eduKateSG · Why Science?
A large open-research index is strongest when the provider route stays visible
Read CORE records through repository aggregation, metadata harmonisation, full-text discovery, identifiers, deduplication, API fields, licences and provider provenance.
Universities and subject communities maintain repositories containing articles, manuscripts, reports and other research outputs. Searching each collection separately is slow. CORE, short for COnnecting REpositories, aggregates and indexes scholarly literature from repositories and journals for discovery and machine access.
CORE describes itself as a not-for-profit, community-governed service hosted by The Open University. Its network includes institutional and subject repositories, preprint servers, and open-access and hybrid journals. The service collects, harmonises and enriches metadata and, where available, full text.
Did you know? An aggregator can show a freely readable copy even when the publisher page is not open. That route is useful only when the reader keeps the repository, version, licence and identifier attached to the copy rather than treating every PDF as interchangeable.
This guide treats CORE as a map into open research. An aggregated result helps locate candidate sources and supports text-mining workflows. The provider record, paper, methods and current corrections still determine what evidence the source can support.
Chapter index
Chapters 1–9: providers, harvesting and harmonisation
- 1. Providers are the foundation
- 2. Aggregation widens discovery
- 3. Harvesting copies metadata
- 4. Harmonisation aligns fields
- 5. Enrichment can add value
- 6. One work can have several copies
- 7. Persistent identifiers improve matching
- 8. Metadata and full text are different layers
- 9. Open access still has conditions
Chapters 10–17: full text, APIs, rights and evidence limits
- 10. Search ranking is a discovery aid
- 11. CORE Discovery finds alternative copies
- 12. The API supports structured retrieval
- 13. API keys and limits are method details
- 14. The dataset supports bulk work
- 15. Repository dashboards support quality
- 16. Full-text mining needs boundaries
- 17. Open-research infrastructure supports careers
An open-research aggregator helps when it preserves the route to each provider. CORE collects records from a global network of repositories and journals. Each provider remains responsible for the objects and metadata it exposes. This makes large-scale discovery understandable to science learners.
Try this
Evidence question: Which repository or journal supplied this record? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Record provider name and source landing page. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for providers are the foundation, its identifier or version context, and one limit that must remain visible.
Repository metadata and full text pass through harvesting, harmonisation and enrichment steps. A shared index lets one query reach many collections. Coverage depends on participating providers, harvest success and indexing decisions. It also separates an aggregated record from the repository that deposited the object.
Try this
Evidence question: Which collections are included in this result set? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Avoid calling the index a complete inventory of all research. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for aggregation widens discovery, its identifier or version context, and one limit that must remain visible.
A search result may represent one scholarly work gathered from more than one source. Repository records are gathered through technical interfaces and then processed. A harvested record is a representation of the provider’s description at a time. A careful reader can use the index as a map and the source record as evidence context.
Try this
Evidence question: When was the provider record last collected? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Keep harvest and publication dates separate. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for harvesting copies metadata, its identifier or version context, and one limit that must remain visible.
Machine access becomes responsible when licences, identifiers and provider context remain visible. Different repositories name and format creators, dates, subjects and identifiers differently. CORE maps these into a common structure for search and API use. That habit supports reproducible reviews, data work and digital-library careers.
Try this
Evidence question: Which source field became this normalised value? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Compare the aggregated and repository metadata. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for harmonisation aligns fields, its identifier or version context, and one limit that must remain visible.
An open-research aggregator helps when it preserves the route to each provider. Aggregation services can enrich language, subjects, links or other fields. Enriched values should remain distinguishable from provider-supplied assertions. This makes large-scale discovery understandable to science learners.
Try this
Evidence question: Was this field supplied or inferred? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Label enrichment provenance in the research note. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for enrichment can add value, its identifier or version context, and one limit that must remain visible.
Repository metadata and full text pass through harvesting, harmonisation and enrichment steps. Co-authors may deposit the same paper in different repositories, and a publisher version may coexist with manuscripts. Records and copies need reconciliation. It also separates an aggregated record from the repository that deposited the object.
Try this
Evidence question: Are these duplicates, versions or distinct outputs? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Compare DOI, title, authors and version statements. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for one work can have several copies, its identifier or version context, and one limit that must remain visible.
A search result may represent one scholarly work gathered from more than one source. DOIs and decentralised identifiers help systems recognise works and resolve repository records. Identifier absence should not trigger an invented match. A careful reader can use the index as a map and the source record as evidence context.
Try this
Evidence question: Which identifier anchors the object? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Resolve it independently before merging records. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for persistent identifiers improve matching, its identifier or version context, and one limit that must remain visible.
Machine access becomes responsible when licences, identifiers and provider context remain visible. A record can be searchable even when CORE does not host a full text, while some entries provide machine-readable text or a PDF route. That habit supports reproducible reviews, data work and digital-library careers.
Try this
Evidence question: Which layer is actually available? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: State metadata access and document access separately. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for metadata and full text are different layers, its identifier or version context, and one limit that must remain visible.
An open-research aggregator helps when it preserves the route to each provider. Freely readable material can carry licences, embargo histories and manuscript-version limits. Aggregation does not erase provider or publisher rights. This makes large-scale discovery understandable to science learners.
Try this
Evidence question: Which licence governs this copy? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Read the source rights statement before reuse. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for open access still has conditions, its identifier or version context, and one limit that must remain visible.
Repository metadata and full text pass through harvesting, harmonisation and enrichment steps. Ranked results reflect query matching and system signals. A high rank does not prove methodological quality or relevance to a precise research question. It also separates an aggregated record from the repository that deposited the object.
Try this
Evidence question: Which terms likely produced this rank? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Test a title phrase, DOI and subject query separately. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for search ranking is a discovery aid, its identifier or version context, and one limit that must remain visible.
A search result may represent one scholarly work gathered from more than one source. The discovery service can help route readers from a paywalled page to a freely available copy. The alternative may be a different manuscript version. A careful reader can use the index as a map and the source record as evidence context.
Try this
Evidence question: Which copy did the tool locate? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Save both publisher DOI page and repository page. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for core discovery finds alternative copies, its identifier or version context, and one limit that must remain visible.
Machine access becomes responsible when licences, identifiers and provider context remain visible. CORE API access exposes searchable metadata and, subject to fields and permissions, full-text information for application workflows. Requests need recorded queries and result windows. That habit supports reproducible reviews, data work and digital-library careers.
Try this
Evidence question: Which endpoint and parameters produced these rows? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Save request, page and field selection. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for the api supports structured retrieval, its identifier or version context, and one limit that must remain visible.
An open-research aggregator helps when it preserves the route to each provider. Programmatic access can require registration and operate under service limits or terms. A reproducible workflow records the access surface without exposing credentials. This makes large-scale discovery understandable to science learners.
Try this
Evidence question: Which plan or limit shaped retrieval? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Document the limit and keep the secret key out of the report. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for api keys and limits are method details, its identifier or version context, and one limit that must remain visible.
Repository metadata and full text pass through harvesting, harmonisation and enrichment steps. CORE offers dataset access for workflows needing a larger snapshot. A dataset release freezes content at a point in time and may differ from live search. It also separates an aggregated record from the repository that deposited the object.
Try this
Evidence question: Which release and package were analysed? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Record release date, files and licence. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for the dataset supports bulk work, its identifier or version context, and one limit that must remain visible.
A search result may represent one scholarly work gathered from more than one source. Provider-facing tools help repositories inspect harvesting, metadata quality and compliance-related information. A dashboard signal still needs local investigation. A careful reader can use the index as a map and the source record as evidence context.
Try this
Evidence question: Which source-field problem caused the warning? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Trace one issue back to the repository record. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for repository dashboards support quality, its identifier or version context, and one limit that must remain visible.
Machine access becomes responsible when licences, identifiers and provider context remain visible. Machine-readable papers can support classification, retrieval and language analysis. Researchers must respect licences, document versions and sampling bias. That habit supports reproducible reviews, data work and digital-library careers.
Try this
Evidence question: Which corpus and permitted use support the analysis? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Write inclusion rules and rights checks before mining. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for full-text mining needs boundaries, its identifier or version context, and one limit that must remain visible.
CHAPTER 17 OF 17 · Return to the source
17. Open-research infrastructure supports careers
An open-research aggregator helps when it preserves the route to each provider. Repository managers, librarians, metadata engineers, developers and data scientists maintain provider networks and discovery services. This makes large-scale discovery understandable to science learners.
Try this
Evidence question: Which hand-off fails when source provenance is lost? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Design a CORE-style record for a school research archive. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for open-research infrastructure supports careers, its identifier or version context, and one limit that must remain visible.
Contents · Previous chapter · Continue at the Science Learning Hub
Use the aggregator to discover, then verify at the provider
CORE can shorten the path from a question to open scholarly literature by bringing repository and journal records into one searchable infrastructure. Its usefulness depends on retaining provider, identifier, version and rights context.
Save the CORE record, provider, persistent identifier, query, selected copy, version, licence and retrieval date. Then read the complete paper and any correction or retraction notice before making a scientific claim.
| Reading object | What it can support | Question to keep visible |
|---|---|---|
| Provider | Repository or journal supplying the record | Where is the authoritative source page? |
| Aggregated record | Harmonised discovery representation | Which fields were transformed or enriched? |
| Full-text route | Readable or machine-accessible copy | Which version and licence apply? |
| API or dataset | Programmatic access surface | Which query, window or release was used? |
Sources and next reading
Substantive resource claims in this article were checked against the primary pages below. A structured record is a route into evidence; it does not by itself prove a scientific conclusion. Recheck the live record, evidence labels and release context before using it in assessed work or research.
For a classroom enquiry, save a short research receipt containing the question, search terms, selected object, exact identifier, version, access date and the evidence boundary you observed. Add the file or page that supports each statement. This small routine turns browsing into accountable science learning and gives a teacher, classmate or future analyst a route they can independently follow.
Before sharing the conclusion, ask another reader to retrace the route without verbal hints. If they reach a different version, file or interpretation, improve the receipt and narrow the claim. Reproducibility begins with this practical check. It teaches careful reading, supports fair collaboration and keeps uncertainty visible while evidence moves from an online service into notes, analysis and a final explanation.
A useful final review separates five questions. What object was found? Which organisation or repository maintains it? What exact evidence was inspected? Which statement does that evidence support? What remains uncertain? Answer each question in one sentence, then connect every answer to a visible field, file or source page. This review catches common mistakes: citing a collection when a file supplied the value, treating the newest release as the analysed release, confusing permission with scientific suitability, or turning a discovery count into a conclusion. It also gives students a calm way to improve work. A missing detail becomes a next research step instead of an invitation to guess. When the route is complete, the final explanation can stay simple because the identifiers, dates, versions, limitations and supporting evidence are already organised underneath it.
Primary sources
- CORE about — accessed 25 October 2026.
- CORE API — accessed 25 October 2026.
- CORE services — accessed 25 October 2026.
- CORE dataset — accessed 25 October 2026.
- CORE frequently asked questions — accessed 25 October 2026.