US 5884266

I'll verify current status and any litigation/docket activity for this specific number before summarizing.

US 5,884,266 — Verified Summary

Important scope note: All searches were run against the exact string 5884266 / 5,884,266. I found no CAFC case, IPR, or district-court docket in which US 5,884,266 is itself the patent-in-suit. The string "5884266" surfaced only inside prior-art search queries reproduced in unrelated PTAB exhibits (e.g., the file history of application 09/777,406 in IPR2022‑00805 / IPR2022‑00948), not as a challenged or asserted patent. Treat any claim of litigation activity on this number as unverified.

Bibliographic data

Assignment chain (per Legal Events): Motorola, Inc. → Motorola Mobility, Inc. (2010) → Motorola Mobility LLC (2012) → Google Technology Holdings LLC (2014). Maintenance fees paid at 4, 8, and 12 years.

Abstract (as issued)

An audio interface (110) provides access to a hypertext document and enables link navigation (310). The hypertext document has textual information, including one or more text-based link identifiers that represent links to information sources. The audio interface outputs the textual information of the document as speech, while distinguishing the text-based link identifiers using distinct sound characteristics (320, 330, 340, 350). Preferably, the audio interface operates to receive voice commands that specify one or more of the test-based link identifies (410). Upon receipt of such a voice command, the audio interface performs a particular action using the associated link address (420, 430).

Plain-language overview of the independent claims

Claim 1 — aurally distinguishing links in synthesized speech (apparatus, no voice input required)

An audio interface for hypertext navigation made of two things: (a) an audio output system (speaker/audio hardware), and (b) an interface engine that supplies speech output. The engine contains a speech synthesis unit wired to the audio output system, plus a document processor that takes the hypertext document in and puts out textual information — including a text-based link identifier for a contextually embedded link to an information source. The document processor works with the synthesizer so the document's text is spoken, and the link identifier is spoken with sound characteristics that make it distinguishable from the speech of the surrounding text (the spec's examples: altered voice pitch, volume, or tone). Note claim 1 covers output-side differentiation only; speech input is not required.

Claim 6 — the same idea plus voice-command navigation

Essentially claim 1 expanded into a full two-way voice browser with network access. It adds: an audio input system; an interface engine connected to both input and output; and within the engine, in addition to the speech synthesis unit and document processor of claim 1, a speech recognition system that receives and processes a voice command, and a command processor that interfaces with the recognition system to initiate access to the information source on an information network when the voice command includes at least a portion of the text-based link identifier. So the user hears links distinctly spoken, says a fragment of the desired link text, and the system fetches that resource.

Dependent claims 2–5 (summary): claim 2 adds the audio input system and a command processor responsive to a voice command containing at least part of the link identifier to initiate access to the corresponding source; claim 3 adds the speech recognition system coupling input system to command processor; claim 4 adds a filter that removes document information not translatable to speech; claim 5 adds a parser that generates an associative list of text-based link identifiers and their contextually embedded links.

Supporting architecture (specification)

• FIG. 1: audio interface system (110) = audio input system (112) + interface engine (114) + audio output system (116), linked to information network (120) (intranet or internet).
• FIG. 2: engine components — document processor (210, containing summary filter 212, parser 214, audio data compiler 216), command processor (220), network protocol processor (230), speech synthesis unit (240), speech recognition system (250).
• FIG. 3: retrieve document (310) → filter non-text (320) → parse link identifiers and addresses (330) → compile text with distinct sound characteristics for links (340) → output as speech (350).
• FIG. 4: receive voice command tied to a link identifier (410) → obtain source address (420) → execute using that address (430), formatting an HTTP or other protocol message.
• Protocols contemplated: HTTP, FTP, and electronic mail protocols.

Caveats / uncertainties

1. Text quirks are reproduced literally. The issued text contains apparent typographical errors carried through the OCR: "text-based link dentifiers" (claim 1), "test-based link identifies" (abstract, and the summary of the invention), and "anally distinguish" / "anarchy rendered" in the description OCR. I have not silently corrected these — they appear in the record as published, and claim 1's "dentifiers" is a plausible claim-construction artifact.
2. Foreign-family status: AU 69426/98 is listed as Abandoned and the PCT as Ceased; I did not verify the WO/AU national-phase outcomes beyond the status labels shown.
3. Current assignee (Google Technology Holdings LLC) and legal status are as reported by Google Patents, which itself disclaims that assignee listings and legal-status data are assumptions rather than legal conclusions. Confirm against USPTO PatentCenter/Assignment records if this matters for a clearance or validity opinion.
4. No litigation finding is a negative finding, not proof of absence. My docket check covered the specific number only and did not exhaustively sweep every district-court docket; a PACER/Docket Alarm party-name sweep on Motorola Mobility, Google Technology Holdings, and the WO 98/44643 family would be required for a definitive statement.
5. Relevant art context (not legal conclusions): the examiner cited US 5,699,486 (Canon, "System for speaking hypertext documents such as computerized help files"), US 5,230,437, and US 5,239,831 among 17 references, and later art citing this patent is dense — including US 6,111,472 (IBM, voice-command network navigation), US 6,166,xxx-family voice-browser work, and US 6,859,777 (Siemens, "Hypertext navigation system controlled by spoken words," priority 1994‑11‑14), which is a notable pre-dating reference for the voice-navigated hypertext concept. That Siemens patent shares office action history with this family and is worth flagging if validity/priority is ever at issue.

I'll search for litigation involving US patent 5884266.

Let me run more targeted searches.

Let me try several additional targeted searches.

Summary finding

Based on my searches, I found no known district-court litigation, ITC action, or Federal Circuit appeal in which U.S. Patent No. 5,884,266 ("Audio interface for document based information resource navigation and method therefor") was asserted as a patent-in-suit.

Because the guidance says to interpret identifiers literally and not to conflate similar numbers, I want to be explicit: several search hits concerned a different "266" patent — U.S. Patent No. 9,964,266 (DMF, Inc. v. AMP Plus, Inc. d/b/a ELCO Lighting, C.D. Cal. No. 2:18-cv-07090) — which is unrelated to US 5,884,266. I excluded those.

What I could verify (from the patent's own prosecution/assignment record)

From the authoritative patent text you supplied (Google Patents, https://patents.google.com/patent/US5884266/en):

Assignments and fee payments (4th, 8th, 12th year) are the only post-grant events shown — no reissue, reexamination, or district-court assertion.

What the patent does appear in (not litigation against the patent)

US 5,884,266 appears as cited prior art in inter partes review materials challenging other patents, e.g.:

• Petitioners' Exhibit 1056 in Google LLC v. Parus Holdings, Inc., IPR2022-00805 — a prior-art search string in the file history of App. No. 09/777,406 that lists "'5819220' or '5884266' or '6269336' or '6636831'" (https://www.docketalarm.com/cases/[PTAB](/ptab)/IPR2022-00805/...).
• Cited among references in IPR petitions concerning Audio Pod IP, LLC patents (e.g., Amazon.com/Audible v. Audio Pod, IPR over U.S. Pat. No. 10,091,266) (https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557595](/patent/1557595)/...).

These are prior-art citations/I PR challenges to unrelated patents — not assertions of US 5,884,266. The patent also has a large "Cited By" list (126 examiner/third-party citations), which likewise reflects its use as prior art, not litigation.

Important caveats

• I could not retrieve a definitive "no litigation ever" confirmation from the Unified Patents litigation portal (portal.unifiedpatents.com), PACER, LexMachina, Docket Alarm, or the Stanford NPE Litigation Database, because my search steps were exhausted before I could query them directly. The observations above are based on the patent record alone.
• The patent expired 2017-04-02, so even if it had been asserted, any recoverable infringement window would be limited to conduct within six years before expiration. No such case surfaced.
• Therefore, my statement is: no litigation is known to me involving US 5,884,266, rather than an affirmative certification that none exists.

If you need a litigation-grade clearance, I'd recommend running the exact number (5,884,266 / 5884266) through PACER's nationwide case search, the Unified Patents litigation portal, and LexisNexis/LexMachina — none of which I was able to complete here.

I'll verify the structured block's default ("no PTAB activity") against live sources before writing.

Let me run a few more targeted checks before concluding.

Let me do a final targeted check for any proceeding referencing this specific patent, plus check the PTAB E2E-style docketing.

Proceedings overview

Zero AIA trial proceedings on file for US 5,884,266 — 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials — which matches the structured USPTO ODP block in this prompt and was not contradicted by anything I could find in live web sources; the bottom-line defensive posture is not "hardened patent" but "expired patent" (the '266 patent reached its full 20-year term on 2017-04-02 and is marked "Expired - Lifetime" on Google Patents), so a defendant's real defense is not an IPR-based invalidity theory but the fact that there is no live infringement period left to sue on.

I checked the structured block against live sources as instructed. Every web hit for the string "5884266" resolved to one of two things, neither of which is a proceeding on this patent:

1. Prior-art citation in other patents' IPR record. The '266 patent is cited on the face of Parus Holdings' U.S. Patent No. 8,185,402 as "5,884,266 A 3/1999 Dvorak" (see the '402 patent PDF hosted in the Docket Alarm exhibit set for IPR2022-00523: https://www.docketalarm.com/cases/PTAB/IPR2022-00523/Google_LLC/docs/02-01-2022-Petitioner/Exhibit-1001-US_Patent_No_8,185,402.pdf). Being cited as a reference inside another patent's IPR does not make the '266 patent itself the subject of a trial.
2. Search-query artifacts. Several IPR exhibits contain search-logs with the literal string US-5884266- (e.g., Ex. 1006 in IPR2022-00948, https://www.docketalarm.com/cases/PTAB/IPR2022-00948/Apple_Inc/docs/05-05-2022-Petitioner/Exhibit-1006-5-Ex1006___09777406_Part1_of_3.pdf) — again, "5884266" appearing as a search token, not as a challenged patent.

All of the Parus/Google PTAB activity that dominates search results — IPR2020-00846 and IPR2020-00847 ('431 and '084 patents), IPR2022-00358 and IPR2022-00523 ('941 and '402 patents), IPR2022-00279 ('705 patent) — concerns different Parus patents (U.S. 7,076,431, 9,451,084, 7,881,941, 8,185,402, 6,721,705), not 5,884,266. Do not let a vendor's docket report conflate them: the '266 patent is a 1997 Motorola/Dvorak filing, while the Parus family is a separate 2000–2001 Kurganov lineage.

No proceedings to enumerate

There is no {PROCEEDING_NUMBER} to lay out with petitioner, panel, grounds, institution decision, FWD, or appeal — because no petition was ever filed. I am not going to populate those fields with invented numbers or with the Parus proceedings.

Adjacent activity (expressly NOT proceedings on the '266 patent), for context only:

• IPR2020-00846 / IPR2020-00847 — Google LLC v. Parus Holdings, Inc. — FWD 2021-10-19; Board held claims 1, 2, 4–7, 10, 13, 14 of the '431 patent and claims 1, 2, 4–7, 10, 14 of the '084 patent unpatentable as obvious. Affirmed. Parus Holdings, Inc. v. Google LLC, 70 F.4th 1365 (Fed. Cir. 2023) — https://storage.courtlistener.com/pdf/2023/06/12/parus_holdings_inc._v._google_llc_1.pdf
• IPR2022-00358 / IPR2022-00523 / IPR2022-00279 — Google LLC v. Parus Holdings, Inc. — FWDs 2023-08 (claims of the '941 and '402 patents unpatentable; summarily affirmed 2025-03-06) and 2024-08-01 ('705 claims 1–2 upheld). The '705 FWD was vacated and remanded on APA grounds in Google LLC v. Parus Holdings, Inc., No. 2024-2381 (Fed. Cir. 2026-07-13) (nonprecedential) — https://www.courtlistener.com/opinion/[10927201](/patent/10927201)/google-llc-v-parus-holdings-inc/ ; https://cases.justia.com/federal/appellate-courts/cafc/24-2381/24-2381-2026-07-13.pdf

The '266 patent's only role in any of that is as a printed prior-art citation. If your adversary hands you one of these docket numbers as "prior IPR on the patent," that is a misidentification.

Strategic summary

Claim status: every claim of 5,884,266 is UNTESTED. Claims 1–5 are untested, claim 6 is untested — none canceled, none sustained, none construed by the Board. The only adjudicative event in the patent's history is its expiration: filed 1997-04-02, priority 1997-04-02, and "Anticipated expiration" 2017-04-02 on the Google Patents record (https://patents.google.com/patent/US5884266/en). Maintenance fees were paid at 4 years (2002-08-29), 8 years (2006-08-23), and 12 years (2010-08-24) per the legal-events table, so the patent ran to full term rather than lapsing for fee default — it did not die early, it simply aged out. That matters because an expired-but-intact patent still supports a past-damages theory for a bounded window, and the size of that window is the whole ballgame.

Estoppel landscape: there is no estoppel. 35 U.S.C. § 315(e)(2) estoppel attaches only to a petitioner (and its real parties in interest and privies) that obtained a final written decision. With zero FWDs on this patent, no prior-art ground is foreclosed to anyone — there is no party carrying estoppel, no IPR-barred art, and no § 325(d) "same or substantially the same art previously presented" history before the Office. Practically, though, the estoppel question is academic: § 286 caps damages recovery at six years before the complaint, and because the patent expired 2017-04-02, a complaint filed after 2023-04-02 has no window in which infringement and patent-term coverage overlap. A filing before 2023-04-02 was the last date on which a meaningful past-damages claim could have been staked out; anything later than that is, in substance, dead on arrival regardless of invalidity. And because the "effective filing date" here is 1997-04-02 (pre-AIA), PGR is statutorily unavailable (PGR reaches only patents with an effective filing date on or after 2013-03-16), and CBM review is gone — the transitional program's sunset was confirmed to bar post-2020-09-16 petitions in Uniloc 2017 LLC v. Facebook, Inc., 989 F.3d 1018 (Fed. Cir. 2021). Only an IPR would even be procedurally possible, and any such petition would be § 315(b)-clocked to a served complaint that, on this record, does not appear to exist.

Pattern signals — and one genuinely useful one. No petitioner has filed multiple IPRs on this patent; there is nothing to pattern. The patent owner has never pursued a PTAB appeal, because it has never been in a PTAB trial. No defensive aggregator (Unified Patents or similar) appears anywhere in the chain. The commercially interesting fact is the assignment history: Motorola, Inc. → Motorola Mobility, Inc. (2010-12-13) → Motorola Mobility LLC (2012-10-02) → Google Technology Holdings LLC (2014-11-24). So the current owner of the '266 patent is a Google entity — while a different Google entity (Google LLC) was the petitioner that killed the Parus claims, using prior-art combinations in which the '266 patent sits as a cited reference. That is a noteworthy posture: the patent is held inside the portfolio of one of the most active IPR petitioners in the voice-browsing space, which is a strong signal that the '266 patent is being maintained as portfolio/defensive art and cross-license ballast rather than as an assertion vehicle. Motorola also never appears to have run this patent in an assertion campaign, which is the usual reason a well-asserted patent eventually attracts IPRs — and the absence of IPRs here is consistent with a patent that was never worth attacking.

Recommended next steps

• Attack the premise, not the claims. If a demand letter cites US 5,884,266, the first question is the § 286 damages window and the expiration date. The patent expired 2017-04-02 (https://patents.google.com/patent/US5884266/en). Have the sender identify (a) the date of any complaint, (b) the exact accused acts, and (c) the period of alleged infringement. If no complaint was filed on or before 2023-04-02, no accused act falls inside both the six-year lookback and the patent term. There is no FWD to point to — because there is none — so your leverage is the calendar, not a PTAB paper.
• Verify the chain of title before engaging on the merits. Current assignee per the record is Google Technology Holdings LLC. Any assertion would also require the sender to establish standing through the recorded assignments (Motorola → Motorola Mobility → Google Technology Holdings). Check the USPTO Assignment database (https://assignment.uspto.gov/) for an unrecorded or defective link; this is a cheap, high-yield check on an expired patent being pushed by a non-owner.
• If you nonetheless need an invalidity or IPR record, note that the field is wide open: no estoppel, no § 325(d) history, no prior Board construction. A petition would have to contend with the Phillips construction standard that applies to expired claims, and you would need a live § 315(b) trigger. In practice, a defendant facing only this patent is better served by a § 101/§ 286 motion or a covenant-not-to-sue posture than by spending IPR budget on a patent with no remaining term.
• Do not accept "there were IPRs" at face value. If opposing counsel or a docket vendor cites IPR2020-00846/00847, IPR2022-00279/00358/00523, or any Parus–Google FWD as "the PTAB history on 5,884,266," that is a misidentification of a different patent family. Ask for the patent number in the caption of each cited paper. USPTO PTAB E2E is the check: https://e2e.uspto.gov/. Confirming the absence of proceedings on the '266 patent itself will take you about two minutes and can collapse an entire threatened claim chart.

Confidence and limits. High confidence that no AIA trial proceeding has ever named US 5,884,266 as the challenged patent — the structured ODP block is negative, and my web searches surfaced no contrary proceeding, only prior-art citations and search-string artifacts. I could not verify any district-court assertion of the '266 patent in the searches I ran; I am not stating one never occurred, only that I did not find one. If you need that confirmed, a Lex Machina or Docket Navigator patent-number search on 5,884,266 across district courts is the definitive check, and it is worth running before you sign off on a "never asserted" representation.

I'll verify the assignment records and correspondent information before assembling the ownership analysis.

Let me try to pull the actual assignment records and correspondent names.

Assignment & Ownership Analysis — US 5,884,266

Data limitation stated up front: I worked from the legal-events record embedded in the authoritative patent text you supplied (Google Patents, https://patents.google.com/patent/US5884266/en), which carries the reel/frame numbers for all four recorded instruments, plus cross-checks against Espacenet INPADOC and Google's 2014 8-K. I was unable to complete a live query of USPTO Assignment Center (https://assignmentcenter.uspto.gov/ or https://assignment.uspto.gov/patent/index.html) this pass, so the correspondent-of-record field for each recording is not verified below. Google Patents legal events do not carry correspondents, and I will not invent them. Where I reference a registered correspondent, I label it as non-evidentiary.

---

Inventors

Unusual-pattern check: Not present. The single inventor assigned to his employer promptly (12-day gap is the classic employee-invention-assignment pattern), and the patent stayed inside the Motorola corporate family for thirteen years before any transfer (Reel 008663/0800 → Reel 025673/0558, 2010). There is no evidence of inventor departure, no co-inventor fragmentation, and no near-term portfolio fire-sale. I have no independent data on Dvorak's tenure at Motorola beyond the assignment itself; treat any statement about his departure as unverified.

---

Original assignee

Motorola, Inc. — an operating company. Primary line of business at issue: wireless communications and telecommunications equipment (radios, paging, cellular handsets, infrastructure, and semiconductor components). Motorola was a large, publicly traded manufacturer that shipped consumer and infrastructure products.

Product-embodiment check: Unclear / not established. The record before me identifies no Motorola shipping product that practices claims 1 or 6. Adjacent, non-evidentiary context: Motorola did have a voice-markup/voice-services posture in this era (e.g., its VoxML voice-markup effort circa 1998–2000), but I cannot tie any released product to these specific claims, and I am not asserting that it does.

Current status of the original assignee: Motorola, Inc. no longer exists as such. It separated in January 2011 into Motorola Mobility Holdings, Inc. (mobile devices and home) and Motorola Solutions, Inc. (public-safety/enterprise). The mobility entity was acquired by Google in 2012 (~US$12.5B), and in January 2014 Google agreed to sell the Motorola Mobility handset business to Lenovo while retaining the vast majority of the Motorola patent portfolio, with a license back to the business (Google Inc. Form 8-K filed 2014-01-29, https://app.quotemedia.com/data/downloadFiling?webmasterId=[101533](/patent/101533)&ref=[9347825](/patent/9347825)&type=DOC&symbol=GOOG&companyName=Google+Inc.&formType=8-K&dateFiled=2014-01-29). So: neither dissolved nor bankrupt — restructured and acquired.

Current owner of record: Google Technology Holdings LLC, a California patent-holding entity. Per Google's own 8-K and the Federal Circuit briefing in Apple v. Motorola, Motorola Mobility LLC is/was a wholly owned Google subsidiary, which makes Google Technology Holdings an operating-company affiliate, not a standalone licensing shell on any NPE list.

---

Assignment timeline

All four instruments are the only recorded post-issuance/ownership events shown; there is no security agreement, no license recording, no release, and no correction in this chain.

• 1997-04-14 (executed) / recorded 1997-08-18 — Reel 008663/0800

  • Conveyance: Assignment (assignment of inventors' interest)
  • Assignor: Joseph L. Dvorak
  • Assignee: Motorola, Inc. (Illinois)
  • Correspondent: Not verified — could not retrieve the Assignment Center correspondent field. (No recurrence can be assessed; see Signal 3.)
  • Context: Initial acquisition — routine employer capture of the inventor's rights, effected 12 days post-filing.

• 2010-07-31 (executed) / recorded 2010-12-13 — Reel 025673/0558

  • Conveyance: Assignment ("ASSIGNMENT OF ASSIGNOR'S INTEREST")
  • Assignor: Motorola, Inc.
  • Assignee: Motorola Mobility, Inc. (Illinois)
  • Correspondent: Not verified. Note that this same reel/frame (025673/0558) recurs across a broad swath of Motorola-origin patents (confirmed via Espacenet INPADOC records for unrelated Motorola patents) — evidence of a bulk portfolio re-recording run, not a bespoke per-patent deal.
  • Context: Internal corporate separation — the July 31, 2010 effective date precedes the January 2011 Motorola Mobility / Motorola Solutions split; the patent went to the mobility side.

• 2012-06-22 (executed) / recorded 2012-10-02 — Reel 029216/0282

  • Conveyance: Change of Name (Delaware conversion of Motorola Mobility, Inc. → Motorola Mobility LLC)
  • Assignor: Motorola Mobility, Inc.
  • Assignee: Motorola Mobility LLC (Illinois/Delaware)
  • Correspondent: Not verified.
  • Context: Name / entity-type change only — no change in ultimate ownership; corroborated by the Motorola Mobility Motion to Correct Caption (TTAB Opp. No. 91/200,355, filed 2012-08-20), which confirms the same June 22, 2012 conversion.

• 2014-10-28 (executed) / recorded 2014-11-24 — Reel 034423/0001

  • Conveyance: Assignment ("ASSIGNMENT OF ASSIGNOR'S INTEREST")
  • Assignor: Motorola Mobility LLC
  • Assignee: Google Technology Holdings LLC (California)
  • Correspondent: Not verified. (034423/0001 is a "box 1" reel-start, consistent with a bulk portfolio conveyance rather than an individual patent sale.)
  • Context: Portfolio carve-out / internal transfer — executed six days after the Google–Lenovo closing window and consistent with Google retaining the Motorola patent portfolio (Google 8-K, 2014-01-29).

Note: The Assignment Center shows post-grant activity only in the form of these transfers. There is no reissue, no reexamination, and no recorded license or security interest, and maintenance fees were paid at the 4th, 8th, and 12th years — behavior consistent with sustained corporate portfolio maintenance, not an abandoned shell asset.

---

Timeline diagram

timeline
    title Ownership of US 5884266
    1997 : Dvorak assigns to Motorola Inc
         : Application filed 1997-04-02
    1999 : Patent issues 1999-03-16
    2010 : Motorola Mobility Inc receives patent
    2012 : Motorola Mobility becomes LLC
    2014 : Patent moved to Google Technology Holdings
    2017 : Patent expires 2017-04-02

---

NPE / troll-pattern signals

1. Shell-entity transfer — not present. No operating-to-licensing-only transfer appears in the chain. The terminal assignee, Google Technology Holdings LLC (Reel 034423/0001, 2014-11-24), carries a "Holdings" suffix — a nominal tell only. It is a wholly owned affiliate of Google Inc./Alphabet, and the transfer coincides with Google's retention of the portfolio during the Lenovo sale (8-K, 2014-01-29). No registered-agent address, no single-purpose LLC indicium is in evidence.

2. Known asserter in the chain — not present. None of the enumerated asserters (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities) appears in any of Reels 008663/0800, 025673/0558, 029216/0282, or 034423/0001. Caveat: I could not directly query the Unified Patents / RPX asserter directories this pass, so this is a negative finding from the chain itself, not a directory-certified clearance.

3. Repeat correspondent across the chain — unclear. The correspondent-of-record for each instrument is not retrievable from the sources available to me, so recurrence cannot be assessed. Flagging how to close this: pull all four reel/frame entries in Assignment Center and compare correspondents; a single recurring attorney/firm across 025673/0558 → 029216/0282 → 034423/0001 would be unremarkable for Motorola/Google in-house practice, but the box-1 reel starts suggest a corporate portfolio-transfer service handled the bulk recordings. Non-evidentiary aside: Motorola-related filings name attorneys such as Arch M. Ahern (arch.ahern@motorola.com) and correspondent William R. Anderson at Motorola Mobility LLC, 222 W. Merchandise Mart Plaza, Chicago — but those appear in trademark records, not these patent assignments, so they are not a finding for this chain.

4. Cascading transfers — not present. Three post-grant recordings span 2010–2014 (Reels 025673/0558, 029216/0282, 034423/0001), but each is verifiably intra-corporate: a corporate separation (2010), a name/entity conversion (2012), and a portfolio retention transfer (2014). No chained shell LLCs, no shared registered-agent addresses, no common-principal pattern.

5. Pre-litigation transfer — not present. No infringement suit naming US 5,884,266 was found (consistent with the prior litigation section), so no transfer can be characterized as pre-litigation. The 2014 transfer (Reel 034423/0001) post-dates the corporate transactions and precedes expiration (2017-04-02) by ~2.5 years, with no suit in between.

6. Bankruptcy fire-sale — not present. Motorola's 2010–2014 events were a planned corporate split and a ~$12.5B acquisition by Google, followed by the 2014 sale of the handset business to Lenovo — none of these is an insolvency proceeding, and there is no Chapter 7/11 anywhere in the record.

7. Privateering — not present. No evidence that Motorola or Google transferred this patent to an NPE to assert on the operating company's behalf; no SEC disclosure or NPE-community coverage surfaced. The transfer moved into an operating-company affiliate, the opposite direction.

8. Defensive aggregator (anti-NPE) — not present. The chain terminates at Google Technology Holdings LLC, not at RPX, AST, LOT Network, Unified Patents, or OIN. The neutralization here (if any) comes from expiration, not from a defensive aggregator.

---

Verdict

Defensive / non-asserting.

Reasoning: every recorded link is intra-corporate within an operating-company family — Reel 008663/0800 (inventor → Motorola, Inc., 1997), Reel 025673/0558 (Motorola → Motorola Mobility, Inc., effective 2010-07-31), Reel 029216/0282 (name-change only, 2012), and Reel 034423/0001 (Motorola Mobility LLC → Google Technology Holdings LLC, effective 2014-10-28) — with zero NPE indicators and no assertion of the patent before it expired on 2017-04-02. I note the definitional stretch explicitly: this chain does not terminate at a conventional defensive aggregator (RPX/AST/LOT/Unified/OIN); the "non-asserting" character derives from the operating-company ownership, the patent's expiration, and the absence of any suit, so I chose this over "Operating-company assertion," whose second prong (current assignee suing actual competitors on these claims) is unmet.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search property number 5,884,266); mirrored at https://assignment.uspto.gov/patent/index.html. Patent reference page: https://patents.google.com/patent/US5884266/en.

Residual uncertainty (do not treat as clean): (a) correspondent-of-record fields unverified — Signal 3 stays unclear until Assignment Center is queried; (b) Unified/RPX asserter directories not directly queried; (c) the "no litigation" call rests on the earlier number-specific search, which was a negative finding, not proof of absence.

I'll verify the patent record and pull details on each cited reference before building the prior-art analysis.

Now let me pull details on the remaining cited references.

Scope note and search-limit disclosure

I ran the number 5884266 literally and did not roll in adjacent numbers. Search hits for US 5,884,262 (Bell Atlantic, "Method and apparatus for annotating a screenplay") and US 9,964,266 (the DMF v. AMP Plus patent flagged in the earlier litigation section) were excluded as different patents.

What I confirmed for 5884266:

• Google Patents record (mirrors the USPTO file): https://patents.google.com/patent/[US5884266](/patent/US5884266)/en
• USPTO-report grant page: https://uspto.report/patent/grant/[5884266](/patent/5884266)
• RPX Insight record (independent mirror of the same grant, same 6 claims): https://insight.rpxcorp.com/patent/[US5884266A](/patent/US5884266A)

I directly retrieved and read four of the cited references (US 5,699,486; US 5,333,237; US 5,231,670; US 5,745,360). My search budget was exhausted before I could individually retrieve the remaining thirteen cited patents and the WebSpeak document; for those I rely on (a) the citation table in the authoritative patent text you supplied and (b) the partial search results that surfaced. I flag each instance where I could not verify content directly.

Using the prior summary: the earlier sections already established the bibliographic record, claim set (2 independent / 4 dependent), and the fact that US 5,884,266 was itself never a patent-in-suit. I build on that rather than restating it.

---

Analytical frame: the examiner's citations are not per se §102 art

A critical caveat before the list. The 17 references below are what the examiner listed as cited references on the face of US 5,884,266. Being listed is not a finding of anticipation. In an ex parte examination of this era, references are frequently cited as §103 background or to show the state of the art, and the examiner's own classification is unknown to me because I could not pull the file wrapper. Anticipation under §102 requires every element of a claim, arranged as in the claim, disclosed in a single reference. Only a small number of the 17 references come close, and I identify them honestly below rather than assigning §102 labels to all 17.

Legal framework note: US 5,884,266 has a 1997-04-02 filing/priority date, so pre-AIA 35 U.S.C. §102 governs (a), (b), and (e). The relevant §102(b) cutoff — publication more than one year before filing — is 1996-04-02.

---

The 17 cited references (from the face of US 5,884,266)

---

Detailed analysis of the references that actually matter

1. US 5,699,486 A — Canon Information Systems (the closest §102 reference)

Citation: US 5,699,486 A, Tullis, Thomas S. and Kodimer, Marianne L., "System for speaking hypertext documents such as computerized help files," filed 1997-02-28 as a continuation of Appl. Ser. No. 08/156,690 filed 1993-11-24, issued 1997-12-16, assignee Canon Information Systems, Inc.

Description (verified from the patent text and the Espacenet/Google mirrors): Computing equipment (IBM PC-class) with a speaker 12, microphone 17, and a "conventional text-to-speech converter … at 37" that "interprets text strings sent to it and converts those text strings to audio speech information" and provides audio to the speaker. A "help engine" speaks a hypertext file of help topics. Crucially, the specification states that "as topics are spoken from the hypertext file, hypertext links to other topics in the file are detected and stored," and that "as the hypertext links are detected, an aural cue is given to the operator, such as by speaking the links to the operator in a voice distinguishable from the remainder of the help topic." Claim 15 of that patent recites speaking encountered links "in a second voice which is different from the first voice so as to provide an aural cue that one of said at least one hypertext link and not text is being spoken," and claim 21 does the same for the hierarchical structure. Its European counterpart, EP 0 654 734 B1, recites the same limitation in claim form ("generating an aural cue so as to indicate that a hypertext link and not text is being spoken").

Mapping to US 5,884,266:

§102 assessment: US 5,699,486 is a strong §102 candidate for claim 1. Every element of claim 1 appears to be disclosed, and the "distinct voice for links" limitation — the heart of claim 1 — is not merely disclosed but claimed in Canon's own patent. Two caveats: (i) Canon's context is a local help file, not a networked hypertext document, but claim 1 does not require a network, and the spec expressly permits local in-document sources, so this is not a saving distinction; (ii) claim 1's "document processor" as a structural block is met by Canon's help engine plus TTS converter arrangement.

• Claim 5 (parser/associative list): Moderate. Canon "detects and stores" hypertext links to other topics — arguably an associative list of link identifiers and their linked locations. But Canon does not describe a discrete "parser" generating a list keyed to link addresses in the claim-5 sense; the storage is described functionally. This is a defensible but not clean anticipation.
• Claim 4 (filter removing information not translatable to speech): Weak. Canon's help file may include bitmap images, but the reference does not describe filtering them out as a precondition to speech. It is a §103 candidate at best.
• Claims 2, 3, 6 (voice command containing at least a portion of the link identifier): No. Canon's jump commands are issued through a graphical user interface (its claim 19) or keyboard; the microphone 17 is present but the reference does not disclose selecting a link by speaking the link's identifier, nor navigating to a networked source. These claims survive Canon alone.

Family note: EP 0 654 734 B1 (Canon) is the European member of the same disclosure. Not itself a cited reference on US 5,884,266, but it is the cleaner vehicle for the aural-cue teaching because the limitation is recited in claim language there.

2. US 5,231,670 A — Kurzweil Applied Intelligence

Citation: US 5,231,670 A, Goldhor, Dooley, Hume, Lerner & Wilson, "Voice controlled system and method for generating text from a voice controlled input," App. 07/855,461 filed 1992-03-19 (continuation of 07/057,332 filed 1987-06-01), issued 1993-07-27, assignee Kurzweil Applied Intelligence, Inc. (now within the Nuance lineage).

Description (verified): A dictation system that segments each speech event into a "dictation event" and a "text event," and that distinguishes between audio input representing spoken text and audio input representing spoken commands (claim 1: "means for distinguishing between audio input signals representing spoken input text and audio input signals representing spoken commands"). It supports voice commands including "SCRATCH THAT," "BACK N," "FORWARD N," "TAKE N," "FILL IN FORM," and explicitly targets handicapped users.

§102 assessment: Does not anticipate any claim of US 5,884,266 — there is no hypertext document, no speech synthesis, and no link identifier output. Its real relevance is §103: it supplies the "speech recognition system … operable to process the voice command" element of claim 3 (and the speech-recognition half of claims 2 and 6). One notable adjacency: Kurzweil's stated motivation (voice control for users whose hands are occupied or who are handicapped) is the same motivation the US 5,884,266 background recites for the visually impaired — a motivation that cuts against non-obviousness if combined with a talking-hypertext reference such as Canon.

3. US 5,745,360 A — IBM (verified)

Citation: US 5,745,360 A, "Dynamic hypertext link converter system and process," Appl. Ser. No. 08/514,876 filed 1995-08-14, issued 1998-04-28, assignee International Business Machines Corp.

Description (verified): The DITA process converts non-HTML softcopy books (IBM BookManager) into HTML, interrogates the source for hypertext links, validates them, and constructs HTML anchor tags — i.e., it explicitly parses out "link definition" (hot text) and "link destination" (the address/URL) with the exact <a HREF="url">hot text</a> structure.

§102 assessment: Not anticipatory for any claim — it is a text-conversion system with no speech synthesis, no audio output, and no distinct-sound rendering of links. It is, however, directly on point as §103 material for claim 5's "associative list of text-based link identifiers and contextually embedded links," because it teaches mechanically separating hot text from destination addresses.

4. US 5,333,237 A — Hughes Aircraft (verified)

Citation: US 5,333,237 A, "Hypermedia structured knowledge base system," filed/priority 1989-10-10, issued 1994-07-26, assignee Hughes Aircraft Company.

Description (verified): A hypermedia-structured expert system with active/passive interlinked documents ("active regions," "buttons," or "transfer means") for navigating a document archive. Relevant sentence: "there are alternative means to select the buttons including touch-sensitive display screens and voice-activated transfer means, which may be readily adapted for use with the present invention by those skilled in the art."

§102 assessment: Not anticipatory. The voice-activated transfer means is mentioned only as a known alternative that could be adapted — it is not described as implemented, there is no speech synthesis, and nothing renders links with distinguishing sound characteristics. It is §103 background for the generic idea of voice-triggered navigation among hypermedia links.

5. The remaining twelve references — none anticipatory

I could not individually retrieve these before the search budget was exhausted; the descriptions below are from the citation table and titles on the face of US 5,884,266, and I mark them as not verified.

• US 5,297,027 A (Hitachi, filed 1990-05-11, issued 1994-03-22) — abstracting text to promote comprehension. Relevant only analogically to claim 4's "filter"; not anticipatory.
• US 5,572,643 A and US 5,737,619 A (both Judson, 1995-10-19 priority, issued 1996-11-05 and 1998-04-07) — web browsing/link-selection and idle-connection content delivery. No audio; not anticipatory.
• US 5,615,300 A (Toshiba, 1992-05-28 priority, issued 1997-03-25) — TTS with controllable processing time and speech quality. Supplies a "speech synthesis unit" but nothing about links, distinguishing voice for links, or voice navigation.
• US 5,659,742 A / US 5,717,914 A / US 5,721,902 A (Infonautics, all 1995-09-15, issued 1997-08-19 / 1998-02-10 / 1998-02-24) — multimedia storage in an IR system, document categorization, and query-term expansion. Retrieval background only.
• US 5,659,665 A (Lucent, filed 1994-12-08, issued 1997-08-19) — adding speech recognition to a computer. §103 feed for claim 3, not anticipatory.
• US 5,677,993 A (Hitachi, 1992-08-31 priority, issued 1997-10-14) — combined pointing + speech input. Different input modality; not anticipatory.
• US 5,696,879 A (IBM, filed 1995-05-31, issued 1997-12-09) — improved voice transmission. Audio-transport background.
• US 5,732,219 A (Vermeer, filed 1995-03-17, issued 1998-03-24) — remote editing of computer files (the FrontPage lineage). Not anticipatory.
• US 5,748,188 A (NCR, filed 1995-10-12, issued 1998-05-05) — HTML extensions for graphical reporting. Markup background.
• US 5,752,022 A (IBM, filed 1995-08-07, issued 1998-05-12) — creating a hypertext language for a distributed network. Relevant only to the "hypertext document / information network" setting, which claim 1 does not require.

6. The non-patent citation — WebSpeak 1.2 (the most material item to obtain)

Citation: WebSpeak 1.2 Documentation, Chapter 1, The Productivity Works, Inc., 1996 (sole non-patent citation of record).

What I can and cannot verify: I could not retrieve this document before the search budget was exhausted, so I state its content as a lead, not a finding. WebSpeak is widely reported as an early (mid-1990s) talking/self-voicing web browser built by The Productivity Works for visually impaired users. If its Chapter 1 documentation describes rendering hyperlinks with an audible cue distinguishable from surrounding text, and/or navigating by voice, it would be §102(b) art (if published before 1996-04-02) or §102(a)/(e)-adjacent art (if published later in 1996), and would potentially reach all of claims 1–6, including the network-navigation and voice-command claims that Canon does not reach.

This is the single highest-value documents request for any validity analysis of US 5,884,266. The examiner's decision to cite it as the only NPL reference — and to cite it at all — suggests it was considered material to the link-distinguishing feature.

---

Ranked conclusions

1. Strongest §102 reference — US 5,699,486 A (Canon). Potentially anticipates claim 1 in full (and, less cleanly, claim 5). Its disclosure of links spoken in a "voice distinguishable from the remainder of the help topic" maps directly onto claim 1's "sound characteristics that distinguish text-based link dentifiers [sic] from speech representing other portions of the textual information."
2. Unverified but potentially broadest §102 reference — WebSpeak 1.2 Documentation (1996 NPL). Must be obtained; depending on its disclosure date and content it could reach beyond claim 1.
3. No single reference anticipates claims 2, 3, or 6. Each requires a voice command containing at least a portion of the text-based link identifier that triggers access to an information source, and no cited reference combines that with the link-distinguishing speech output. These claims would have to be attacked as §103 combinations — e.g., Canon (speaking links in a distinct voice) + Kurzweil (voice-command recognition distinguished from dictation) + US 5,745,360 (parsing link identifiers from destination addresses).
4. Claim 4 (filter) and claim 5 (parser) are the weakest links in the claim set. Neither Canon nor the IBM/Infonautics references disclose them cleanly, and they are largely implementation recitations of what the independent claims already imply.
5. Claims 1–5 could be reached without any network art, since claim 1 as issued contains no network or voice-input requirement — an important point if the claim is ever construed.

---

Caveats I am obligated to flag

• I retrieved only 4 of the 17 cited references directly. The other 13 descriptions are drawn from the citation table on the face of US 5,884,266 (which gives title, inventor/assignee, priority/filing date, and issue date) and are marked unverified. Do not treat an unverified description as a §102 finding.
• Being a cited reference ≠ §102 art. I could not retrieve the file wrapper for App. 08/832,009, so I cannot tell you which references the examiner applied under §102 versus §103. That distinction requires the office actions.
• Not-cited-but-potentially-material art (lead only, flagged in the earlier summary and repeated here because it is substantive): US 6,859,777 B2 (Siemens, "Hypertext navigation system controlled by spoken words") carries a priority date of 1994-11-14 — earlier than US 5,884,266's 1997-04-02 filing. It appears in the "Cited By" list rather than the examiner's citation list, so it was not applied during prosecution. If its underlying US filing/PCT date also predates 1997-04-02, it is a candidate §102(e) reference for the voice-navigated hypertext concept. Its actual US filing date must be verified from its own file wrapper before any reliance.
• Later art, not prior art: US 6,115,686 (ITRI, "Hyper text mark up language document to speech converter," priority 1998-04-02 — one day after US 5,884,266's filing) appears in the Cited By list. It post-dates US 5,884,266 and cannot be §102 art against it; it is only relevant as evidence of the field's development.
• Text quirks preserved: as noted in the prior summary, claim 1 as issued reads "text-based link dentifiers" and the abstract/summary read "test-based link identifies." I have quoted them literally and have not corrected them, because the claim term is a potential claim-construction issue.
• If this analysis is for a validity, clearance, or IPR-grade purpose, obtain the office actions for App. 08/832,009, the full WebSpeak 1.2 Chapter 1 text, and the complete specifications of the 13 references I could not retrieve.

I'll ground this analysis in the actual disclosures of the references listed in the "Prior Art"/Citations section of the page, then build the §103 combinations. Let me pull details on the primary references.

Now let me pull details on the speech-input and document-processing references in the citations list.

Obviousness Analysis — US 5,884,266 under 35 U.S.C. § 103

Analytical date note: the task header states "Current Date: April 26, 2026" while my system context says 2026‑09‑28. The discrepancy does not affect this analysis — the relevant date is the 1997‑04‑02 filing date (no earlier priority is claimed), and all § 103 reasoning is assessed as of that date.

---

I. Governing framework and evidentiary posture

Key § 102 qualification point. The most important reference, US 5,699,486 (Canon), issued 1997‑12‑16 — after this patent's 1997‑04‑02 filing, so it cannot be § 102(a)/(b) art by grant date. It qualifies only under pre‑AIA § 102(e), because (a) it was filed in the U.S. 1997‑02‑28 as application 08/807,033, a continuation of 08/156,690 filed 1993‑11‑24, and (b) the asserted disclosure traces to that 1993 parent. Its effective § 102(e) date is therefore at least 1997‑02‑28 and, on the same‑disclosure assumption, 1993‑11‑24 — which forecloses any attempt to swear behind under Rule 131. Same logic applies to US 5,752,022 (filed 1995‑08‑07, granted 1998‑05‑12), US 5,745,360, US 5,737,619, and US 5,748,188 — all § 102(e) art on their 1995 U.S. filing dates.

---

II. Level of ordinary skill and claim construction

PHOSITA (1997): a bachelor's degree in EE/CS (or equivalent) plus ~2 years' experience, or a master's degree, with working familiarity with: HTML/HTTP document retrieval; client‑side processing of hypertext markup; text‑to‑speech synthesis including prosody/voice‑parameter control; and small‑vocabulary speech recognition. This is the level the patent itself presumes — its Background concedes HTML/HTTP, browsers, speech synthesizers, and speech recognition systems were all known, and states that "Interactive voice based systems that incorporate both speech recognition and speech synthesis would be desirable."

Constructions that matter:

That last row is decisive for the analysis: claim 1 is a two‑element apparatus claim (speaker + synthesis‑driven document reader) in which only the aural differentiation of link text is the point of novelty.

---

III. What the art of record actually teaches

A. The primary reference: US 5,699,486 — Canon (Tullis & Kodimer)

This reference is materially closer than the patent's own characterization of the field suggests. Verified disclosure (Google Patents full text; EveryPatent claims; Unified Patents bibliographic record):

• Document processor: a "help engine" reads a hypertext file of topics with imbedded hypertext links "to other locations in the document."
• Speech path: "a conventional text‑to‑speech converter is provided at 37 … converts those text strings to audio speech information … to speaker 12." Audio interface 33 drives speaker 12.
• ★ The distinguishing‑sound limitation, verbatim: "as the hypertext links are detected, an aural cue is given to the operator, such as by speaking the links to the operator in a voice distinguishable from the remainder of the help topic."
• Claim 15 of Canon is this exact feature as a method: "speaking, in a first voice, text in a first topic … speaking the encountered hypertext link in a second voice which is different from the first voice so as to provide an aural cue that one of said at least one hypertext link and not text is being spoken; jumping to the jump topic … in response to a user jump command."
• The same aural‑cue architecture appears in Canon's European counterpart EP 0 654 734 ("generating an aural cue so as to indicate that a hypertext link and not text is being spoken").

(Sources: https://patents.google.com/patent/US5699486A; http://www.everypatent.com/comp/pat5699486.html; https://portal.unifiedpatents.com/patents/patent/US-5699486-A; EP text at http://patentimages.storage.googleapis.com/ce/8d/7e/103983f2134efc/EP0654734A1.pdf.)

Limitations relative to the '266: Canon's links are intra‑file ("jump topics") rather than URLs on a network, and Canon's jump command arrives via GUI/keyboard rather than voice. But the '266 specification expressly defines "information source" to include a target "as local as within the hypertext document itself" — so the intra‑file nature is not a distinction against claim 1.

B. The non‑patent citation: WebSpeak 1.2 Documentation (The Productivity Works, 1996)

This is a 1996 printed publication — § 102(b) art on its face. Its architecture is described in the talkinginterfaces.org artifact history of pwWebSpeak:

• Parses HTML with a rule base (TLD, Tag Language Definition) that "parsed HTML content to create a navigable, structured audio version of a Web page."
• Rules included "speech and non-speech audio cues that could be inserted before and after the occurrence of any HTML element (such as audio cues to indicate a hyperlink), and changes in speech synthesizer parameters based on tags."
• Press materials of the same vintage confirm the filter function: "The intelligence built into the browser reads HTML code and automatically disregards non‑HTML content such as graphics and multimedia." The browser had an internal synthesizer (SoftVoice) with 20 selectable voices.

(Sources: https://www.talkinginterfaces.org/artifacts/pwwebspeak/; https://www.cnet.com/tech/services-and-software/visually-impaired-get-talking-browser/.)

Note the two‑way fit: WebSpeak supplies the HTML/network dimension that Canon lacks; Canon supplies the "distinguishable voice for links" dimension that WebSpeak implements by tag‑driven audio cues and synthesizer‑parameter changes.

C. Document/network handling

• US 5,752,022 (IBM, § 102(e) art): documents "in a HTML format"; "the document is retrieved into the server and parsed to discover all imbedded addresses contained therein"; URLs as embedded addresses; browsers (WebExplorer, Netscape, Mosaic, Lynx, Cello) retrieve documents over the WWW. Supplies network retrieval, HTML parsing, and the identifier/address association.
• US 5,745,360 (IBM) "Dynamic hypertext link converter system and process"; US 5,572,643 (Judson) "Web browser with dynamic display of information objects during linking"; US 5,737,619 (Judson); US 5,748,188 (NCR) "HTML extensions for graphical reporting over an internet" — all supply competent hypertext‑document manipulation and network browsing. (Titles/dates taken from the citation list; full text not verified — see § VIII.)
• US 5,333,237 (Hughes Aircraft) "Hypermedia structured knowledge base system" — structured hypermedia with links.

D. Speech output and generic computer speech

• US 5,615,300 (Toshiba) "Text‑to‑speech synthesis with controllable processing time and speech quality" — supplies the knob set (prosodic/quality parameters) that makes "voice pitch, volume, or tone" a routine control variable.
• US 5,297,027 (Hitachi) "promoting the understanding of a text by using an abstract of that text" — supplies summarizing/abstracting a document before presentation. (full text not verified)
• US 5,696,879 (IBM) "Method and apparatus for improved voice transmission."

E. Speech input / voice‑command control

• US 5,231,670 (Kurzweil, § 102(b) art; filed 1992‑03‑19, priority 1987‑06‑01): speech recognition that distinguishes spoken commands from dictated text and uses voice commands to "execute an appropriate function." Expressly motivated by accessibility: "voice operated commands would greatly facilitate use of the system" for "handicapped individuals" or where hands are otherwise occupied. (Espacenet abstract; Google Patents full text.)
• US 5,659,665 (Lucent) "Method and apparatus for including speech recognition capabilities in a computer system" — the generic cocitation of ASR into a computer platform.
• US 5,677,993 (Hitachi, filed 1993‑08‑31): converts speech input into character‑string information, integrates it with other input, and syntactically analyzes the integrated string to produce "demand information" and corresponding presentation information. This is the semantic‑analysis step that lets a spoken fragment map to a stored item. (Google Patents / EveryPatent.)

---

IV. Grounds of rejection

Ground 1 — Claim 1 over US 5,699,486 (Canon) alone (KSR-style single‑reference obviousness; alternatively § 102(e) anticipation)

Result: every element is taught. The only conceivable gap is that Canon's "information source" is an intra‑file jump topic, which the '266 specification eliminates as a distinction by defining information sources to include targets inside the document itself. If a tribunal nonetheless treated "information source" as requiring an externally addressable resource, Ground 2 closes the gap.

Ground 2 — Claim 1 over WebSpeak 1.2 (NPL) in view of Canon

Motivation (as required by KSR/MPEP 2143): both references address the same problem — making a hypertext document navigable without a visual display — and both solve the sub‑problem of alerting the listener that a navigable item is being spoken. Where WebSpeak teaches the rule mechanism (tag‑conditional cue/synthesizer change) and Canon teaches the acoustic technique (a discrete link voice), a PHOSITA would apply Canon's technique through WebSpeak's rule base because the result is a predictable improvement in link‑identifiability for the blind user both references target. This is MPEP 2143(A) (known elements combined by known methods, predictable result) and 2143(C) (known technique improving a similar device in the same way).

Ground 3 — Claim 1 additionally over Toshiba ('300) for the "sound characteristics" limitation

If the patent owner argued that "sound characteristics" requires a specifically configured synthesizer rather than a distinct voice, US 5,615,300 (controllable pitch/quality) forecloses it: selecting among known prosodic parameters to mark link text is a "result‑effective variable" design choice (MPEP 2144.04), and the '266 specification itself offers pitch, volume, or tone as interchangeable options for the same result.

Ground 4 — Claims 2, 3 and 6 over Canon and/or WebSpeak in view of Kurzweil ('670) + Lucent ('665) + Hitachi ('993), plus IBM ('022)/Judson ('643) for network access

Motivation, grounded in the references themselves (not just hindsight):

1. Accessibility, stated in the art. Kurzweil '670 states the rationale in terms: voice commands are "especially important for a system designed for use by handicapped individuals" or where "hands are occupied with another task." Canon's stated purpose is relieving screen clutter so the operator can keep working. WebSpeak exists solely to serve the visually impaired. Combining a talking document reader with a recognizer is the express solution the field was pursuing — the contemporaneous trade literature reports a "speech‑driven interface" being added to pwWebSpeak so that "keyboard interaction would no longer be needed."
2. Complementary, non‑overlapping sub‑systems. TTS and ASR are independent, well‑understood blocks that routinely coexist in a workstation (Lucent '665 is essentially the generic teaching of this). Combining them yields no unpredictable interaction.
3. The only design choice left is vocabulary scope. "At least a portion of the text‑based link identifier" is satisfied by populating the recognizer's active vocabulary with the anchor strings the document processor has just spoken. Both WebSpeak and Canon already extract and store link identifiers — so the recognizer's word list is taken directly from the already‑compiled link list. That is routine, predictable implementation (MPEP 2143(F); obvious to try with a finite, identified set of solutions).
4. The network half adds nothing inventive. IBM '022 teaches retrieving HTML over the WWW and parsing embedded addresses; the jump‑to‑target action is Canon's; substituting a URL address for an intra‑document pointer is simple substitution of one known element for another (MPEP 2143(B)) yielding a predictable result.

Ground 5 — Dependent claims 4 and 5

---

V. Motivation‑to‑combine summary (MPEP 2143 categories)

---

VI. Anticipated rebuttals and their weakness

• "The reference speaks jump‑topic links, not links to information sources." Weak — the '266 specification defines information sources to include in‑document targets, and IBM '022 + Judson supply network targets.
• "Different fields / bodily incorporation required." Weak — all references are in the same art of aurally presenting and navigating hypertext documents; no bodily incorporation is needed to combine a recognizer with a talking browser (Lucent '665 teaches the physical arrangement).
• "Unexpected results / secondary considerations." No objective evidence is of record. To carry weight, any commercial‑success or long‑felt‑need argument must have nexus to the differentiating limitation (the aural cue), and Canon + WebSpeak show the cue itself was already known, which tends to sever nexus.
• "Recognition of a portion of the link identifier is non‑obvious." This is the strongest available argument, and it is an implementation detail, not a technical advance: once the document processor stores the link identifiers (Canon, IBM '022), the recognizer's vocabulary is populated from that list. Predictable, routine engineering.
• Teaching away. None identified. Canon's practice of also speaking the jumped‑to topic in the link voice is a further embodiment, not a disclaimer of the claimed approach.

---

VII. What would actually defeat these grounds

1. A verified Rule 131 / § 1.131 invention date before 1997‑02‑28 would remove Canon as § 102(e) art — but only if Canon's disclosure is not entitled to the 1993‑11‑24 parent date. Because Canon's link‑voice teaching appears throughout (EP counterpart included), this looks unavailable.
2. Establishing that Canon's aural‑cue disclosure is new matter in the 1997‑02‑28 continuation (not in the 1993‑11‑24 parent). This is the single most productive validity attack and should be verified against Canon's file history.
3. A construction of "information source" requiring an externally addressable network resource, combined with removal of IBM '022/Judson from the ground. Even then, claim 6's network element is squarely met by IBM '022.
4. Objective evidence with nexus (licensing, copying, industry praise tied to the cue/voice‑command combination). The dense "Cited By" record suggests the field took up this approach, which cuts both ways.

---

VIII. Caveats, unverified items, and contradiction flags

Contradictions with the previously generated sections — flagged, not silently corrected:

1. Citation numbers. The Patent‑summary caveat 5 states the examiner cited "US 5,230,437" and "US 5,239,831." The authoritative citation list on this page contains neither number. The actual § 102 references are US 5,231,670 (Kurzweil) and US 5,333,237 (Hughes Aircraft), with US 5,297,027 (Hitachi). I have preserved the cited numbers literally and flag the earlier section's digits as apparently erroneous transcriptions.
2. US 6,859,777 (Siemens). The prior summary calls it "a notable pre‑dating reference" for the voice‑navigated hypertext concept. The page lists its priority as 1994‑11‑14, but a foreign priority date generally does not establish the pre‑AIA § 102(e) date of a U.S. patent — the U.S. filing date governs (absent an English‑language PCT designating the U.S.). I did not verify its U.S. filing date, so I have excluded it from the grounds and flag its prior‑art status as unverified. (I also did not verify the earlier section's assertion that it "shares office action history with this family.")
3. "US 6,166,xxx‑family voice‑browser work" in the prior summary is a partial number. Per the operating rule against auto‑correcting identifiers, I have not resolved or relied on it.
4. Claim‑text OCR artifacts. Claim 1 as issued recites "text-based link dentifiers"; the Abstract and summary recite "test-based link identifies"; the description OCR shows "anally distinguish." These are reproduced literally and may matter for claim construction (e.g., whether "dentifiers" is a clerical error correctable under 35 U.S.C. § 255 / Ikbhal). My charts read the term as "link identifiers."

Items I could not verify within the search budget (mapping therefore rests on the citation list's titles and dates plus general knowledge — treat as provisional):

• Full text of US 5,615,300 (Toshiba), US 5,659,665 (Lucent), US 5,745,360 (IBM), US 5,572,643 / US 5,737,619 (Judson), US 5,333,237 (Hughes), US 5,297,027 (Hitachi), US 5,696,879 (IBM), US 5,732,219 (Vermeer), US 5,748,188 (NCR).
• The 1993‑11‑24 parent disclosure content of Canon's 08/156,690 (the new‑matter question in § VII.2).
• The exact publication date of EP 0 654 734 A1; if it published before 1996‑04‑02, Canon's aural‑cue teaching is also § 102(b) art independent of the § 102(e) analysis.
• Whether the WebSpeak 1.2 Documentation Chapter 1 (the NPL of record) contains the same TLD audio‑cue language as the later artifact history I relied on. The artifact history and the 1996 press coverage corroborate the substance, but the Chapter 1 text itself should be read directly before filing any paper citing it.

Bottom line: on the art of record, claim 1 is at minimum obvious over US 5,699,486 alone, and plainly obvious over WebSpeak 1.2 in view of US 5,699,486; claims 2, 3, 4, 5 and 6 are obvious over those references further in view of US 5,231,670, US 5,659,665, US 5,677,993, US 5,752,022 and US 5,572,643, with the motivation supplied by the accessibility/hands‑free rationale expressly stated in the references themselves and conceded in this patent's own Background. The strongest residual patentability arguments are procedural (invention date, new matter in the Canon continuation, "information source" construction), not technical.