I85PREP
⌘ K
Informatica Data Engineering Foundation study materials
Informatica Data Engineering Foundation study resources

Informatica Data Engineering Foundation practice questions

One question at a time. Build your Informatica Data Engineering Foundation understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Data Integration

In an Informatica PowerCenter mapping, which transformation is used to route rows to different output groups based on conditions evaluated in order, where the first true condition wins?

  1. Aggregator
  2. Filter
  3. Joiner
  4. Router Correct answer.

Explanation

The Router transformation evaluates each group's condition in order and sends a row to the first group whose condition evaluates true, and can also send unmatched rows to a default group. A Filter transformation only passes or drops rows against a single condition and cannot split a stream into multiple outputs.

Question 2Data Integration

A developer needs to generate a sequence of surrogate key values inside a mapping without reading from a source table. Which transformation provides this capability?

  1. Expression
  2. Normalizer
  3. Sequence Generator Correct answer.
  4. Lookup

Explanation

The Sequence Generator transformation produces sequential numeric values using settings such as Start Value, Increment By, and End Value, and it can be reused across mappings. It is the standard way to create surrogate keys without an external sequence object.

Question 3Data Integration

Which PowerCenter object stores the reusable connection and source/target definitions that a mapping references, and is the container a mapping is saved into?

  1. Folder within a repository Correct answer.
  2. Session
  3. Mapplet
  4. Mapping

Explanation

A mapping is saved into a folder inside the repository, and that folder organizes the mappings, mapplets, and other metadata for a project. The repository folder is the organizational container, while sessions and workflows are scheduling objects stored alongside them.

Question 4Data Integration

In an Informatica PowerCenter workflow, which task type actually executes a mapping against sources and targets?

  1. Assignment task
  2. Decision task
  3. Timer task
  4. Session task Correct answer.

Explanation

A Session task binds a mapping to a workflow and defines how that mapping runs, including source and target connections and load settings. Other task types control workflow logic such as branching, waiting, or setting variables but do not execute the mapping itself.

Question 5Data Integration

Which transformation should be used when a developer needs to combine two source pipelines in the same mapping based on a common key, without going to the database?

  1. Aggregator
  2. Union
  3. Joiner Correct answer.
  4. Sorter

Explanation

The Joiner transformation performs a join between two pipelines inside the mapping using a join condition and a join type such as inner, master outer, or detail outer. Union appends rows without matching keys, so it does not satisfy a key-based combine requirement.

Question 6Data Integration

A mapping must reject source rows where the CUSTOMER_ID is null and write them to a separate error table. Which transformation configuration is the most direct approach?

  1. Use a Sequence Generator to replace the nulls.
  2. Use a Filter transformation and route rejected rows to a separate target. Correct answer.
  3. Use an Update Strategy to mark null rows for deletion.
  4. Use an Aggregator to count null rows.

Explanation

A Filter transformation with a condition such as NOT ISNULL(CUSTOMER_ID) passes valid rows and drops the rest; adding a second output target for the rejected stream captures the bad records. This keeps valid and invalid data separated without altering the business data.

Question 7Data Integration

What is the primary purpose of a reusable transformation in PowerCenter?

  1. To encrypt source data in transit.
  2. To allow the same transformation logic to be reused across multiple mappings. Correct answer.
  3. To schedule workflows at defined intervals.
  4. To automatically generate target DDL.

Explanation

A reusable transformation is stored once in the repository and can be referenced by multiple mappings, which keeps logic consistent and reduces maintenance. Non-reusable transformations are local to a single mapping and cannot be shared.

Question 8Data Integration

In an Informatica PowerCenter session, what does the 'Treat source rows as' setting control?

  1. The number of partitions used by the Integration Service.
  2. The commit interval used during loading.
  3. The row operation (insert, update, delete, or data-driven) applied to target rows. Correct answer.
  4. The error threshold before the session fails.

Explanation

The 'Treat source rows as' property determines whether rows are processed as inserts, updates, deletes, or data-driven operations, and it interacts with the Update Strategy transformation. It governs the operation applied to each row rather than performance or error handling.

Question 9Data Integration

A mapping joins a large fact pipeline to a small dimension using a Lookup transformation. To reduce repeated database calls, which lookup option should be enabled?

  1. Dynamic lookup with insert
  2. Unconnected lookup
  3. Lookup caching Correct answer.
  4. Sorted input

Explanation

Enabling lookup caching loads the lookup condition values into memory so the Integration Service can resolve matches without querying the database for every row. This is the standard optimization when the lookup table is small relative to the fact data.

Question 10Data Integration

Which statement correctly describes the difference between a connected and an unconnected Lookup transformation?

  1. A connected lookup cannot be reused in multiple mappings.
  2. A connected lookup can only use static caches, while an unconnected lookup must use dynamic caches.
  3. A connected lookup receives input from a pipeline, while an unconnected lookup is called from an expression using :LKP. Correct answer.
  4. A connected lookup returns all matching rows, while an unconnected lookup returns multiple rows.

Explanation

A connected lookup is part of the data flow and receives its input ports directly from upstream transformations, returning one value per row. An unconnected lookup is not in the pipeline and is invoked explicitly from an Expression transformation using the :LKP reference.

Question 11Data Integration

A developer wants the Aggregator transformation to process data more efficiently by avoiding a sort step. Which session or transformation setting supports this?

  1. Set the Aggregator to 'Unsorted Input' and disable caching.
  2. Use a Filter transformation before the Aggregator.
  3. Increase the commit interval in the session.
  4. Enable 'Sorted Input' on the Aggregator and ensure upstream data is sorted on the group-by ports. Correct answer.

Explanation

When the incoming data is already sorted on the group-by ports, enabling Sorted Input lets the Aggregator process groups sequentially without building a full sort, improving performance. The upstream pipeline must genuinely deliver sorted data for this to be valid.

Question 12Data Integration

In a PowerCenter workflow, a Decision task evaluates a workflow variable. What happens if the condition is false and no false branch is defined?

  1. The session is marked as skipped.
  2. The workflow continues along the default path.
  3. The Integration Service retries the task.
  4. The workflow fails immediately. Correct answer.

Explanation

A Decision task requires both true and false conditions to be defined; if the evaluated condition is false and no false condition is configured, the task fails and the workflow errors out. The workflow does not silently continue or retry.

Question 13Data Integration

A mapping uses a Router transformation with three groups. A row matches none of the group conditions. What happens to that row?

  1. It is written to the first group by default.
  2. It is duplicated to all groups.
  3. It is written to the default group if the default group is enabled. Correct answer.
  4. It causes the session to fail.

Explanation

The Router transformation has an optional default group that receives rows matching none of the defined conditions. If the default group is not enabled, unmatched rows are discarded.

Question 14Data Integration

Which approach best implements slowly changing dimension Type 2 in PowerCenter?

  1. Insert a new versioned row and expire the previous row using an Update Strategy and a Lookup. Correct answer.
  2. Overwrite the dimension row with the new attribute values.
  3. Store only the latest value in a separate history table.
  4. Delete the old row and insert the new one.

Explanation

Type 2 preserves history by inserting a new row for each change and marking the prior row as expired, typically tracked with effective dates or a current-flag. An Update Strategy plus Lookup is the standard PowerCenter pattern for detecting changes and applying the correct insert/update operations.

Question 15Data Integration

What is the role of the Workflow Manager's 'Link' feature between tasks?

  1. It enables partitioning across nodes.
  2. It defines the order in which tasks execute and the conditions under which they run. Correct answer.
  3. It maps source columns to target columns.
  4. It creates a database connection for sessions.

Explanation

Links connect tasks in a workflow and determine execution order, including conditional links that run a task only if a prior task succeeds or a condition is satisfied. They are the control-flow mechanism of the workflow.

Question 16Data Integration

A developer needs to convert a single source row containing repeating groups into multiple rows. Which transformation is appropriate?

  1. Rank
  2. Aggregator
  3. Normalizer Correct answer.
  4. Expression

Explanation

The Normalizer transformation flattens repeating groups in a source row into multiple rows, generating an occurrence identifier for each. It is used when source data is denormalized and needs to be normalized for a relational target.

Question 17Data Integration

In PowerCenter, which session property determines how many rows are processed between commits to the target?

  1. Buffer Block Size
  2. Commit Interval Correct answer.
  3. Recovery Strategy
  4. Error Threshold

Explanation

Commit Interval sets the number of rows processed before the Integration Service commits the transaction to the target. It balances recovery granularity against commit overhead.

Question 18Data Integration

A developer wants to reject rows where the target update fails due to a constraint violation without stopping the entire session. Which feature supports this?

  1. Set the session to fail on the first error.
  2. Enable 'Treat source rows as delete'.
  3. Use a Sequence Generator to renumber keys.
  4. Configure a reject file and set the error threshold appropriately. Correct answer.

Explanation

PowerCenter can write failed rows to a reject file while continuing, provided the error threshold allows the session to proceed. This lets a load complete while capturing problematic rows for later review.

Question 19Data Integration

Which statement accurately describes the difference between a mapplet and a reusable transformation?

  1. A mapplet is a single transformation, while a reusable transformation is a set of transformations.
  2. A mapplet is a reusable set of transformations, while a reusable transformation is a single transformation reused across mappings. Correct answer.
  3. A mapplet can only be used in one mapping, while a reusable transformation can be used in many.
  4. A mapplet executes as a standalone workflow, while a reusable transformation does not.

Explanation

A mapplet packages a group of transformations and can be reused as a unit in multiple mappings, whereas a reusable transformation is a single transformation object shared across mappings. Both promote reuse but differ in scope.

Question 20Data Integration

A developer needs to implement a lookup against a flat file that changes during the session and must reflect newly inserted rows. Which lookup configuration is required?

  1. Dynamic cache with insert enabled Correct answer.
  2. Lookup with sorted input
  3. Static cache with a persistent cache file
  4. Unconnected lookup without cache

Explanation

A dynamic lookup cache with insert enabled allows the lookup to detect new or changed rows and update the cache during the session, so subsequent rows see the inserted values. A static cache is built once and does not reflect in-session changes.

Question 21Data Integration

In a PowerCenter mapping, a Joiner transformation is used with a master outer join. Which statement about the result is correct?

  1. All rows from the detail pipeline are returned, plus matching rows from the master pipeline.
  2. Only rows matching in both pipelines are returned.
  3. All rows from the master pipeline are returned, plus matching rows from the detail pipeline. Correct answer.
  4. Only unmatched rows from the master pipeline are returned.

Explanation

A master outer join returns every row from the master pipeline and includes matching detail rows where they exist, leaving detail columns null when there is no match. This differs from a detail outer join, which preserves all detail rows.

Question 22Data Integration

A developer must ensure that a session processes rows in a deterministic order for a Rank transformation to produce correct results. What is the most reliable approach?

  1. Increase the commit interval.
  2. Sort the input on the rank ports before the Rank transformation. Correct answer.
  3. Enable partitioning on the Rank transformation.
  4. Rely on the source's natural order without sorting.

Explanation

Rank depends on the ordering of input rows to assign ranks consistently, so the pipeline must sort on the rank key before ranking. Without a defined sort, results can vary between runs and across partitions.

Question 23Data Integration

A mapping uses a connected Lookup with a dynamic cache and 'Insert Else Update' enabled. A row arrives whose lookup condition matches an existing cached row. What operation is performed on the lookup table?

  1. A new row is inserted.
  2. The matching row is updated with the incoming values. Correct answer.
  3. No operation is performed because the row already exists.
  4. The matching row is deleted.

Explanation

With Insert Else Update, a matching row triggers an update using the incoming values, while a non-matching row triggers an insert. This keeps the lookup table synchronized with the incoming stream.

Question 24Data Integration

Which scenario best justifies using a persistent lookup cache in PowerCenter?

  1. The lookup table is large and static across multiple session runs. Correct answer.
  2. The lookup table changes frequently during each session.
  3. The mapping runs only once and the lookup table is tiny.
  4. The lookup must reflect inserts made earlier in the same session.

Explanation

A persistent cache is saved to disk and reused across session runs, which benefits large, relatively static lookup tables by avoiding a full rebuild each time. If the table changes frequently, the cache must be refreshed, reducing the benefit.

Question 25Data Integration

A workflow must run a session only if a prior session completed successfully, and must send an email if it failed. Which combination of workflow features achieves this?

  1. Two independent sessions started in parallel.
  2. A Timer task before the first session plus a Decision task.
  3. A conditional link from the first session plus an email task on the failure path. Correct answer.
  4. An Assignment task plus a Command task.

Explanation

Conditional links let the workflow branch based on the outcome of the preceding task, so the second session runs only on success and an email task can be triggered on the failure path. This models dependency and notification within the workflow.

Question 26Data Integration

In PowerCenter, which transformation is used to generate a sequence of numeric values that can be assigned to a surrogate key column in the target?

  1. Aggregator transformation
  2. Sorter transformation
  3. Expression transformation
  4. Sequence Generator transformation Correct answer.

Explanation

The Sequence Generator produces monotonically increasing numeric values at run time and exposes them through its NEXTVAL and CURRVAL output ports. It is the standard mechanism for assigning surrogate keys during a session, with options for start value, increment, and end value.

Question 27Data Integration

A developer needs to combine two source pipelines into a single output where every row from the master pipeline is retained even when no match exists in the detail pipeline. Which configuration accomplishes this?

  1. Joiner transformation with a detail outer join
  2. Union transformation with duplicate removal enabled
  3. Joiner transformation with a master outer join Correct answer.
  4. Lookup transformation with the default inner join behavior

Explanation

In a Joiner transformation the master outer join keeps all master rows and fills detail ports with nulls where no match is found. Choosing which pipeline is master therefore determines which side is preserved.

Question 28Data Integration

Which PowerCenter object stores reusable metadata logic that can be dropped into multiple mappings and can itself contain multiple transformations?

  1. Session configuration template
  2. Reusable transformation
  3. Mapplet Correct answer.
  4. Worklet

Explanation

A mapplet is a reusable object that groups a set of transformations and ports and can be inserted into many mappings. It can include input, output, and other transformations, making it broader than a single reusable transformation.

Question 29Data Integration

A source table contains a single column holding comma-separated values that must be split into separate output rows. Which transformation is designed for this?

  1. Router transformation
  2. Update Strategy transformation
  3. Normalizer transformation Correct answer.
  4. Rank transformation

Explanation

The Normalizer transformation converts repeating groups, including delimited occurrences in one row, into multiple output rows. It is the standard choice for horizontal-to-vertical row expansion.

Question 30Data Integration

A mapping reads from an Oracle source and writes to a partitioned target table. The developer sets the session's target load type to bulk and enables partitioning on the target. Which consideration is most important?

  1. Bulk loading automatically disables all target partitioning and forces a single loader process
  2. The session must be run in a workflow with the 'Recover as restartable' option disabled
  3. The Aggregator transformation must be removed because it is incompatible with partitioned targets
  4. Each partition must write to a distinct target partition, so the partition key in the session must align with the target table's partition scheme Correct answer.

Explanation

When the session writes to a partitioned target, the partition type and key defined in the session must correspond to how the target table is physically partitioned. Mismatched keys cause rows to be routed incorrectly or loading to fail.

Question 31Data Integration

A developer wants a session to stop and mark itself failed if more than 5 percent of source rows are rejected by a downstream transformation. Which mechanism best enforces this?

  1. Define a session-level error threshold using the 'Error threshold' option for the session Correct answer.
  2. Add a Router transformation that discards rejected rows silently
  3. Configure a workflow Decision task that queries the session log after completion
  4. Set the session's 'Stop on errors' property to a fixed number of errors

Explanation

The session error threshold lets the Integration Service compare the number or percentage of rejected rows against a configured limit and fail the session when exceeded. It evaluates rejection counts during the run rather than after the fact.

Question 32Data Integration

A mapping must detect new source rows and update existing target rows in a single pass, using the target table itself as the reference. Which transformation supports this behavior natively?

  1. Connected Lookup transformation with dynamic cache and Update Dynamic Cache enabled Correct answer.
  2. Expression transformation that compares source and target ports
  3. Filter transformation placed before the target
  4. Static Lookup transformation with the default cache settings

Explanation

A dynamic lookup cache lets the Integration Service insert new cache entries and update existing ones as rows flow through, so subsequent rows see the changes. Combined with Update Strategy logic, this supports insert-or-update processing against the target.

Question 33Data Integration

A developer needs to compute a running total of transaction amounts per customer, ordered by transaction date, within a single mapping. Which transformation and configuration is appropriate?

  1. Rank transformation with the amount port as rank port
  2. Expression transformation using a variable port that accumulates the amount and resets on customer change Correct answer.
  3. Aggregator transformation grouped by customer with the amount port summed
  4. Sorter transformation on customer and date

Explanation

A running total requires row-by-row accumulation, which an Expression transformation achieves with a variable port that adds the current amount to the prior value. Resetting the variable when the customer changes yields a per-customer running total.

Question 34Data Integration

A developer wants to prevent a session from reading the same source rows twice if it is restarted after a failure. Which session property is most directly relevant?

  1. Recovery strategy Correct answer.
  2. Session sort order
  3. Source-based commit
  4. Commit interval

Explanation

The recovery strategy determines how the Integration Service resumes a failed session, including whether it uses stored recovery state to avoid reprocessing committed rows. Configuring it correctly is what makes restart behavior safe.

Question 35Data Integration

A mapping uses an unconnected Lookup transformation. How is it invoked within the mapping?

  1. By placing it directly in the data flow between source and target
  2. By calling it from an Expression transformation using the :LKP reference qualifier Correct answer.
  3. By linking it to the target definition with a dynamic port
  4. By adding it to a workflow as a separate task

Explanation

An unconnected Lookup is not part of the main pipeline; it is called explicitly from an Expression transformation through the :LKP reference qualifier, passing the lookup condition ports as arguments. The returned value is then available to the expression.

Question 36Data Integration

A developer must apply the same business rule to reject invalid email formats across ten mappings. Which approach best supports centralized maintenance?

  1. Add the rule to the target table as a check constraint
  2. Create a reusable transformation and reference it in each mapping Correct answer.
  3. Copy the same Expression transformation into each of the ten mappings
  4. Create ten separate sessions that each validate email format

Explanation

A reusable transformation is defined once and shared, so a change to the validation logic propagates to every mapping that uses it. This avoids divergence between copies of the same rule.

Question 37Data Integration

A mapping must load a dimension table using Type 2 history, keeping prior versions with effective dates and a current-flag indicator. Which design element is essential?

  1. A Filter transformation that discards rows whose attributes differ from the target
  2. An Aggregator that groups the dimension by natural key
  3. A Sequence Generator that overwrites the existing surrogate key on every run
  4. A Lookup on the natural key plus logic that expires the current row and inserts a new version when attributes change Correct answer.

Explanation

Type 2 history requires comparing incoming attributes with the current version, then closing the old version and inserting a new row with a new surrogate key and effective dates. A lookup on the natural key supplies the comparison.

Question 38Data Integration

A mapping contains a Router transformation with two groups and the default group disabled. A row satisfies neither group condition. What is the outcome for that row?

  1. The row is dropped from the Router output Correct answer.
  2. The row is passed to the target unchanged
  3. The row is written to the first group by default
  4. The row causes the session to fail

Explanation

When the default group is disabled, rows that match no group condition are discarded by the Router transformation. Enabling the default group is what captures such rows for further handling.

Question 39Data Integration

A developer observes that a session with an Aggregator transformation produces different group totals on repeated runs. Which cause is most likely?

  1. The session commit interval is set too high
  2. The target table has an index on the group-by column
  3. The incoming data is not sorted on the group-by ports and the session's sorted input assumption is incorrect Correct answer.
  4. The Aggregator's group-by ports are sorted, so results are stable and the difference is unrelated

Explanation

The Aggregator relies on incoming rows being grouped together for correct results when sorted input is assumed. If the data is not sorted on the group-by ports, the Integration Service may treat the same key as multiple groups across the stream, producing inconsistent totals.

Question 40Data Integration

A workflow must run a session and, only if it succeeds, trigger a shell script that archives the output; if the session fails, the script must not run. Which workflow construct achieves this?

  1. Set the command task to run unconditionally after the session
  2. Add both tasks to the same worklet without links
  3. Link the session to the command task with a conditional link whose condition tests the session status Correct answer.
  4. Place the session and the command task in parallel branches from a Start task

Explanation

A conditional link from the session to the command task evaluates the predecessor's status, so the command runs only when the session succeeded. This is the standard way to express success-dependent sequencing in a workflow.

Question 41Data Quality & Observability

In an Informatica Data Quality workflow, which object defines reusable business logic that can be mapped to multiple columns and reused across mappings?

  1. A worklet
  2. A rule specification Correct answer.
  3. A session
  4. A mapplet

Explanation

A rule specification is the reusable DQ asset that expresses business logic (for example a completeness or pattern rule) and can be bound to different input columns in different mappings. Mapplets and worklets are transformation-reuse constructs, not declarative DQ rule definitions.

Question 42Data Quality & Observability

Which score expresses the percentage of records in a column that contain a non-null, non-blank value?

  1. Timeliness
  2. Completeness Correct answer.
  3. Validity
  4. Uniqueness

Explanation

Completeness measures whether expected values are present, so it counts records with populated values against the total. Uniqueness measures duplicate suppression, validity measures conformance to a rule or pattern, and timeliness measures currency of the data.

Question 43Data Quality & Observability

A data steward needs to flag rows where the postal code does not match the national pattern. Which Informatica DQ rule type is most appropriate?

  1. A pattern-based validation rule Correct answer.
  2. An address reference match
  3. A duplicate record rule
  4. A data domain discovery rule

Explanation

Pattern-based validation checks each value against a defined character pattern such as a numeric or alphanumeric format, which is exactly what a postal-code format check requires. Duplicate and matching rules compare records to each other, and discovery profiles data rather than enforcing a format.

Question 44Data Quality & Observability

What is the primary purpose of a column profile in Informatica Data Quality?

  1. To summarize value distributions, null rates, and patterns in source data Correct answer.
  2. To encrypt sensitive columns at rest
  3. To generate the target DDL for the warehouse
  4. To schedule the workflow at a fixed interval

Explanation

Profiling analyzes source data and reports statistics such as distinct counts, null percentages, and inferred patterns so analysts can understand data before writing rules. It does not create DDL, schedule workflows, or perform encryption.

Question 45Data Quality & Observability

Which exception management action records a rejected record together with the rule it violated so a steward can review it later?

  1. Writing the record to an exception table with rule metadata Correct answer.
  2. Dropping the record silently
  3. Promoting the record to production
  4. Aggregating the record into a count metric

Explanation

Exception management captures failing records with context such as the violated rule and severity so they can be triaged and remediated. Silently dropping them loses the audit trail, and a count metric alone cannot support record-level review.

Question 46Data Quality & Observability

A mapping must send rows that fail a mandatory-field check to a reject path while passing valid rows downstream. Which transformation pattern accomplishes this?

  1. A Router transformation with two output groups Correct answer.
  2. An Aggregator transformation grouped by the key
  3. A Sequence Generator transformation
  4. A Lookup transformation on a reference table

Explanation

A Router evaluates conditions and routes each row to one of several output groups, which cleanly separates pass and reject flows. Aggregators summarize, lookups enrich, and sequence generators create surrogate keys, none of which split rows by condition.

Question 47Data Quality & Observability

Which metric best indicates whether a customer dimension contains duplicate customer records?

  1. Uniqueness score Correct answer.
  2. Null rate
  3. Pattern conformance rate
  4. Referential integrity rate

Explanation

Uniqueness quantifies how many values or records are distinct, so a low uniqueness score signals duplicates. Null rate addresses missing values, pattern conformance addresses format, and referential integrity addresses orphaned keys.

Question 48Data Quality & Observability

What does a data quality scorecard provide to business stakeholders?

  1. A consolidated view of quality scores across dimensions and objects over time Correct answer.
  2. A physical data model of the warehouse
  3. A list of ETL session log errors
  4. A schema-diff report between environments

Explanation

A scorecard aggregates quality metrics, often by dimension and subject area, and trends them so stakeholders can track improvement. It is not a data model, a raw log listing, or a schema comparison.

Question 49Data Quality & Observability

A rule specification is bound to a source column whose values include trailing spaces, causing valid values to fail a pattern check. What is the most appropriate remediation within the DQ logic?

  1. Lower the rule threshold to zero
  2. Disable the rule for that column
  3. Convert the column to a numeric data type
  4. Apply a trim or standardization step before evaluating the pattern Correct answer.

Explanation

Normalizing values, such as trimming whitespace, before applying the pattern ensures the rule tests the intended content rather than incidental formatting. Changing thresholds or disabling the rule hides the problem, and a type change does not remove embedded spaces.

Question 50Data Quality & Observability

Which comparison technique detects near-duplicate person records where names are spelled slightly differently?

  1. Range validation
  2. Exact string equality
  3. Referential integrity checking
  4. Fuzzy matching with a similarity threshold Correct answer.

Explanation

Fuzzy matching uses similarity algorithms and thresholds to identify records that are close but not identical, which suits variant name spellings. Exact equality misses variants, referential integrity checks key relationships, and range validation checks numeric bounds.