CDMP-RMD practice question 10 of 35

Matching or candidate identification is the process called similarity analysis. One approach is called deterministic which relies on:

Full set: 100 questions
Answer options
A: Statistical techniques for assessing the probability that any pair of records represents the same entity
B: Taking data samples and looking at results for a subset of the records
C: Being able to determine the similarity between two data models
D: Algorithms for parsing and standardization and on defined patterns and rules for determining similarity Correct answer
E: Finding two references that are linked with a single entity
Correct Answer:
D. Algorithms for parsing and standardization and on defined patterns and rules for determining similarity
Explanation

Deterministic matching, also known as rule-based matching, is an approach to entity resolution that relies on a predefined set of exact rules. This process involves first standardizing and parsing data elements (e.g., names, addresses). Then, specific algorithms apply defined patterns and rules to determine if records match. For instance, a rule might state that two records represent the same entity if they have an exact match on Social Security Number and Last Name. The outcome is binary (match or no-match) and does not involve probabilities.

Why Incorrect

A. This describes probabilistic matching, which calculates the likelihood that two records refer to the same entity using statistical weights (e.g., the Fellegi-Sunter model).
B. This describes sampling, a technique used for data profiling or quality assessment, not the core logic of a matching methodology.
C. This refers to schema matching or data model integration, which operates at the metadata level, not on individual instance-level data records.
E. This describes the overall goal of entity resolution (linking references to an entity), not the specific technique of how deterministic matching achieves it.

References

1. DAMA International. (2017). DAMA-DMBOK: Data Management Body of Knowledge (2nd ed.). DAMA International. Chapter 11, Section 4.3.2.2, "Matching," p. 369. The text states, "Deterministic matching uses pre-configured rules to match records... For example, a rule might be defined to create a match key based on exact matching on a set of selected attributes."

2. Christen, P. (2012). Data Matching: Concepts and Techniques for Record Linkage, Entity Resolution, and Duplicate Detection. Springer. Chapter 2, "Deterministic Record Linkage," pp. 15-16. This chapter defines deterministic record linkage as a process where "a set of rules is defined that determines when a pair of records is classified as a match, a non-match, or a potential match... These rules are often based on the exact agreement of one or more attributes."

3. Herzog, T. N., Scheuren, F. J., & Winkler, W. E. (2007). Data Quality and Record Linkage Techniques. Springer. Chapter 5, "Record Linkage," p. 70. The text contrasts the two main approaches: "The first is usually called deterministic record linkage... The second approach is called probabilistic record linkage." It describes the deterministic approach as relying on defined rules for agreement on identifying fields.

Scroll to Top