Web of Science Author Search: Finding and Disambiguating an Author’s Records

Searching Web of Science for a specific person is a different task from a topic search, and Clarivate has built a dedicated tool for it. Rather than typing a name into the general search box and hoping for the best, Web of Science’s Author Search (surfaced in the platform’s interface as Researcher Search) walks a searcher through a name lookup, groups the platform’s best guess at which records belong to the same person into Author Records, and lets that searcher refine, combine, or verify those groupings before treating the result as a complete publication list. This guide covers that workflow specifically: how Author Search differs from a plain AU= name query, what an Author Record is and why one person can have several of them, how to refine and combine results, and how the whole process connects to a claimed Web of Science Researcher Profile, ResearcherID, and ORCID.

This is a companion to, not a replacement for, CASRAI’s How to Search Web of Science Effectively guide, which covers Advanced Search field tags (including AU= and AI=), Boolean and proximity operators, and Cited Reference Search in depth. That guide’s own author-disambiguation section explains the identifier fields; this guide goes one level deeper into the specific Author Search / Researcher Search interface, the Author Record grouping it produces, and the merge/combine workflow for cleaning up a fragmented result set. For what Web of Science is as a platform, see the Web of Science dictionary entry and the Scopus vs Web of Science vs OpenAlex comparison.

Why a Plain Name Search Isn’t Enough

A name typed into a general search field matches literally: it returns records where that string of characters appears in the author field, nothing more. That approach breaks down for two well-known reasons that anyone doing an author search needs to plan around from the start:

• Common-name collision — a name search for a common surname and initial combination can pull together records from dozens of unrelated researchers who happen to share that name string.
• Name-variant fragmentation — the same real person’s publications can appear under multiple name forms over a career: initials-only versus a full given name, a name change, or an inconsistently transliterated spelling. A literal string match misses records that don’t match the exact form searched.

Web of Science’s Author Search tool exists specifically to manage both problems at once, using algorithmic name-matching plus optional identifier evidence rather than relying on a single exact string.

How Author Search Works: Author Records

Running an Author Search does not return a single flat list of publications the way a topic search does. Instead, it returns a set of Author Records — each one a system-generated grouping of publications the platform’s matching algorithm judges likely to belong to one real person, based on signals such as co-author overlap, subject-area consistency, affiliation history, and any linked identifiers. For a common or ambiguous name, a single real person can easily be split across more than one Author Record, and conversely two different people can occasionally be merged into one if their publication patterns overlap closely. The practical implication: an Author Record is a well-informed algorithmic best guess, not a verified, guaranteed-accurate profile, and should be checked rather than taken as ground truth for anything consequential.

Refining an Author Search

When an initial Author Search returns more than one candidate record, or a record that looks too broad or too narrow, refinement tools narrow it down without abandoning the search and starting over with a manual query. Typical refinement dimensions include:

• Name variants — selecting which alternate forms of the name (different initials, spelling variants) should be included or excluded from a given record.
• Affiliation — filtering to records associated with a particular institution, useful when disambiguating between same-named researchers at different universities.
• Subject area / research field — filtering by the Web of Science subject categories associated with the publications, useful when two same-named researchers work in unrelated fields.
• Publication years — narrowing to a career window, useful for isolating a specific period (e.g., publications since joining a particular institution).
• Claimed status — whether the record has been claimed and verified by the researcher it belongs to, which is one of the stronger signals that a given grouping is accurate.

Combining and Merging Author Records

Where a genuine researcher’s publications have been split across two or more Author Records — the fragmentation problem described above — Web of Science provides ways to bring that person’s output back together into a single view:

• Viewing records as combined, without permanently altering the underlying records, by selecting multiple Author Records from a result set and choosing to view them together as one combined set of publications and citation metrics.
• Merging researcher profiles, where multiple profile-level records genuinely belong to the same person, into a single consolidated profile.

Both operations exist to solve the same underlying problem — a fragmented publication history for one real person — but they act at different levels (a temporary combined view for browsing versus a more durable profile merge), so which one is appropriate depends on whether the goal is a one-off check or a lasting, correct profile.

Author Search, ResearcherID, ORCID, and Claimed Profiles

Author Search results connect directly to Web of Science’s identifier and profile layer:

• Web of Science ResearcherID is Clarivate’s own author-profile identifier (introduced in 2008, originally under Thomson Reuters). A researcher who claims and maintains a profile gives Web of Science a verified anchor point, which materially improves how reliably that person’s Author Record gets assembled and kept correct over time, rather than depending entirely on algorithmic matching.
• ORCID iD (see CASRAI’s ORCID iD entry and the companion How to Register, Find, and Search for ORCID iDs guide) is the non-proprietary, publisher-and-funder-agnostic persistent identifier for individual researchers. A Web of Science profile can be linked to a researcher’s ORCID record, and metadata can flow between the two systems, which is one of the more reliable ways to keep an author’s Web of Science publication list accurate across name changes or affiliation moves.
• Claiming a profile is the step that turns an algorithmically-assembled Author Record into a researcher-verified one — see CASRAI’s Web of Science Researcher Profile guide for the full claiming and verification workflow. An unclaimed record reflects the platform’s best automated guess; a claimed and maintained one reflects a person actively confirming which publications are theirs and correcting the ones that aren’t.

The practical takeaway for anyone conducting an author search for a consequential purpose — a tenure dossier, a funder-mandated publication list, a systematic bibliometric analysis, or verifying a co-author’s track record — is the same one CASRAI’s broader Web of Science search guide makes: search or filter by a known identifier wherever one exists, and treat an unverified, name-only Author Record as a starting point to check against the researcher’s own claimed profile or ORCID record, not as a final answer.

Author Search vs. the AU=/AI= Field-Tag Approach

These are two complementary ways to search for an author, not competing ones:

• Author Search / Researcher Search is a guided, purpose-built workflow: it groups results into Author Records, offers refinement and merge tools, and surfaces claimed-profile and identifier status directly in the interface. It is the more approachable starting point for finding everything published by a specific person.
• The AU= and AI= Advanced Search field tags, covered in CASRAI’s How to Search Web of Science Effectively guide, let a searcher build an author search as one clause inside a larger, more complex query — combined with topic terms, journal restrictions, or date ranges using Boolean logic. This is the better route when the author search is only one piece of a bigger, reproducible search strategy (a systematic review, a saved current-awareness alert) rather than the whole task.

Merging Duplicate Author Records: What You Can Change, and What Only Clarivate Can

The single most common reason an author search looks wrong is that one researcher’s output has been split across two or more Author Records. What you can do about that depends entirely on whether the records are claimed, and it is worth being clear about the three distinct operations people tend to lump together as “merging”:

• Viewing records combined (temporary, anyone can do it). From an Author Search result set, select the Author Records that appear to belong to the same person and view them as a single combined set. This gives you a consolidated publication list and combined citation metrics for the purpose of the check you are doing right now. It changes nothing in the underlying data, it does not persist for the next person who searches that name, and it should never be described in a dossier as the researcher’s Web of Science profile — it is your own ad-hoc union of records.
• Merging into your own claimed profile (durable, the researcher does it). Once you have claimed a Researcher Profile, you can pull additional Author Records and individual publications that are genuinely yours into that profile, and remove ones that are not. This is the operation that actually fixes the fragmentation for everyone who searches your name afterwards, because it replaces an algorithmic guess with a researcher-verified list. It is also the only version of “merging” that survives the next re-run of Clarivate’s matching algorithm.
• Correcting someone else’s record, or a record you cannot claim (Clarivate does it). If two different people have been conflated into one Author Record, or publications have been attributed to a researcher who is deceased, has left research, or simply is not you, there is no self-service merge. Those are data corrections and have to be raised with Web of Science support through Clarivate’s help centre, with the specific record identifiers and the evidence of the error. Research offices doing bulk clean-up for an institution should expect this route, not the self-service one.

A practical sequencing point for research administrators: do the claiming first and the merging second. Merging Author Records into an unclaimed identity has no durable effect, so a department-wide “clean up our Web of Science records” exercise that starts with merges rather than claims produces work that quietly reverts.

Claiming Your Web of Science Researcher Profile

Claiming converts an algorithmically-assembled Author Record into a profile the researcher controls. A Web of Science Researcher Profile is free to create and maintain, and creating one does not require an institutional Web of Science subscription — though searching the underlying Core Collection database still does. The workflow, in the order it actually happens:

1. Register or sign in at webofscience.com with a personal account. If your institution subscribes, sign in through your institutional access so the platform associates your profile with the right affiliation from the start.
2. Find your Author Record via Author Search / Researcher Search, exactly as described above, and confirm which record (or records) the algorithm has built around your name.
3. Claim it from the profile area of the platform. Clarivate verifies that the claim is genuine before the record becomes yours — typically by checking an institutional email address against the affiliations on the publications, or by authenticating a linked ORCID iD.
4. Correct the publication list. Publications can be added from the Web of Science Core Collection by selecting from the matched-publications list the platform proposes, by identifier (DOI or PubMed ID), by uploading a RIS, CSV or BibTeX file exported from EndNote, Zotero or Mendeley, or by synchronising an ORCID record. Removing wrongly-attributed papers matters as much as adding missing ones — a claimed profile that still carries another person’s work is worse than an unclaimed one, because it now carries the researcher’s own endorsement.
5. Turn on ORCID synchronisation in Profile Settings, after authorising Web of Science to access your ORCID record. Automatic sync is what stops the two lists diverging again after a name change or an institutional move, which is the failure mode that created the duplicate records in the first place.

For anyone assessing a researcher rather than being one: a claimed profile is a materially stronger evidence base than an unclaimed Author Record, and the claimed/unclaimed status is visible in the interface. Treat it as the first thing to check before quoting a publication count or an h-index from Web of Science in a tenure case, promotion file or funder report.

ResearcherID: What It Is Now, and Why Old Guidance Still Mentions It

ResearcherID is Clarivate’s proprietary author identifier for the Web of Science ecosystem — an alphanumeric code in a letter(s)–numbers–year pattern (the shape of A-1234-2008, illustrative rather than a real registration), assigned when a person creates a Web of Science Researcher Profile. Its relevance to author searching is direct: a record anchored to a ResearcherID is being assembled from a verified identity rather than from name-matching alone, which is why identifier-anchored results are the ones to trust.

The reason it confuses people is that the product around it has been renamed twice:

• January 2008 — Thomson Reuters launches ResearcherID as a standalone author identifier and profile site, addressing author-name ambiguity in Web of Science.
• 2012 — ResearcherID begins integrating with the then-new ORCID registry, so the two identifiers can exchange and verify author information.
• April 2019 — Clarivate merges ResearcherID into Publons, its peer-review and editing-record platform. Existing ResearcherIDs carry over onto Publons profiles.
• April 2022 — Publons is retired as a separate destination and its profiles migrate into native Web of Science Researcher Profiles. ResearcherID survives as the identifier component inside that unified profile, not as a product you sign up for separately.

So an institutional policy, CV template, or library guide that asks a researcher for their “ResearcherID number,” or points at researcherid.com or publons.com, is asking for the identifier now surfaced on their Web of Science Researcher Profile. It is not a defunct identifier and it is not the same thing as an ORCID iD: ResearcherID has meaning only inside Clarivate’s products, while ORCID is vendor-neutral and travels with the researcher across publishers, funders and repositories. Record both where a form allows it, and link them so the sync described above can run.

Frequently Asked Questions

How do I search for a specific author on Web of Science?

Use the Author Search (Researcher Search) tool rather than a plain keyword search: enter the author’s name, then review the Author Record(s) returned, refining by affiliation, subject area, or name variant as needed to isolate the right person’s publications.

Why does my Web of Science author search return more than one record for the same person?

Web of Science groups publications into Author Records using algorithmic name-matching. A researcher whose name is common, has changed, or appears under inconsistent initials across their career can end up split across multiple Author Records. Use the refine and combine/merge tools, or a linked ORCID iD or ResearcherID, to bring those records together correctly.

Is a Web of Science Author Record the same as a claimed Researcher Profile?

No. An Author Record is an algorithmically-assembled grouping of likely-same-author publications; it can be accurate but is not verified by the person it describes. A claimed Researcher Profile means the researcher has actively confirmed and maintained which publications belong to them, which is a stronger, verified signal.

What is the difference between ResearcherID and ORCID for author searching?

ResearcherID is Clarivate’s own author identifier, specific to the Web of Science ecosystem. ORCID iD is a non-proprietary, publisher-and-funder-agnostic identifier used across the broader scholarly ecosystem, not just Web of Science. A Web of Science profile can be linked to an ORCID iD, and the two can exchange metadata, but they are separate systems maintained by separate organizations.

Do I need a Web of Science subscription to claim a Researcher Profile?

No. Creating and maintaining a Web of Science Researcher Profile is free, and claiming your author record does not require an institutional subscription. Searching the underlying Web of Science Core Collection in full still does, so an unaffiliated researcher can curate their own profile without necessarily being able to run every search described in this guide.

Is ResearcherID still used, now that Publons has been retired?

Yes. ResearcherID was merged into Publons in April 2019, and Publons itself was retired as a separate destination in April 2022 when its profiles migrated into native Web of Science Researcher Profiles. ResearcherID survives as the identifier component inside that unified profile rather than as a standalone product or site, so a form asking for a “ResearcherID number” is asking for the identifier shown on the Researcher Profile.

What if two different researchers have been merged into one Web of Science author record?

That is a data correction rather than a self-service merge. Where a single Author Record conflates two genuinely different people, or attributes publications to someone they do not belong to, raise it with Web of Science support through Clarivate’s help centre with the record identifiers and evidence of the error — the combine and merge tools in the interface only let you consolidate records, not split them apart.

Can I merge two Web of Science author records that both belong to the same person?

Yes. Where research administrators or the researcher themselves confirm that multiple Author Records or researcher profiles genuinely belong to one person, Web of Science provides tools to view them combined or merge them into a single consolidated profile.

Should I use Author Search or the AU= field tag to find an author’s publications?

Use Author Search as the default starting point for a general search for everything by a specific person, since it is purpose-built for disambiguation. Use the AU= (or AI=) Advanced Search field tag instead when the author search is one clause inside a larger Boolean query combined with topic, journal, or date restrictions.