Anticipation vs. obviousness: §102 and §103 in practice
How anticipation and obviousness differ, what each requires under Net MoneyIN, Graham and KSR, and a worked example on claim 1 of US 6,285,999.
By the One Transform team7 min read
A claim is anticipated under § 102 when a single prior-art reference discloses every limitation, arranged the way the claim arranges them. A claim is obvious under § 103 when no single reference has all of it, but the differences between the claim and the prior art are such that the claimed invention as a whole would have been obvious to a person of ordinary skill in the art. Anticipation is a matching exercise. Obviousness is an argument about what a skilled person would have done, and it needs reasons, not just references.
Anticipation: one reference, every limitation
The test is strict. Under Net MoneyIN, Inc. v. VeriSign, Inc., 545 F.3d 1359 (Fed. Cir. 2008), anticipation requires a single reference that discloses all the limitations arranged or combined in the same way as in the claim. Taking one limitation from one example in a document and another from a different, unconnected example in the same document does not meet that test.
One gap can be filled without a second reference: inherency. Under Schering Corp. v. Geneva Pharmaceuticals, Inc., 339 F.3d 1373 (Fed. Cir. 2003), a reference anticipates inherently when the missing feature is necessarily present in what it discloses. "Necessarily" is the operative word. A feature that is probably present, or present in some embodiments, does not qualify.
The USPTO's guidance on anticipation is collected in MPEP 2131. In practice, an anticipation case is one claim chart with one reference column, and every limitation has a cited passage behind it. How to read an element-by-element claim chart covers what to check in one.
Obviousness: what a skilled person would have done
35 U.S.C. § 103 bars a patent if the differences between the claimed invention and the prior art are such that the invention as a whole would have been obvious, before the effective filing date, to a person having ordinary skill in the art.
The framework comes from Graham v. John Deere Co., 383 U.S. 1 (1966). The analysis rests on:
- the scope and content of the prior art;
- the differences between the prior art and the claims;
- the level of ordinary skill in the art;
- secondary considerations, such as commercial success, long-felt but unsolved needs and the failure of others.
KSR International Co. v. Teleflex Inc., 550 U.S. 398 (2007) rejected a rigid requirement that the references themselves teach, suggest or motivate the combination. The Court said "a combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results." It also held that "obvious to try" can apply where there is a finite number of identified, predictable solutions, and it warned against hindsight: reasoning backward from the claim as a roadmap. The USPTO's treatment is in MPEP 2141.
An obviousness case is therefore a combination chart plus a written reason a skilled person would have made the combination and expected it to work.
Worked example: claim 1 of US 6,285,999
US 6,285,999 B1, "Method for node ranking in a linked database" (inventor Lawrence Page, filed 9 Jan 1998, priority 10 Jan 1997, issued 4 Sep 2001, now expired) claims:
A computer implemented method of scoring a plurality of linked documents, comprising: obtaining a plurality of documents, at least some of the documents being linked documents, at least some of the documents being linking documents, and at least some of the documents being both linked documents and linking documents, each of the linked documents being pointed to by a link in one or more of the linking documents; assigning a score to each of the linked documents based on scores of the one or more linking documents and processing the linked documents according to their scores.
Split into limitations (how to break a patent claim into limitations):
- 1a A computer-implemented method of scoring a plurality of linked documents.
- 1b Obtaining documents that are linked, linking or both, each linked document pointed to by a link in one or more linking documents.
- 1c Assigning each linked document a score based on the scores of the linking documents.
- 1d Processing the linked documents according to their scores.
Now two hypothetical references, invented for this example:
- Reference A describes a hypertext search system that collects a set of pages, records which pages link to which, counts the incoming links to each page, and orders search results by that count.
- Reference B is a paper describing a computer program that ranks scientific journals by citations, weighting each citation by the citing journal's own score and recomputing the weights until they settle.
| Limitation | Reference A | Reference B |
|---|---|---|
| 1a Computer-implemented scoring of linked documents | Present | Partially present: scores journals, not hypertext documents |
| 1b Linked, linking and both | Present | Partially present: journals cite and are cited, but through citations rather than links |
| 1c Score based on scores of linking documents | Not found: every incoming link counts the same | Present: each citation weighted by the citing journal's score |
| 1d Processing according to scores | Present: results ordered by count | Partially present: outputs a ranked table, no document processing |
Why this is not a § 102 case. Reference A has every limitation except 1c. A plain count of incoming links is not a score derived from the linking pages' scores, and inherency does not fill the gap, because nothing about counting links necessarily involves the linking pages' own scores. Reference B has 1c but describes citations between journals, not linked documents that are processed. Neither reference, alone, has every limitation arranged as claimed.
Why it is a § 103 case. Applying the Graham factors:
- Scope and content: A teaches ranking hypertext by link counts; B teaches weighting each citation by the importance of its source, computed iteratively.
- Differences: A lacks only 1c, and B supplies the weighting that 1c describes.
- Ordinary skill: someone with working experience in information retrieval and hypertext systems.
- Reason to combine: a link from one page to another and a citation from one paper to another are both directed references from one document to another. Under KSR, a challenger would argue that applying B's known weighting method to A's link counts is a combination of familiar elements by known methods yielding a predictable result.
The patent owner's responses show where the real dispute sits. B comes from a different field that a hypertext engineer would not have consulted. Applying iterative weighting to a large web graph raises problems B never addresses, such as pages with no outgoing links, so the result was not predictable. And the challenger may be using the claim as a roadmap, the hindsight KSR warns against. Evidence of commercial success or long-felt need would also be weighed; it carries most force when tied to what the claim actually requires.
Two caveats. This patent was filed before March 16, 2013, so a real challenge would apply the pre-AIA versions of §§ 102 and 103, whose timing rules differ; the logic shown here is the same. And the references are invented: this says nothing about the actual patent's validity.
Side by side
| Anticipation (§ 102) | Obviousness (§ 103) | |
|---|---|---|
| References | Exactly one | One or more |
| Test | Every limitation, arranged as claimed | Claimed invention as a whole obvious to a person of ordinary skill |
| A missing limitation | Fatal unless necessarily present (inherency) | Can come from another reference or from reasoning about ordinary skill |
| Reasoning required | Mapping and citation | A stated reason to combine and an expectation that it would work |
| Secondary considerations | Not part of the test | Part of the Graham analysis |
| Main risk for the challenger | One limitation is not quite there | Hindsight, non-analogous art, unpredictable results |
| Key sources | Net MoneyIN; Schering; MPEP 2131 | Graham; KSR; MPEP 2141 |
Both are available grounds in inter partes review, but only on the basis of patents and printed publications (35 U.S.C. § 311(b)).
Building each case in practice
- Chart before you argue. Put every candidate reference into one limitation-by-reference table. A column with no gaps is an anticipation candidate; columns with complementary gaps are combination candidates.
- Lead with a primary reference. For § 103, start from the reference that covers the most limitations and add as few others as possible, each tied to a named gap.
- Write the reason to combine down. One paragraph per combination: the problem, why the secondary reference would have been consulted, and why the result was predictable.
- Answer the rebuttal early. Address analogous art, predictability and secondary considerations before the other side raises them.
- Search the gap, not the claim. When one limitation keeps coming up uncovered, search for it alone; prior-art search beyond keywords covers the strategies.
Where One Transform fits
One Transform's Validity search builds a prior-art record and analyzes anticipation (§ 102) and obviousness (§ 103) alongside definiteness and enablement (§ 112), eligibility (§ 101), prosecution history and double-patenting exposure. It scores the patent claim by claim on a 0–10 scale and cites the references and reasoning behind each conclusion. Scores come from claim-element coverage: if an element is not found in the art, the score says so.
Next step
Chart your best reference against the claim you are assessing and see whether you have a § 102 or a § 103 case. To build the record and the analysis together, start a Validity search from the One Transform product overview.
This article is general information about patent practice, not legal advice. For advice on a specific matter, consult a registered patent attorney or agent.
