US 9436354

I'll search for authoritative information on US patent 9436354, including any 2026 CAFC docket activity.

Let me check the litigation history and any 2026 CAFC activity more specifically.

Let me check for any 2026 court activity and confirm the post-IPR status of the claims.

US Patent 9,436,354 — Analyst Summary

Source note: The authoritative patent text is the Google Patents full text (https://patents.google.com/patent/[US9436354](/patent/US9436354)/en), corroborated by FreePatentsOnline (https://www.freepatentsonline.com/9436354.html), Justia (https://patents.justia.com/patent/9436354) and the PTAB record. Where sources conflict I say so below.

Bibliographic Data

Abstract (verbatim)

"A database for selecting database items, where the database items are indexed by a list of item identifiers. The item identifiers may be in the form of text. An initial display is generated which includes one or more parts of the item identifiers. Selection of the one or more parts may be made and results in the generation of a display of a further one or more parts for selection. The further one or more parts may be selected in order to add to the selected one or more parts to build a larger part or whole of an item identifier. Selection from a large list of item identifiers may be carried out in a relatively short time period."

Independent Claims — Plain-Language Overview

The patent has 14 claims, with only one independent claim (claim 1); claims 2–14 all depend on claim 1. (Broader independent claims — e.g., claims 15, 17, 32 — appear in the related family publications, not in this issued patent.)

Claim 1 (the sole independent claim) — a method:

1. Initial display of one or more "parts" of item identifiers (e.g., partial strings such as "An" or "John") for a user to pick from.
2. User selects one of those parts.
3. In response, the system displays a further set of parts, where those further parts are chosen at least partly under a ranking scheme, and are laid out on the screen in relative positions corresponding to at least up, down, left, or right — with the highest-ranked part placed in one of those up/down/left/right positions.
4. User selects from the further parts, and the selections are appended to what was already selected, progressively building a larger part or the whole item identifier.
5. Two structural limits define the claim's scope:
   • the selected parts (both initial and further) are shorter than a complete word — i.e., the system offers partial strings, not whole-word completions; and
   • selection happens via a key press mapped to the on-screen relative arrangement (e.g., a directional/joystick key press matching a menu position).

In essence: an incremental, prediction-based text selection UI where the user drills down through ranked partial-string candidates, each successive set positioned directionally around a common control (joystick/arrow keys).

Dependent claims (2–14) — mostly narrowing:

• 2: iterate the selection steps until the full identifier is built, then enable selecting the item itself.
• 3–5: initial-display parts chosen by "predetermined criteria" — specifically frequency of selection (4) or ranking within a tree-hierarchy classification of the parts (5). (The specification describes a stored hierarchical word tree whose branch nodes are ranked by the sum of child rankings.)
• 6–7: display plural parts; may display an entire identifier.
• 8–10: identifiers are alphanumeric characters / a text string / a word.
• 11–14: database stored in a computing device's memory, specifically a mobile telephone (12), selections made via a joystick (13), or a PDA (14).

Litigation and PTAB History (relevant to the ID you asked about)

• Kannuu Pty Ltd. v. [Samsung Electronics Co.](/litigations/by-defendant/Samsung%20Electronics%20Co.), No. 1:19-cv-04297-ER (S.D.N.Y., filed May 10, 2019). The '354 patent was one of five patents asserted (the others: 8,676,852; 8,996,579; 9,697,264; 8,370,393), alongside an NDA-breach claim.
• IPR2020-00737 (Samsung Electronics America, Inc. v. Kannuu Pty Ltd) — petition filed March 17, 2020; instituted Sept. 23, 2020. Final Written Decision: all challenged claims 1–14 of the '354 patent unpatentable (Paper, Sept. 21, 2021; panel: APJs Droesch, Chung, Kaiser). See https://www.docketalarm.com/cases/PTAB/IPR2020-00737/Samsung_Electronics_America_Inc._v._KANNUU_PTY_LTD/. The dispute centered on claim 1 limitation [1D] (the "highest ranked" part placed in an up/down/left/right position) over the Perlman, Pu, Krohn and Dostie references.
• CAFC 21-1638 (Kannuu v. Samsung) — decided Oct. 7, 2021; affirmed the S.D.N.Y.'s denial of Kannuu's preliminary injunction seeking to bar Samsung from the IPRs (the NDA forum-selection-clause issue). http://cafc.uscourts.gov/opinions-orders/21-1638.opinion.10-7-2021_1845799.pdf
• CAFC 22-1526 / 22-1527 (appeals from IPR2020-00737 and IPR2020-00738) — affirmed (nonprecedential per curiam judgment, Oct. 11, 2023). https://www.courtlistener.com/opinion/[9431910](/patent/9431910)/kannuu-pty-ltd-v-samsung-electronics-america-inc/
• IPR certificate: Google Patents legal events record an "Inter Partes Review Certificate" (kind code K1, event dated 2024-02-13), reflecting the cancellation of claims by operation of the FWD/allied appeal termination.

2026 CAFC Docket Check — Explicit Uncertainty

I searched the Federal Circuit's 2026 dockets/opinions for anything naming this patent number 9436354. I found no 2026 CAFC proceeding concerning US 9,436,354. The known appellate activity for this patent concluded in 2021 (21-1638) and 2023 (22-1526/-1527). Note the patent lapsed for failure to pay maintenance fees (effective 2024-09-06), which makes further enforcement-driven appellate activity unlikely. I cannot rule out a very recent or not-yet-indexed docket entry, so this should be treated as "no 2026 activity found in the sources searched," not as a certified negative.

Caveats / Data Conflicts Observed

• Priority date: Google Patents lists 2005-08-12; one Unified Patents page displays "2005-08-11" — likely a time-zone artifact. The AU provisional is 2005904378, filed Aug 12, 2005, consistent with the Google entry.
• Anticipated expiration is listed by Google Patents as 2026-08-11, but the patent actually lapsed in 2024 for fee non-payment, so the listed expiration is moot.
• The Darts-IP "litigation" links on the Google page (S.D.N.Y., Fed. Cir. 21-1638 and 22-1526) are consistent with the above; no separate 2026 matter is listed there.
• I could not independently re-verify the exact wording of the 2024 IPR certificate from the USPTO patent/publication database in the searches performed; the cancellation of all claims follows from the affirmed FWD, but treat the certificate's precise text as unconfirmed.

I'll search for litigation records on US 9,436,354.

Let me get a few more specifics — the CAFC 22-1526 appeal, the IPR final outcome, and any 2026 developments.

Litigation Involving U.S. Patent No. 9,436,354

Bottom line: Every litigation/proceeding I could identify involving U.S. 9,436,354 traces back to a single dispute — Kannuu Pty Ltd. v. Samsung — comprising one district court action, two Federal Circuit appeals, and one PTAB inter partes review. I found no litigation involving the '354 patent against any defendant other than Samsung. Note that the patent's claims were ultimately cancelled, and the patent itself has since lapsed (see caveats below).

Important interpretive note: the '354 patent is one member of a large Kannuu family (the '393, '852, '579, '264, and related patents). Several of the family-related proceedings below did not actually involve the '354 patent (e.g., IPR2020-00738 concerned the '393 patent; the ex parte reexaminations concerned the '852, '579, and '264 patents). I have flagged those distinctions.

1. District Court Litigation (the core infringement case)

Sources: S.D.N.Y. Opinion & Order, Doc. 131 (July 29, 2025); Lexology summary (Aug. 5, 2025).

2. PTAB Inter Partes Review of the '354 Patent

Kannuu's principal defense — that the parties' NDA forum-selection clause barred Samsung from filing the IPR — was rejected by the Board, and its request for rehearing/POP review was denied Jan. 14, 2021.

Sources: RPX Insight – IPR2020-00737; Docket Alarm – IPR2020-00737 (’354).

3. Federal Circuit Appeals

(a) Appeal No. 2022-1526 (consolidated with 2022-1527) — merits of the IPR

• Caption: Kannuu Pty Ltd. v. Samsung Electronics America, Inc. (Director of the USPTO as intervenor)
• Appeals from: PTAB Nos. IPR2020-00737 ('354) and IPR2020-00738 ('393)
• Decided: Oct. 11, 2023
• Outcome: AFFIRMED (nonprecedential, Fed. Cir. R. 36), per curiam (Moore, C.J.; Reyna and Hughes, JJ.). This left the '354 claims cancelled.
• CourtListener opinion; Fed. Cir. judgment PDF

(b) Appeal No. 2021-1638 — forum-selection clause / preliminary injunction

• Caption: Kannuu Pty Ltd. v. Samsung Electronics Co., Ltd. and Samsung Electronics America, Inc.
• Appeals from: S.D.N.Y. No. 1:19-cv-04297-ER
• Decided: Oct. 7, 2021 — 15 F.4th 1101 (Fed. Cir. 2021)
• Issue: Case of first impression — whether the NDA's forum-selection clause barred Samsung from petitioning for IPR of the Kannuu patents (including the '354).
• Outcome: AFFIRMED the district court's denial of Kannuu's preliminary injunction (majority: Chen, J., joined by Prost, J.; Newman, J., dissenting). The court held the forum-selection clause in a mere NDA did not reach the PTAB proceedings.
• Fed. Cir. opinion PDF

4. Related Kannuu–Samsung proceedings NOT involving the '354 patent (for completeness)

• IPR2020-00738 — Samsung's IPR of U.S. 8,370,393; companion case to 2022-1526/1527.
• Ex parte reexaminations (requests filed by Samsung May 28, 2021) against U.S. 8,676,852; 8,996,579; and 9,697,264.
• The PTAB denied institution of Samsung's IPR petitions on the '264, '852 and '579 patents (Sept. 22, 2020).

Caveats and confidence

• Confidence level: High for the existence and identities of the cases above (primary documents located). Medium for the very latest 2026 status of the S.D.N.Y. action — my searches surfaced records only through the July 2025 opinion; I have no verified information about events after that point. I am stating this explicitly rather than speculating.
• The '354 patent is no longer enforceable independently: Its claims were cancelled by the IPR certificate (Feb. 9, 2024), and per the patent's official record the patent expired for failure to pay maintenance fees, with lapse effective around Sept. 6, 2024. So the '354 itself is not a live asserted patent in the ongoing S.D.N.Y. case, which now rests on two newer replacement patents.
• Do not auto-correct identifiers: I have used the patent and case numbers exactly as they appear in the sources (e.g., 1:19-cv-04297-ER; IPR2020-00737). Note the family listing also shows a separate matter, IPR2020-01022, but I did not confirm that it concerns the '354 patent, so I have not attributed it to it.
• If you need a sworn/verifiable litigation-history chart, the authoritative sources are PACER (S.D.N.Y. 1:19-cv-04297), the PTAB Patent Center / PTAB API (IPR2020-00737), and CAFC dockets for 2021-1638 and 2022-1526/1527.

I'll research the PTAB proceedings on this patent. Note upfront: the ODP block says "no proceedings," but the patent's own legal events show an IPR (IPR2020-00737, Samsung) and a Feb 2024 IPR certificate — so I need to reconcile that.

Let me dig into the FWD specifics, the appeal, and whether any additional petitions target the '354 patent.

Let me get the FWD's actual reasoning and the Federal Circuit disposition.

PTAB proceedings overview

Reconciliation note before anything else: the structured "PTAB proceedings on file" block says the ODP API returns no AIA trial proceedings for US 9,436,354. That is inconsistent with the patent's own prosecution/legal-events record, which shows an instituted IPR (IPR2020-00737), a denied Director-review request, and an Inter Partes Review Certificate issued 2024-02-09 cancelling the claims. I treat the ODP block as stale/incomplete and rely on the PTAB record itself. One IPR was filed against this patent; it reached a Final Written Decision holding all 14 claims unpatentable, was affirmed by the Federal Circuit, and the certificate has issued.

Bottom line: all claims of US 9,436,354 (claims 1–14) are cancelled. Any demand letter or infringement assertion citing this patent is asserting already-cancelled claims — that is sanctionable conduct under Rule 11 / § 285 if pursued.

IPR2020-00737 — [Samsung Electronics America, Inc.](/litigations/by-plaintiff/Samsung%20Electronics%20America%2C%20Inc.) v. Kannuu Pty Ltd

• Type: Inter Partes Review (35 U.S.C. §§ 311–319)
• Filed: 2020-03-17 (petition); Patent Owner Preliminary Response 2020-06-24 (corrected 2020-07-15); Preliminary Sur-Reply 2020-07-29
• Status: Final Written Decision — all challenged claims unpatentable; now terminated via IPR certificate (K1), effective 2024-02-09. The patent itself lapsed 2024-09-06 for failure to pay maintenance fees.
• Judge panel: Kristen L. Droesch, Minn Chung (opinion author), Jessica C. Kaiser, Administrative Patent Judges
• Petition grounds (seven grounds, all claims 1–14 challenged):
  • Ground 1 — claims 1–4, 6–10 anticipated by Perlman (US 2002/0113825 A1) under § 102
  • Ground 2 — claim 5 obvious over Perlman + Dostie under § 103
  • Ground 3 — claims 11–14 obvious over Perlman + Pu under § 103
  • Ground 4 — claims 1–4, 6–14 anticipated by Pu (US 7,152,213 B2) under § 102
  • Ground 5 — claim 5 obvious over Pu + Dostie under § 103
  • Ground 6 — claims 1–4, 6–14 obvious over Krohn (US 6,593,913 B1) + Pu under § 103
  • Ground 7 — claim 5 obvious over Krohn + Pu + Dostie under § 103
• Institution decision: Instituted on all challenged claims, 2020-09-23 (Paper 23). The panel found a reasonable likelihood that Petitioner would prevail on at least one claim and, per SAS v. Iancu, instituted on all grounds and all claims. Notably, the Board declined to give effect to the forum-selection clause in the parties' April 5, 2012 NDA, holding that making such determinations "would be contrary to Board cases expressly declining to consider forum selection clauses." Kannuu's 2020-10-07 rehearing request (§ 42.71(d)) and simultaneous POP request were denied, and its parallel effort to enjoin the IPR in SDNY was denied on 2021-01-19 and affirmed on appeal (see Appeal, below).
• Final Written Decision: 2021-09-21 — "For the reasons discussed below, we determine Petitioner has shown by a preponderance of the evidence that claims 1–14 of the '354 patent are unpatentable." The judgment caption reads: "Final Written Decision Determining All Challenged Claims Unpatentable — 35 U.S.C. § 318(a)." The Board also granted both parties' motions to seal and granted Patent Owner's Motion to Strike.
  • Claim-level granularity: claim 1 is the sole independent claim; claims 2–14 depend from it. All of claims 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 and 14 were held unpatentable. No claim was sustained.
  • Caveat on ground allocation: the public FWD is a redacted version, and I could not confirm from the accessible text precisely which of Grounds 1–7 carried the judgment for each claim. Kannuu's Sur-Reply framed Grounds 1, 2, 3, 6 and 7 as the live disputes (i.e., the anticipation-by-Pu grounds and the Krohn+Pu ground were addressed primarily in the Reply), which suggests the Board's dispositive reasoning rested principally on the Perlman and/or Krohn+Pu theories. In the companion '393 FWD (IPR2020-00738) the Board expressly held claim 1 anticipated by Perlman and claims 1 and 8 anticipated by Pu. I note this as context only — I will not attribute a specific ground to a specific '354 claim without the unredacted text.
  • Kannuu's non-obviousness case rested heavily on secondary considerations (Samsung's internal "industry praise" of Kannuu's predict-next-letter technology, alleged copying, commercial success, failure of others). The Board did not find it sufficient to overcome the art.
• Settlement / termination: No settlement. The proceeding ran to a merits FWD; the parties did litigate discovery aggressively (Kannuu's motion for additional discovery from Samsung Electronics Co., Ltd. (Korea) and its 30(b)(6) topics on Samsung's 2012–13 predict-next-letter development effort, which Petitioner opposed as a "fishing expedition").
• Director review: Samsung/Kannuu sought Director review of the FWD (Ex. 3100). Denied 2022-01-07 by Commissioner for Patents Andrew Hirshfeld, performing the functions of the Director — the order states the FWD in IPR2020-00737 (and IPR2020-00738) "is the final decision of the agency."
• Appeal: Yes — Federal Circuit Nos. 2022-1526 and 2022-1527 (the '354 and '393 FWDs). Judgment issued 2023-10-11, AFFIRMED, per curiam (Moore, Chief Judge, and Reyna, JJ., plus a third panel member), nonprecedential. The en banc/panel challenge to the Board's refusal to enforce the NDA forum-selection clause failed both in the separate § 1292(a)(1) appeal (Kannuu Pty Ltd. v. Samsung Elecs. Co., 15 F.4th 1101 (Fed. Cir. 2021), aff'g the SDNY denial of a preliminary injunction, decided 2021-10-07) and on the merits appeal of the FWDs.
• Certificate: Inter Partes Review Certificate (kind code K1) for the '354 patent, trial no. IPR2020-00737, effective 2024-02-09 (recorded 2024-02-13) — the certificated cancellation of claims 1–14.
• Defensive value: Total. There is no live claim to infringe. Claims 1–14 are cancelled as of the certificate; the patent also lapsed for non-payment of maintenance fees on 2024-09-06. Any assertion of this patent today is meritless on its face, and a demand letter citing claims 1–14 is a strong candidate for a § 285 fee motion or a Rule 11 letter.

Strategic summary

Claim status. Every claim of US 9,436,354 is CANCELED: claims 1–14, with claim 1 the sole independent claim and claims 2–14 dependent on it. There are no surviving claims and no untested claims. This was not a partial win — the FWD was styled "Determining All Challenged Claims Unpatentable," the Federal Circuit affirmed, Director review was denied, and the K1 certificate issued 2024-02-09. The patent's own status is now "Expired – Fee Related" (lapse recorded 2024-10-14, effective 2024-09-06), which independently moots any infringement theory going forward.

Estoppel landscape. Under § 315(e)(2), Samsung Electronics America, Inc. and its privies are estopped in the SDNY action and in any ITC proceeding from asserting any ground they raised or reasonably could have raised in IPR2020-00737 — i.e., Perlman, Pu, Krohn, Dostie, and any obviousness combination reasonably available from them. That estoppel has no practical bite now that the claims are cancelled. For a new defendant facing an assertion of the '354 patent, the estoppel runs the other way: you are not bound by Samsung's IPR, but you also don't need one — the K1 certificate is dispositive, and § 315(e)(1) estoppel plus the cancellation certificate foreclose the patent owner from reviving these claims at the PTAB.

Pattern signals. Samsung filed five IPR petitions against the five patents-in-suit on the same day (IPR2020-00736 through -00740, petition dated 2020-03-17): the PTAB denied institution on three (the '264, '852 and '579 patents — IPR2020-00736, -00739, -00740, all denied 2020-09-22 on the merits of failing to show unpatentability) and instituted two (the '354 and '393). Samsung then pivoted to ex parte reexamination on the three patents that survived institution ("852, '579, '264); the district court case, Kannuu Pty Ltd. v. Samsung Electronics Co., No. 1:19-cv-04297-ER (S.D.N.Y.), was stayed pending the IPRs (extended 2022-01-26). The petition was not filed by a defensive aggregator — it was Samsung's own campaign, driven by the underlying NDA/licensing relationship and Kannuu's SDNY suit, though Unified Patents is separately linked to the family's litigation-portal data. Kannuu litigated aggressively on procedure (POP request, rehearing request, discovery motions, motion to strike, the § 1292(a)(1) injunction appeal, and the FWD appeal) — and lost at every stage on this patent.

One unverified item. A search surfaced a PTAB petition listing ("IPR2020-01022") that appears in the same document bundle as the '354 patent number. I could not confirm that IPR2020-01022 is a separate challenge to US 9,436,354, and I will not assert it is. Treat the proceedings list above (IPR2020-00737 only) as complete pending confirmation from PTAB E2E.

Recommended next steps

1. If you are a defendant being asserted against US 9,436,354 — do not build any defense on the merits. Pull the certificate and the FWD, and quote the disposition: "we determine Petitioner has shown by a preponderance of the evidence that claims 1–14 of the '354 patent are unpatentable." Cite the IPR2020-00737 Final Written Decision (2021-09-21), the Federal Circuit judgment of 2023-10-11 in Nos. 2022-1526/2022-1527 (affirmed), the Director-review denial of 2022-01-07, and the Inter Partes Review Certificate (K1) effective 2024-02-09. Demand that any assertion be withdrawn; if it is not, consider a Rule 11 safe-harbor letter and a § 285 fee request. Sources:
   • PTAB E2E / PTAB Center for IPR2020-00737: https://ptacts.uspto.gov/ptacts/public-informations/petitions/... (search "IPR2020-00737")
   • FWD (redacted) via Docket Alarm: https://www.docketalarm.com/cases/PTAB/IPR2020-00737/Samsung_Electronics_America_Inc._v._KANNUU_PTY_LTD/
   • Federal Circuit judgment (2022-1526 / 2022-1527), 2023-10-11: https://www.courtlistener.com/opinion/[9431910](/patent/9431910)/kannuu-pty-ltd-v-samsung-electronics-america-inc/
   • Forum-selection-clause appeal (2021-1638), 2021-10-07: http://cafc.uscourts.gov/opinions-orders/21-1638.opinion.10-7-2021_1845799.pdf
2. If you are evaluating the broader Kannuu family (the '393, '852, '579, '264, '852, and the continuation US 11,573,939) — note that the '354 result is not representative. The '393 patent's claims were likewise cancelled (IPR2020-00738 FWD, affirmed), but the '264, '852 and '579 patents survived institution, and Samsung resorted to ex parte reexamination on those. Validity exposure must be assessed patent-by-patent.
3. No active proceedings to monitor. IPR2020-00737 is fully terminated (FWD → affirmance → K1 certificate). There is no institution deadline, oral hearing, or statutory § 316(a)(11) FWD deadline outstanding. The absence of further PTAB activity is explained by the fact that the patent's claims are already gone and the patent has lapsed — there is nothing left to challenge.

Confidence flags: proceeding number, parties, panel, filing/institution/FWD dates, claim-level outcome, certificate date, and appeal disposition are all well-supported by the record above. Not confirmed: (a) the precise ground-by-ground allocation in the redacted FWD; (b) the identity of the third Federal Circuit panel member on 2022-1526; (c) the status of IPR2020-01022.

I'll research the assignment record and litigation history for US 9,436,354 before building the analysis.

Let me gather more on the assignment records and Kannuu's corporate status.

Let me verify the assignment records and Kannuu's status as an operating company vs. asserter.

Ownership & NPE-Pattern Analysis — US 9,436,354

Sourcing note (read first): I could not open a live session against assignmentcenter.uspto.gov in this run, so the ownership entries below are reconstructed from the Google Patents legal-events mirror of the USPTO assignment record (which republishes reel/frame, recordation date, and conveyance text) plus the PTAB/SDNY/CAFC record. Every reel/frame cited should be re-verified at the Assignment Center by patent number. The "correspondent of record" field is not exposed in the mirrored data, so I have not named one — fabricating an attorney would be worse than reporting the gap.

Inventors

• Kevin William Dinn — sole named inventor on the '354 patent (and on the entire Kannuu family). He is the founder/principal behind the patentee, Kannuu Pty Ltd, and is named individually on the assignment document (reel 034999/0042, "ASSIGNOR: DINN, KEVIN WILLIAM").
  • Employer at filing time: Kannuu Pty Ltd (the PCT applicant of record for WO2007019610, the parent of this family). Rights were not formally papered to Kannuu until an assignment executed 2008-06-30 — i.e., executed by Dinn in favour of his own company, not by a departing employee in favour of a third party.

Unusual patterns — noted, but not the classic "inventor exodus" tell:

• There is no multi-inventor team, so the "all inventors departed within 12 months of filing" fire-sale precursor cannot apply. The single inventor and the assignee are the same economic interest.
• There is a related-party shuffling pattern worth flagging: Dinn's earlier patent filings sit with other Dinn-controlled vehicles — Besdi Pty Ltd (WO2002039244 "Device interface" 2000; WO2003077102 2002; WO2004017188 2002) and K W Dinn Holdings Pty Limited (AU provisional PS1072, priority 2002-03-13, "Improved device interface"). So the same inventor's IP has moved among several personally-controlled entities over two decades. That is a technology-holding-vehicle habit, not evidence of NPE shelling.

Original assignee

Kannuu Pty Ltd (Australia) — named as assignee on the face of the '354 patent and as record owner from 2008 to the present.

• Primary line of business: development/licensing of predictive search-and-navigation input technology for constrained devices (mobile handsets, remotes, and later smart TVs / Blu-ray players). Per the CAFC record, Kannuu built demonstration platforms it showed to Samsung in 2012–2013 and pursued a license-and-supply business model rather than mass-market device sales. The 2021 CAFC opinion describes it as "a small company with limited employees and resources."
• Did it ship a product embodying the claims? No evidence in the record of a commercially shipped US product reading on the claims. Its monetization route was OEM licensing/evaluation deals. Treat "operating company that ships the claimed product" as unproven.
• Current status: No record of acquisition, merger, dissolution, or bankruptcy. Google Patents still lists Kannuu Pty Ltd as current assignee. The company let this patent lapse (maintenance-fee lapse recorded 2024-10-14), while continuing to press two newly-issued patents in the same SDNY action after the originals were invalidated (2024-07-29 leave to amend granted).

Possibly-relevant ownership nuance: the 4th-year maintenance fee was paid in 2020-02-20 as a LARGE ENTITY payment. A small Australian developer paying large-entity fees is a data point that could indicate a large-entity license relationship or a status/classification irregularity — but it is not evidence of an assignment or of NPE status, and I flag it only as a curiosity.

Assignment timeline

Only one assignment is of record for this patent. There are no post-issuance transfers — no shell LLC, no aggregator, no security interest, no name change.

• 2008-06-30 (executed) / recorded 2015-02-20 — Reel 034999/0042
  • Conveyance: Assignment ("ASSIGNMENT OF ASSIGNORS INTEREST")
  • Assignor: Kevin William Dinn
  • Assignee: Kannuu Pty Ltd., Australia
  • Correspondent: Not disclosed in the fetched record — must be verified in the Assignment Center; cannot be reported without fabrication.
  • Context: Founder-to-company assignment — the inventor conveying his rights in the family to his own operating company. Recorded in 2015 contemporaneously with the filing of this continuation (application 14/627,914, filed 2015-02-20), i.e. a re-recordation against a new continuation rather than a fresh commercial transfer.

Because there is at least one record, I proceed — but note that the ownership chain is a single link, and that is itself the central finding.

Timeline diagram

timeline
    title Ownership of US 9436354
    2005 : Priority AU provisional filed
    2006 : PCT AU2006001151 filed by Kannuu
    2008 : Dinn assigns rights to Kannuu
    2015 : Continuation filed and assignment recorded
    2016 : Patent issued to Kannuu
    2019 : Kannuu sues Samsung in SDNY
    2021 : PTAB invalidates the claims
    2024 : IPR certificate issued
         : Patent lapses for unpaid fees

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only recorded transfer is Dinn → Kannuu Pty Ltd (reel 034999/0042, executed 2008-06-30), i.e. inventor to his own operating company. There is no downstream move to an "IP / Holdings / Ventures / Licensing" LLC, no registered-agent address, and no single-member Delaware/Texas vehicle anywhere in the record.

2. Known asserter in the chain — not present. The sole assignee, Kannuu Pty Ltd, does not appear on any of the enumerated lists (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities). To the contrary, Kannuu is the original owner asserting its own R&D output — an operating-side posture. (Note: Kannuu was the patent owner / appellee in the PTAB and CAFC proceedings; the PTAB petitioner/assailant was Samsung.)

3. Repeat correspondent across the chain — not assessable. Only one assignment is of record, so there is no "same lawyer on every link" pattern to detect. The correspondent field is not present in the mirrored data; no finding either way.

4. Cascading transfers — not present. One assignment, no chained LLCs, no shared correspondent addresses, no <24-month serial transfers.

5. Pre-litigation transfer — not present. The only assignment (executed 2008-06-30) predates the first infringement suit (Kannuu v. Samsung, SDNY 1:19-cv-04297, filed 2019-05-10) by roughly eleven years. Nothing in the chain was arranged near the suit date; standing was never in question.

6. Bankruptcy fire-sale — not present. No Chapter 7/11, no court-supervised patent sale. The company instead voluntarily let this patent lapse for non-payment (lapse recorded 2024-10-14).

7. Privateering — not present. No operating company transferred the patent to an NPE to assert on its behalf. The asserter is the original owner.

8. Defensive aggregator — not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. The patent was asserted, then invalidated, then abandoned — it was not neutralized by aggregation.

Verdict

Insufficient data — for the specific question of NPE/troll-pattern signalling.

Justification: the ownership record for US 9,436,354 contains only the original assignment — inventor Kevin William Dinn to Kannuu Pty Ltd at reel 034999/0042, executed 2008-06-30 and recorded 2015-02-20 — with the original operating assignee remaining record owner to expiration. None of the transfer-based NPE signals (shell-entity conveyance, cascading LLCs, pre-litigation assignment, bankruptcy fire-sale, privateering) can be evaluated because no such transfers exist in the chain, and Kannuu does not match any published high-frequency-plaintiff list.

That said, two facts should temper a purely dismissive reading: (a) the patent was genuinely asserted — Kannuu sued Samsung in SDNY on 2019-05-10, and the '354 claims were cancelled in IPR2020-00737 (final written decision 2021, IPR certificate K1 issued 2024-02-09, CAFC affirmed 2022-1526 on 2023-10-11) — after licensing negotiations collapsed; and (b) the assertion posture is that of a small, product-light licensor suing a giant, which is NPE-like behavior without the NPE-shell paper trail. That is a behavioral observation, not a recorded-assignment finding, and I decline to upgrade it to "NPE — moderate confidence" on naming/posture alone.

Verify at: https://assignmentcenter.uspto.gov/ (search patent 9,436,354) — confirm reel 034999/0042 and check for any additional security-interest or name-change records not surfaced in the Google Patents mirror; and https://patents.google.com/patent/US9436354/en for the legal-events feed on which this reconstruction is based.

I'll start by verifying the patent's particulars and then examine its cited prior art.

Let me dig into the IPR2020-00737 proceeding, which identifies the actual prior-art grounds used against this patent.

Let me get details on the four references actually asserted in the IPR proceedings.

Prior-Art Analysis — U.S. Patent No. 9,436,354 B2

Note on scope/method: I verified the patent itself against the USPTO/Google Patents record. I was able to identify and characterize the references the examiner cited on the face of the patent and, importantly, the four references Samsung actually asserted in the inter partes review. I was not able to independently verify the full bibliographic citations of three of the four IPR references (Pu, Krohn, Dostie) before my search budget ran out — I flag this explicitly below rather than guess. No patent numbers have been altered or auto-corrected.

1. Confirmation of the patent (literal identifier: 9436354)

Claim 1 is the whole ballgame. Any §102 anticipation analysis reduces to whether a single reference discloses all of:

• (a) database items indexed by a list of item identifiers;
• (b) generating an initial display of one or more parts of item identifiers;
• (c) enabling selection of those parts;
• (d) in response, generating a display of further part(s) selected based at least in part on a ranking scheme;
• (e) the further parts arranged relative to one another corresponding to at least an up, down, left or right position;
• (f) the highest ranked of those parts positioned in one of the up/down/left/right positions;
• (g) enabling selection to add parts together to build a larger part or whole identifier;
• (h) the parts being shorter than a complete word;
• (i) selection enabled by a key press corresponding to the relative arrangement on the display.

Claims 2–14 are dependent and add: iteration/selection of completed identifier (2); predetermined criteria for initial display (3); frequency-of-selection criterion (4); tree-hierarchy ranking criterion (5); plurality of parts (6); whole identifier displayed (7); alphanumeric (8), text string (9), word (10); database in computing-device memory (11); mobile telephone (12); joystick selection (13); PDA (14).

Which §102 applies: The application was filed 2015-02-20 but all claims are entitled to the 2005-08-12 priority date, so the pre-AIA §102 framework governs. The §102(b) critical date is one year before the U.S. filing date in the chain (Aug. 11, 2006) ≈ Aug. 11, 2005; references published before that date are §102(b) art, references published between then and the invention date may be §102(a) art, and U.S. patents/publications filed before but published after the priority date are candidates for §102(e). Because claim 1 has a single integrated combination of limitations, most of these references are "potentially anticipatory" only for a subset of claim 1's elements — the true §102 risk is concentrated in the four references Samsung litigated, which are the ones I weight most heavily.

2. The most relevant prior art — the references actually asserted in IPR2020-00737

Samsung's petition asserted §102 anticipation grounds (not just §103) against US 9,436,354, based on four references not considered during prosecution:

(Grounds and claim groupings taken from Patent Owner's Response and Sur-Reply in IPR2020-00737, Kanuu Pty Ltd., U.S. Pat. No. 9,436,354, discussion at §§ III.B–III.H.)

2.1 Perlman — the primary §102 reference

• Full citation: U.S. Patent Application Publication No. US 2002/0113827 A1, Stephen G. Perlman, "Apparatus and method for selecting data."
• Filing date: 2001-02-22 (priority 2001-02-22).
• Publication date: 2002-08-22.
• §102 posture: Published well before the Aug. 11, 2005 critical date → §102(b) art.
• Brief description: Discloses a user interface/apparatus for selecting data in which candidate data items are presented for user selection in a spatial arrangement navigable by a multi-directional input device. This is the closest single-reference teaching of the directional-arrangement aspects of claim 1.
• Claims potentially anticipated: Claims 1–4 and 6–10 (as Samsung asserted in Ground 1). Its disclosure bears most directly on claim 1 elements (b), (c), (e), (g) and (i), and on claim 1's "up/down/left/right" arrangement. The disputed element — and the one Patent Owner argued Perlman does not disclose — is element (f), the "highest ranked" part being placed in one of the up/down/left/right positions (i.e., ranking-driven spatial placement), which the Board adopted as a limiting construction ("Claim 1 is Limited to Placing the 'Highest Ranked' Parts of Item Identifiers In the 'Up, Down, Left or Right' Positions").

2.2 Pu — the second §102 reference

• Full citation: Not verified. The record refers to this reference only as "Pu." It is treated as a single anticipating reference (Ground 4, claims 1–4 and 6–14), and also supplies the device/memory aspects in Grounds 3, 5, 6 and 7 (claims 11–14, mobile telephone/PDA/joystick).
• Dates: Not verified from the materials retrieved.
• Brief description (inferred from role in the grounds): A single patent or printed publication disclosing selection of items from an indexed list by user selection of parts, asserted to disclose the claim 1 combination and the computing-device environment of claims 11–14.
• Claims potentially anticipated: Claims 1–4 and 6–14 (Ground 4); combined for claims 11–14 (Ground 3) and claims 1–4, 6–14 (Ground 6).
• ⚠️ Verification needed: Pull IPR2020-00737 Paper 1 (Petition) and Exhibits 1003–1006 at the PTAB portal to obtain the exact publication number, date and title before relying on this entry.

2.3 Krohn — secondary combination reference

• Full citation: Not verified (record refers only to "Krohn").
• Brief description: Asserted only in combination with Pu (Grounds 6 and 7). Patent Owner argued Krohn does not disclose a "database" and does not disclose claim elements [1C] and [1D].
• Claims potentially implicated: claims 1–4, 6–14 (Ground 6); claim 5 (Ground 7) — as a combination, not alone.
• ⚠️ Verification needed (same source as above).

2.4 Dostie — secondary combination reference (claim 5 only)

• Full citation: Not verified (record refers only to "Dostie").
• Brief description: Asserted only for claim 5 (Grounds 2, 5, 7), i.e., to supply the "ranking of the one or more parts in relation to a tree hierarchy classification" limitation. It is not asserted as an anticipating reference against any independent claim.
• Claims potentially implicated: claim 5 only, and only in combination.
• ⚠️ Verification needed (same source as above).

2.5 Outcome of the IPR (context for a §102 analysis)

The Board instituted review on the '354 patent on 2020-09-23, held a final written decision, and the case terminated 2021-09-21. An inter partes review certificate (kind code K1) issued 2024-02-09 for trial IPR2020-00737, and the Federal Circuit affirmed in Kannuu Pty Ltd. v. [Samsung Electronics America, Inc.](/litigations/by-plaintiff/Samsung%20Electronics%20America%2C%20Inc.), Nos. 22-1526/22-1527 (nonprecedential, Oct. 11, 2023), with the USPTO Director intervening. The issuance of a §318(b) certificate and the posture of the appeal indicate that the challenged claims did not survive the IPR challenge. Given I could not re-verify the certificate's substance in this session, I flag this as the record's strong implication rather than a quotation.

3. Examiner-cited references on the face of US 9,436,354 (the "Citations" list)

The patent faces 107–111 citations (with substantial overlap between the "Citations" and "Patent Citations" tables). The following are the substantively relevant ones, grouped by the claim-1 element they most threaten. Dates are given as (priority date / publication date) from the record.

Group A — Hierarchical / menu-based selection (threaten claim 1 preamble, 1(b), 1(g); claim 5)

Group B — Predictive / partial-word completion and ranking (threaten elements 1(b), 1(d), 1(g), 1(h); claims 3–4)

Group C — Directional / navigation-key and key-press selection (threaten elements 1(e), 1(f), 1(i); claims 12–13)

Group D — Same-inventor / same-family prior filings (highly material)

These are the applicant's own earlier filings, published before the critical date, and are therefore §102(b) art regardless of common inventorship:

Group E — References that are not §102 prior art

Several documents in the citation tables post-date the Aug. 12, 2005 priority date and are therefore not available as §102 art. The clearest example is US 7,797,228 B2 (Apple, priority 2006-10-13, published 2010-09-07), "Method, system, and graphical user interface for text entry with partial word display" — this appears in the "Citations" table as a later document and is relevant only as a §103/state-of-the-art or litigation context reference. Similarly, items in "Families Citing this family" (16 entries, e.g., the Kno gesture-control publication, Google's Predictive virtual keyboard, Parikh's US 8,930,181) are forward citations, not prior art.

4. Non-patent literature cited

5. Bottom line — where the §102 risk is concentrated

1. Perlman (US 2002/0113827 A1) is the single most dangerous §102 reference, and it is the only reference Samsung used alone to assert anticipation against the independent claim (Ground 1, claims 1–4 and 6–10). It was not before the examiner.
2. "Pu" is the second independent-claim anticipation reference (Ground 4, claims 1–4 and 6–14) — again, not before the examiner. I could not verify its full citation; treat it as the critical unknown.
3. Krohn and Dostie are combination-only references (Grounds 6/7 for claims 1–4, 6–14 and claim 5 respectively) and are therefore §103 material rather than standalone §102 art.
4. Among examiner-cited art, the most relevant under §102 are US 5,267,174 (Symantec, hierarchical menus), US 5,787,417 (Microsoft), US 5,892,321 and US 6,377,965 (Microsoft, text completion), US 5,953,541 (Tegic, frequency ordering → claim 4), WO 2005/064445 A1 (Nokia, four-way input → element (f)/(i)), and the Dinn/Besdi family publications (WO 2002/039244, WO 2003/077102, WO 2004/017188).
5. No single examiner-cited reference appears to disclose the full claim 1 combination — the combination of ranking-driven selection plus the "highest ranked placed in the up/down/left/right position" limitation (element (f)) is precisely the limitation Patent Owner successfully argued was missing from Perlman. That is why the challenger reached for Art not considered during prosecution.

Caveats: (i) The full citations for Pu, Krohn and Dostie are unverified — retrieve IPR2020-00737 Paper 1 and Exhibits 1003–1006 at https://portal.unifiedpatents.com/ptab/case/IPR2020-00737 before relying on them. (ii) A "potentially anticipates" designation is a §102 risk flag, not a finding: pre-AIA §102 anticipation requires every element in a single reference, and the Board's claim construction required the "highest ranked" part to occupy a directional position. (iii) The §102(e) characterizations depend on exact §102(e) filing dates, which I derived from the priority-date column of the citation table and should be confirmed against the actual US filing dates.

I'll research the specific prior-art references cited on this patent's page before building the § 103 analysis.

Now let me pin down the other IPR references (Pu, Krohn, Dostie) and the key non-patent art.

Let me identify the Krohn and Dostie references and confirm the Board's ultimate findings.

Obviousness Analysis — US 9,436,354 B2 under 35 U.S.C. § 103

Cross-reference note: This analysis builds on the previously generated sections (bibliographic data, abstract, plain-language claim overview, and litigation/PTAB history) and does not repeat them. Where the earlier sections already established a fact — e.g., that claim 1 is the sole independent claim and that the FWD in IPR2020-00737 held claims 1–14 unpatentable — I treat that as authoritative and reason forward from it.

Governing law and date. The '354 patent's claims all enjoy an effective filing date of 12 August 2005 (AU provisional 2005904378), and no claim in the chain has an effective filing date on or after 16 March 2013. Pre-AIA § 103(a) therefore governs (consistent with the Board's use of pre-AIA § 102(b)/§ 103(a) in the ground structure I retrieved). The statutory framework is Graham v. John Deere (scope/content of the prior art; differences; PHOSITA level; secondary considerations) and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) (combination of known elements with predictable results; design incentives and market forces).

1. Person Having Ordinary Skill in the Art (PHOSITA)

Based on the patent's own field framing (https://patents.google.com/patent/US9436354/en, Background) and the art cited on that page, the PHOSITA would be a software engineer with a bachelor's degree in computer science/electrical engineering (or equivalent experience) and roughly 2–3 years of experience designing text-entry and menu-selection interfaces for devices with constrained input (mobile phones, PDAs, TV/media remotes, in-vehicle systems). Such a person would be familiar as of Aug 2005 with: predictive/disambiguating text entry (T9/iTAP class), soft keyboards, trie/prefix-tree dictionaries, probability-ranked candidate lists, and directional-key (joystick/D-pad) navigation of on-screen menus. This is not a crowded or esoteric art — it is the routine design space of consumer electronics text input.

2. The Prior-Art Record Relied Upon

I restricted the primary analysis to references listed in the Prior Art / Citations section of the Google Patents page for US 9,436,354, and to the four references actually litigated in IPR2020-00737 (three of which appear on the page).

Perlman claim text retrieved from https://patents.justia.com/patent/20020113827; Masui/POBox text from https://dlnext.acm.org/doi/10.1145/[274644](/patent/274644).[274690](/patent/274690) and http://www.pitecan.com/papers/HUC99/HUC99.pdf; IPR ground structure from the Patent Owner's Response/Sur-reply listings at https://www.docketalarm.com/cases/PTAB/IPR2020-00737/Samsung_Electronics_America_Inc._v._KANNUU_PTY_LTD/ and the IPR2020-00738 FWD excerpt at http://bannerwitcoff.com/wp-content/uploads/2022/08/IPR2020-00738.pdf.

3. Claim 1 Element-by-Element: What the Art Discloses

Claim 1 (see the full text at https://patents.google.com/patent/US9436354/en) breaks into seven elements. Patent Owner's own IPR papers conceded all elements except [1D] — a strong signal that only one limitation was genuinely in dispute.

Net result of the mapping: Perlman alone supplies 1A–1C, 1E–1G and ranking (part of 1D), and only arguably leaves a gap on the placement of the highest-ranked item specifically in an up/down/left/right position. Pu supplies multi-character "parts shorter than a complete word" plus frequency/tree ranking and the mobile/PDA/keypad context. Masui supplies the express rationale for partial-word entry on few-key devices. That is precisely the shape of a § 103 case.

4. Combinations That Render Claim 1 Obvious

Combination A (primary): Perlman + Pu (optionally + Masui/POBox)

What each brings.

• Perlman teaches the complete architectural skeleton: a database of items indexed by text identifiers, an initial on-screen group of selectable character options spatially arranged to mirror a small physical button set, ranking of each successive group by probability within the database, and a "more" function to retrieve the next-most-common set. See Perlman claims 1–2, 7–12 and the passages: "a new set of characters are mapped to the character-mapped buttons … which correspond to the most common second letters"; "display logic causes a group of buttons to be displayed having a spatial configuration related to that of said group of buttons on said remote control device" (claim 8).
• Pu teaches the same technique at the level of parts/prefixes shorter than a complete word, with ranking by relative frequency, and expressly frames the field as spanning "telephones, cellular telephones, internet phones, pagers, portable computers, navigational devices," with a telephone-keypad embodiment.
• Masui/POBox supplies both 1F and the motivation: text is entered by selecting from dynamically filtered dictionary candidates after "a user enters a small part of the word," and the technique is expressly extended to "PDAs and cellular phones" using "a jog-dial with a small number of keys."

Motivation to combine (KSR factors).

1. Same field, same problem. Both Perlman and Pu address selecting/entering items on devices lacking a full keyboard — a remote control and a phone keypad respectively. Combining a known predictive-mapping scheme with a known multi-character candidate set is the use of a known technique for its known purpose.
2. Express suggestion in the references. Perlman itself is premised on the fact that a remote has few buttons; its display already mirrors a directional button cluster. Pu's specification itself points to the broader device class (phones, PDAs). Masui explicitly states the same approach works on a cellular phone with a small number of keys. No reference teaches away from ranked, positionally-mapped predictions.
3. Design incentive / market forces. By 2005 the industry was under strong, well-documented pressure to reduce keystrokes on one-handed and 4-direction-key devices (TV remotes, phones) — exactly the pressure the page's own Besdi/Dinn references (WO 02/039244; WO 03/077102; WO 2004/017188) and the Siemens '270 and Griffin '330 references address.
4. Predictable result. Placing the highest-ranked candidate in a directional (up/down/left/right) slot is the natural, if not inevitable, way to map ranked options onto a 4-way control; the art already taught both the ranking (Apple '672; Tegic '541/'554) and the positional mapping (Perlman; Siemens '270).

Reasonable expectation of success. Both references were in the same technical discipline, had compatible architectures (a ranked candidate table driving a small on-screen menu), and combining them required no new hardware. The expected benefit — fewer presses to reach a target identifier — was the very benefit each reference already claimed. This is a classic "predictable variation" under KSR.

Combination B (as instituted): Krohn + Pu (Ground 6 of IPR2020-00737)

The Board instituted on this combination for claims 1–4 and 6–14. Here Krohn supplies the character-selection-from-a-tree mechanism (with "more" paging), while Pu supplies the database, the ranking by relative frequency, the multi-character identifiers, and the mobile/PDA/keypad context. Patent Owner attacked it on two fronts — that Krohn lacks a "database," and that Krohn/Pu lack elements [1C]/[1D] — which, again, confirms that the combination otherwise met the claim. The motivation is the same as Combination A: both references address character entry on button-limited devices, and Pu's own disclosure of predefined lists, relative-frequency ranking, and telephone-keypad implementations supplies the missing "database" and ranking elements that a PHOSITA would naturally pair with Krohn's tree-based character selection.

Combination C (for the directional-placement sub-limitation): Perlman or Pu + Siemens '270 and/or Griffin '330

If one were to treat element [1D]'s "highest ranked … in one of the up, down, left and right positions" as the only unmet limitation (the position Patent Owner took), the secondary references cure it directly: US 2005/0270270 A1 is expressly directed to "navigation key-based predictive text entry" on a hand-held communication device, and US 2005/0188330 A1 to predictive text input on a mobile communication device. A PHOSITA seeking to let a user pick a predicted next element with a 4-way key would have had every reason to place the top-ranked option at a directional position and to bind the key press to that position. Motivation: the entire purpose of a 4-way control is positional correspondence.

5. Dependent Claims 2–14

Because claims 2–14 all depend from claim 1, they fall with it; but each also has independent support in the cited art:

(Note: the Pu mapping for claims 2, 6–12 above mirrors the Board's analysis in the sibling IPR2020-00738 FWD concerning the same Pu reference, retrieved at http://bannerwitcoff.com/wp-content/uploads/2022/08/IPR2020-00738.pdf.)

6. Secondary Considerations (Objective Indicia)

Patent Owner relied heavily on objective indicia — commercial success of Samsung's accused products, industry praise in Samsung's internal documents, copying, unexpected results, failure of others, and a Nokia "Mobile Rules!" award — see the Patent Owner's Response and Sur-reply in IPR2020-00737 (https://www.docketalarm.com/cases/PTAB/IPR2020-00737/Samsung_Electronics_America_Inc._v._KANNUU_PTY_LTD/). Two points follow from the earlier-generated section:

1. The Board's final written decision nonetheless held claims 1–14 unpatentable, meaning the objective indicia were not sufficient to rebut the § 103 showing as the Board weighed the evidence. That outcome is itself strong corroboration that the combinations above supply a legally sufficient motivation to combine plus a reasonable expectation of success.
2. The indicia are also vulnerable on nexus. Kannuu itself tied the praise to the "highest ranked… up, down, left and right" placement corresponding to a TV remote's 4-way keys (Sur-reply ¶17) — i.e., the very limitation at issue — but the commercial-success and praise evidence tied to Samsung's "Predict Next Letter" feature also implicates features (ring-shaped prediction around a tapped letter) that are not coextensive with the claim. Where the objective evidence is not tied to the claimed combination, it cannot carry the day.

7. Bottom Line

Under pre-AIA § 103(a), claim 1 would have been obvious over (i) Perlman in view of Pu, (ii) Krohn in view of Pu (the instituted Ground 6), or (iii) either primary combination further in view of Siemens '270 / Griffin '330 for the directional-placement limitation and Masui/POBox for the "parts shorter than a complete word" limitation and the few-key motivation. Each combination pairs references from the same field addressing the identical problem — rapid selection of text-indexed items on devices with limited input — with a predictable result, and the page's own prior-art record (Apple '672; Tegic '541/'554; Microsoft '321/'965; Nokia '4423; the Besdi/Dinn device-interface publications; the iTAP paper) supplies redundant support for every individual element. Dependent claims 2–14 add only conventional implementation details (iteration, frequency/tree ranking, device types, and joystick/keypad selection) that the same art discloses or renders obvious. This is consistent with — and explains — the Board's FWD cancelling claims 1–14 and the subsequent IPR certificate.

8. Caveats and Explicit Uncertainties

• Krohn and Dostie identities unconfirmed. The IPR abbreviation table I retrieved defined only "Pu" (= US 7,152,213). I did not independently confirm the full bibliographic identities of the "Krohn" and "Dostie" references (I saw only a truncated excerpt of a deposition discussing a "character selection"/plurality-of-characters reference with a tree and a "hello"/car-model example, which is consistent with Krohn but not proof). I therefore reason from their role in the instituted grounds rather than asserting citations I could not verify. This should be treated as an open item, not a settled fact.
• FWD text not re-retrieved this session. I relied on the previously generated section's statement that the FWD held all of claims 1–14 unpatentable, corroborated by the docket metadata (FWD 21 Sept 2021; panel Droesch/Chung/Kaiser). I did not retrieve the FWD's own claim-by-claim reasoning in this session, so I cannot state precisely which ground the Board adopted for claim 1 (Perlman-based vs. Krohn+Pu-based).
• No 2026 appellate activity found (per the earlier section); the patent lapsed for fee non-payment effective 2024-09-06, and the claims stand cancelled by the IPR certificate. This analysis is therefore largely a retrospective reconstruction of the § 103 case the Board accepted.
• Minor date discrepancy: the task header states the current date as 26 April 2026, whereas the session timestamp is 29 September 2026. I have treated the analysis as date-independent (the operative legal date is the 12 August 2005 priority date), so the discrepancy does not affect any conclusion.
• Perlman element [1D]. Perlman's specific placement of the highest-ranked character in an up/down/left/right position is the one genuinely debatable point; my conclusion that claim 1 is obvious rests on reading Perlman's star-layout mapping together with the expressly directional art (Siemens '270, Griffin '330) and the natural design imperative to map the top-ranked candidate to a directional key — consistent with the Board's ultimate determination, but I flag it as the load-bearing inference.