" (Non-match), or "designate the pair as a possible link and hold for clerical review" (Likely Match).

3. Christen, P. (2012). Data Matching: Concepts and Techniques for Record Linkage, Entity Resolution, and Duplicate Detection. Springer.

Reference: Chapter 2, Section 2.4.2, "Classification," p. 28. The text states, "The outcome of the classification step is to assign each pair of records... into one of three classes: matches, non-matches, and potential matches."

Scroll to Top