Question at position 4 According to lecture, data integration often requires which of the following?Defining a standard or common data modelTransforming data into a common modelCombining data at the schema levelCombining data at the record levelAll of the above

Question

Question at position 4 According to lecture, data integration often requires which of the following?Defining a standard or common data modelTransforming data into a common modelCombining data at the schema levelCombining data at the record levelAll of the above

BlackTom AI · Accepted Answer

Data integration requires establishing a common understanding and reconciliation of differing data structures, formats, and records, so all listed activities are typically necessary.

1) Relevant knowledge points and concepts
- Standard/common data model: a canonical schema or set of agreed-upon entity definitions (e.g., canonical Customer, Product) used as the target for integrated data.
- Data transformation (ETL/ELT): converting source data into the common model (type conversions, normalization, unit conversion).
- Schema-level combination: aligning and merging different schemas (attribute mapping, resolving naming mismatches, hierarchical vs. flat schemas).
- Record-level combination: deduplicating, linking, and merging individual records that represent the same real-world entity (entity resolution, record linkage, survivorship rules).

2) Step-by-step reasoning
- Start with heterogeneous sources: different systems use different schemas and formats.
- Define a common model to provide a single target and semantics for integrated data.
- Map each source schema to the common model (schema-level integration) so attributes correspond correctly.
- Apply transformations to change data types, units, value formats, and to conform to the common model.
- After transformation and schema mapping, perform record-level integration to detect and merge duplicate or overlapping records and enforce consistency.

3) Intermediate steps (illustrative)
- Identify source attributes: sourceA.firstname, sourceB.given_name → map both to Common.first_name.
- Transform values: "01/02/2020" (MM/DD) → ISO date 2020-01-02.
- Record linkage: match customer A (email matches) to customer B (phone matches) → merge with rule: prefer most recent address.

4) Why the correct answer is correct
Each listed activity addresses a different layer of the integration problem: semantic (common model), syntactic/format (transformation), structural (schema-level merging), and instance-level (record merging). Real-world integration typically requires all these steps.

5) Why other options alone are incorrect
- Defining a model only sets semantics but does not convert data.
- Transforming without schema mapping misses structural alignment.
- Schema-level combining without record-level deduplication leaves duplicate entities.
- Record-level combining without schema/format alignment cannot reliably match fields.
Thus none of the single options by themselves is sufficient.

6) Conclusion
All of the above is the correct choice, matching the provided answer.

Question at position 4 According to lecture, data integration often requires which of the following?Defining a standard or common data modelTransforming data into a common modelCombining data at the schema levelCombining data at the record levelAll of the above单项选择题

类似问题

What are typical use cases for data integration?

Identify the preprocessing steps necessary for integrating multiple data sources: (Select all that apply)

Which one of these statements explains what data integration is?

更多留学生实用工具

智能学习助手

风格化写作助手

论文查重助手

文献引用助手

课堂转译助手

课堂笔记助手

Quiz搜索助手

学校历年真题

智能刷题助手

智能匹配练习

希望你的学习变得更简单