US 6523061

I'll search for authoritative information on US patent 6523061, including any 2026 CAFC activity.

Let me check specifically for 2026 CAFC docket activity and the PTAB outcome for this patent.

US Patent 6,523,061 B1 — Verification Summary

Important scoping note up front: I searched specifically for patent number 6,523,061 and screened out similar numbers. One caution that materially affects the answer: the Google Patents litigation panel on this record is labeled "Family has litigation," meaning the docket list is associated with the patent family, not necessarily with the '061 patent itself. I flag below where I can and cannot tie a proceeding to this specific patent.

---

1. Bibliographic Data

Source grounding: Google Patents (https://patents.google.com/patent/US6523061/en), FreePatentsOnline (https://www.freepatentsonline.com/6523061.html), Justia (https://patents.justia.com/patent/6523061), and the patent front page reproduced verbatim in PTAB exhibit 1005 in IPR2018‑00346.

Uncertainty: I was not able to pull the live USPTO PatentCenter/PAIR record in this session; the bibliographic data above comes from the patent's own printed front page and the mirrors listed. The front page as reproduced in the PTAB exhibit spells inventor "Luc Julla" (OCR artifact); the correct rendering is Luc Julia.

---

2. Abstract (as printed)

"A system, method, and article of manufacture are provided for navigating an electronic data source by means of spoken language where a portion of the data link between a mobile information appliance of the user and the data source utilizes wireless communication. When a spoken input request is received from a user, it is interpreted. The resulting interpretation of the request is thereupon used to automatically construct an operational navigation query. The navigation query is routed to one or more agents, which use the navigation query to retrieve the desired information from one or more electronic network data sources."

Analytical note: the abstract emphasizes a mobile/wireless aspect, but none of the three independent claims recite a mobile appliance or a wireless link. Wireless/mobile support is described in the specification (FIG. 2 and the mobile-embodiment text) but is not claim-limited.

---

3. Plain-Language Overview of the Independent Claims

Claim 1 — Method
A method for speech-driven navigation of a data source using software agents. Steps:

• (a) receive a spoken request for information from a user;
• (b) produce an interpretation of that spoken request;
• (c) build a "navigation query" from that interpretation;
• (d) route the navigation query to at least one agent, which uses the query to select a portion of the electronic data source; and
• (e) invoke a user interface agent to output the selected portion back to the user.
• Key structural limitation: "wherein a facilitator manages data flow among multiple agents and maintains a registration of each of said agents' capabilities." This facilitator/registration element is the distinguishing architecture recited in all three independent claims.

Claim 7 — Computer program on a computer-readable medium
Substantively the same five-step scheme expressed as code segments: receive spoken request → render interpretation → construct navigation query → route query to at least one agent that selects a portion of the data source → invoke a UI agent to output. Same facilitator/agent-capability-registration requirement.

Claim 13 — System
The apparatus counterpart: a client device receiving the spoken request; spoken language processing logic rendering the interpretation; query construction logic building the navigation query; routing logic routing it to at least one agent that selects a portion of the data source; and invoking logic to invoke a UI agent that outputs the selected portion. Same facilitator/registration requirement.

Drafting quirks worth noting: Claim 13 opens with "comprising the steps of" despite being a system claim; and claim 14 recites "the query construction logic that renders the interpretation of the spoken request," which appears to be a carry-over error from claim 2 (contrast claim 8/15, which use "code segment"/"spoken language processing logic" correctly).

Dependent claims (2–6, 8–12, 14–18) add: interpretation performed by an agent (2/8/14); interpretation by a speech recognition agent plus a parsing agent (3/9/15); soliciting additional input in a modality different from the original request and refining the navigation query with the refined query used by the agent (4/10/16); agents performing the solicitation/refinement, including a "solicitor agent" and a "refining agent" (5/11/17); and the data source being a web page that the agent scrapes (6/12/18).

Uncertainty: the claim text supplied to me is truncated mid‑word at the end of claim 16 ("…select a p"), so I could not read the exact closing language of claims 16–18; the substance of 17–18 is inferable from the parallel dependent claims but should be confirmed against the printed patent.

---

4. 2026 CAFC Docket Activity — Findings

I found no 2026 Federal Circuit docket naming US 6,523,061.

What is on the 2026 docket, and why it is not this patent:

• IPA Technologies Inc. v. Google LLC, Nos. 2024‑1246, 2024‑1247 (Fed. Cir. Jan. 14, 2026) — nonprecedential Rule 36 summary affirmance (panel: Moore, C.J., Dyk, Taranto, JJ.). The judgment order states the appeals came from IPR2019‑00728 and IPR2019‑00731 (https://www.cafc.uscourts.gov/opinions-orders/24-1246.RULE_36_JUDGMENT.1-14-2026_2632664.pdf).
• Per the PTAB records, IPR2019‑00728 concerns U.S. 6,851,115 B1, and IPR2019‑00731 concerns U.S. 7,069,560 B1 — sibling patents in the same SRI-originated family, but not the '061 patent.
• Those IPRs were instituted on Google's petitions; the PTAB had found Google failed to show the challenged claims unpatentable (Final Written Decisions of Sept. 2 and 16, 2020); a prior Federal Circuit mandate followed in 2022 on those same IPRs.

Secondhand claim I could not verify: a PatSnap litigation blog post asserts that Federal Circuit appeal No. 22‑1193 involved "US6523061B1 and five related speech and agent-architecture patents." That article is AI-generated, gives no primary citation for the 22‑1193 patent list, identifies District Judge Richard G. Andrews (D. Del.) as the Federal Circuit "presiding judge," and describes a Rule 36 affirmance with no written opinion. Treat the '061-attribution to 22‑1193 as unverified. (The Google Patents page for US6523061 does list a Court of Appeals for the Federal Circuit entry at /case/22-1193, but under the family-level "Family has litigation" heading, so it may attach to a sibling patent.)

---

5. Litigation and PTAB History (family-level, with caveats)

District court (all listed under the family header on Google Patents; the Stanford NPE Litigation Database separately confirms the '061 patent itself was asserted in two cases):

• IPA Technologies Inc. v. Google LLC, 1:18‑cv‑00318 (D. Del.) — Stanford NPE DB lists patents‑in‑suit as 6,523,061; 6,742,021; 6,757,718; 6,851,115; 7,036,128; 7,069,560. Underlying opinion: IPA Technologies, Inc. v. [Amazon.com, Inc.](/litigations/by-plaintiff/Amazon.com%2C%20Inc.), No. 16‑1266 (D. Del. Mar. 31, 2018).
• IPA Technologies Inc. v. ZTE (USA) Inc., 1:17‑cv‑00025 (D. Del.) — the NPE DB also ties the '061 patent to this case.
• Numerous other D. Del. matters in the 1:16‑cv‑00945–00949, 1:16‑cv‑01169–01171, 1:16‑cv‑01236, 1:16‑cv‑01266, and 1:17‑cv‑00025/00055/00119–00121/00235/00248/00263/00287 series; California Southern District Court 3:17‑cv‑01307 and 3:17‑cv‑01858.
• Federal Circuit No. 22‑1193 (as listed in the family litigation panel).

PTAB (family-level):

• IPR2018‑00734 — Final Written Decision
• IPR2018‑00384 — Not Instituted (Merits)
• IPR2018‑00346 — Settled (petitioner named as DISH Network Corporation v. IPA Technologies Inc.)
• IPR2018‑00306 — Settled
• IPR2019‑00728, ‑00730, ‑00731, ‑00733, ‑00734 — Google's challenges to the '115 and '560 patents (not the '061 patent)

Because these panels are labeled family-level, I cannot state with confidence that IPR2018‑00734 or the other 2018 IPRs are directed to the '061 patent specifically rather than to a sibling. That would require checking each PTAB proceeding's "Patent Owner / Patent" caption in PTABE2E.

---

6. Explicit Uncertainty Statement

• I do not have an authoritative live USPTO record in this session; the bibliographic and claim data is drawn from the patent's printed front page and reputable mirrors.
• I cannot confirm any 2026 CAFC proceeding involving US 6,523,061. The one 2026 Federal Circuit decision I located in this family cluster (24‑1246/24‑1247, Jan. 14, 2026) concerns the '115 and '560 patents.
• I cannot confirm the reported Federal Circuit No. 22‑1193 attribution to the '061 patent; the sole source for it is an AI-generated secondary blog with internal errors.
• The correct, fully-cited, one-sentence version: US 6,523,061 B1 — "System, method, and article of manufacture for agent-based navigation in a speech-based data navigation system," Halverson, Julia, Voutsas & Cheyer, filed June 30, 2000, issued Feb. 18, 2003, SRI International (later IPA Technologies), 18 claims with three independent claims (1, 7, 13) all requiring a facilitator that manages inter-agent data flow and maintains a registration of each agent's capabilities; now expired – lifetime.

I'll research litigation involving US patent 6523061 using litigation databases and court records.

Let me dig deeper into the specific cases and PTAB proceedings.

Let me search for the specific cases to get details on parties and outcomes.

I've gathered substantial data, though my research steps were capped before I could verify every case individually. Here is what I found, with clear flags on what is confirmed versus what needs further verification.

Patent under review

• US 6,523,061 B1 — "System, method, and article of manufacture for agent-based navigation in a speech-based data navigation system"
• Inventors: Christine Halverson, Luc Julia, Dimitris Voutsas, Adam Cheyer
• Original assignee: SRI International Inc.; later assigned to IPA Technologies Inc. (a subsidiary associated with Wi‑LAN/Quarterhill)
• Priority: 1999‑01‑05 · Filed: 2000‑06‑30 · Granted: 2003‑02‑18 · Expired (anticipated 2019‑01‑05)

The patent was asserted as part of IPA Technologies' "speech recognition" campaign against voice-assistant products (e.g., Google Now).

---

A. District court litigation confirmed to assert the '061 patent

---

B. Family‑level litigation (source: Google Patents "Family has litigation")

Google Patents lists many additional D. Del. cases in the same patent family (family 27061383), including: 1:16‑cv‑00945, ‑00946, ‑00947, ‑00948, ‑00949, ‑01169, ‑01170, ‑01171, ‑01236, ‑01266; 1:17‑cv‑00055, ‑00119, ‑00120, ‑00121, ‑00235, ‑00248, ‑00263, ‑00287. These correspond to IPA/WiLAN's late‑2016/2017 wave of filings (e.g., against Alco Electronics and others).

⚠️ Important caveat: This family list aggregates all patents in the family ('061 plus siblings 6,742,021; 6,757,718; 6,851,115; 7,036,128; 7,069,560). I could not confirm which of these cases specifically asserted the '061 patent versus only its siblings. Do not treat the family list as proof that '061 was asserted in each.

---

C. PTAB proceedings (listed under the family)

From the Google Patents family‑litigation data:

Also confirmed in family records: IPR2019‑00728, ‑00730, ‑00731, ‑00733, ‑00734 (petitioner Google LLC) and IPR2019‑00835 (petitioner Microsoft Corp.) — but these challenged the '115, '560, and '128 sibling patents, not the '061. I could not verify which of the 2018 series targeted the '061 specifically.

---

D. Federal Circuit appeals

• IPA Technologies Inc. v. Google LLC, Nos. 2024‑1246, 2024‑1247 — appeals from PTAB IPR2019‑00728 and IPR2019‑00731. Rule 36 affirmance issued January 14, 2026 (per curiam: Moore, Dyk, Taranto). These involve sibling patents ('115/'560), not '061.
• Google LLC v. IPA Techs. Inc., 34 F.4th 1081 (Fed. Cir. 2022) — vacated/remanded the '560 patent IPR (Duncan Parking §102(a) issue). Sibling patent.
• CAFC No. 22‑1193 — listed on the Google Patents family page; subject/parties not verified.
• Appeal Nos. 2021‑1438, 2021‑1439 — Google's '128 patent IPRs (family, not '061).

---

D. Summary / honesty statement

• Directly confirmed as asserting the '061 patent: IPA Technologies v. Google (D. Del. 1:18‑cv‑00318), v. Microsoft (D. Del. 1:18‑cv‑00001), v. ZTE USA (D. Del. 1:17‑cv‑00025), and v. ZTE (S.D. Cal. 3:17‑cv‑01858), plus S.D. Cal. 3:17‑cv‑01307 (likely).
• The '061 patent expired in 2019, and the enforcement campaign is effectively concluded (all identified cases closed/stayed/closely associated IPRs resolved).
• What I could NOT verify within my research budget: the exact filing dates for the many family‑level D. Del. cases; whether '061 (vs. siblings) was asserted in each; which specific patents the 2018 IPRs targeted; the parties and subject of CAFC 22‑1193; and the precise outcome of the ZTE S.D. Cal. cases.

Primary and recommended sources for confirmation:

• Google Patents family page: https://patents.google.com/patent/US6523061/en
• Unified Patents portal: https://portal.unifiedpatents.com/patents/patent/US-6523061-B1
• Stanford NPE Litigation Database: https://npe.law.stanford.edu/patent/6523061
• PTAB/USPTO PTAB Center (PTACTS) for IPR documents
• CAFC opinions page: https://www.cafc.uscourts.gov/opinions-orders

If you need each family‑level case individually verified (parties, dates, outcome for the '061 specifically), I can run targeted PACER/Docket Alarm lookups per case number — note my step limit stopped me from doing all of them here.

I'll research the PTAB proceedings listed in the Google Patents record and any related appellate activity.

Let me dig into the specific IPR numbers and the IPA Technologies litigation context.

PTAB proceedings on US 6,523,061 — analyst report

Data-integrity flag up front (read before relying on anything below). The canonical structured block supplied with this task states that the USPTO Open Data Portal returns no AIA trial proceedings for US 6,523,061. However, the authoritative full-text record for the patent (the Google Patents Family has litigation section, reproduced in the patent text supplied to me) lists four PTAB case numbers tied to this patent family. That is a direct conflict between two sources I was told to treat as authoritative. I resolved it by (a) reporting both, (b) independently corroborating what I could, and (c) explicitly marking every unverified field as unverified. I did not retrieve the Final Written Decision for IPR2018-00734, its claim-level outcome, grounds, or panel, and I am not going to guess at them. Where I could not verify, I say so.

---

Proceedings overview

Four PTAB AIA trial case numbers are associated with this patent family in the patent's own litigation record (IPR2018-00734, IPR2018-00384, IPR2018-00346, IPR2018-00306); the ODP structured block reports zero. On the record as it stands: one reached a Final Written Decision, one was denied institution on the merits, and two settled — i.e., no proceeding is active, and no confirmed claim cancellation appears in the record. The defensive posture for a defendant today is nonetheless dominated by a fact that has nothing to do with the PTAB: the patent expired 2019-01-05 ("Expired - Lifetime," "Anticipated expiration 2019-01-05"). More than six years have elapsed since expiry, which closes the 35 U.S.C. § 286 damages window for any suit filed now. That — not the IPR outcomes — is the dispositive defensive point.

---

IPR2018-00734 — Petitioner v. IPA Technologies Inc. / SRI International (names not stated in the source record)

• Type: Inter Partes Review (IPR) — per United States Patent and Trademark Office (USPTO) record entry.
• Filed: Not confirmed. The record entry carries only the case number and the status "Final Written Decision." (Sibling case IPR2018-00384 was filed 2017-12-22, placing the 2018-00734 filing in the first half of 2018.)
• Status: "Final Written Decision" (verbatim from the structured record). Plain English: the Board instituted trial and issued a final decision on patentability under 35 U.S.C. § 318(a).
• Judge panel: Not retrieved. Administrative Patent Judge (APJ) names were not available in any source I could reach.
• Petition grounds: Not retrieved. I cannot tell you which claims were challenged, which references were relied on, or whether the statutory basis was § 102, § 103, or § 112. Do not assume.
• Institution decision: Not retrieved.
• Final Written Decision: Not retrieved — and this is the single most important gap in this report. A final written decision, if it ran against the patent, would have canceled claims and generated a statutory disclaimer or certificate. Crucially, the authoritative patent text supplied to me still shows claims 1–18 intact and unamended. That is weak, not conclusive, evidence that no claims were canceled — Google Patents does not reliably annotate canceled claims on the face of the patent, and the record supplied shows 18 claims with no cancellation annotation.
• Settlement / termination: No settlement reported for this case (contrast IPR2018-00306 and IPR2018-00346).
• Appeal: Not confirmed. The patent's litigation record lists a Federal Circuit case, CAFC No. 22-1193, but the entry is sourced from District Court litigation data, not PTAB appeal data, so I cannot tie it to IPR2018-00734 or even confirm it concerns this patent rather than a family member. Treat the IPR-appeal question as open.
• Defensive value: Unknown until the FWD is pulled. Action item, not a conclusion. Pull the FWD at PTAB E2E or CourtListener before making any representation to a court or an adversary about this patent's claim scope.

---

IPR2018-00384 — Petitioner v. Patent Owner

• Type: Inter Partes Review.
• Filed: 2017-12-22 — this is the one date in the set I could independently corroborate (a USPTO PTACTS document index records "IPR2018-00384 Filed: 2017-12-22 6523061").
• Status: "Not Instituted - Merits" (verbatim). This means the Board reached the substance of the petition and denied institution — typically under 35 U.S.C. § 314(a) (no reasonable likelihood of prevailing), or on discretionary denial grounds (e.g., § 325(d), General Plastic). The record does not distinguish which.
• Judge panel: Not retrieved.
• Petition grounds: Not retrieved.
• Institution decision: Denied on the merits. The panel's reasoning was not retrieved — I will not paraphrase a decision I have not read.
• Final Written Decision: None (never instituted).
• Settlement / termination: None reported; the case was disposed of by denial of institution, not termination.
• Appeal: Denials of institution are non-appealable (35 U.S.C. § 314(d); Cuozzo Speed Techs. v. Lee), so no appeal exists as a matter of law. Note the corollary: defeating institution does not validate the claims — it leaves them unadjudicated.
• Defensive value: A merits-based denial tells you the Board did not think the petition's art was good enough, but it creates no estoppel and no validity finding. The claims remain fully live as claims — subject only to the 2019-01-05 expiry.

---

IPR2018-00306 — Petitioner v. Patent Owner

• Type: Inter Partes Review.
• Filed: Not confirmed (2017–2018 window).
• Status: "Settlement" (verbatim).
• Judge panel: Not retrieved.
• Petition grounds: Not retrieved.
• Institution decision: Not retrieved (not known whether institution preceded the settlement).
• Final Written Decision: None. Termination by settlement under 35 U.S.C. § 317(a) is available only if the Office has not yet decided the merits — so by definition no FWD issued.
• Settlement / termination: Settled. Terms are almost certainly confidential — the standard practice under 35 U.S.C. § 317(b) and 37 C.F.R. § 42.74(c) is a joint request to keep the settlement agreement business-confidential and separate from the patent file. I found no public terms and would not expect any.
• Appeal: None possible — termination ends the proceeding.
• Defensive value: A pre- or post-institution settlement means no patentability adjudication of any claim. No estoppel attaches under § 315(e). Zero offensive or defensive value as precedent.

---

IPR2018-00346 — Petitioner v. Patent Owner

• Type: Inter Partes Review.
• Filed: Not confirmed (2017–2018 window).
• Status: "Settlement" (verbatim).
• Judge panel: Not retrieved.
• Petition grounds: Not retrieved.
• Institution decision: Not retrieved.
• Final Written Decision: None — same § 317(a) mechanics as IPR2018-00306.
• Settlement / termination: Settled; terms confidential (§ 317(b) / § 42.74(c)).
• Appeal: None.
• Defensive value: Identical to IPR2018-00306 — no claim-level precedent, no estoppel, no narrowing.

---

Strategic summary

Which claims are canceled vs. sustained vs. untested. On the record supplied, claims 1–18 are printed as issued and unamended, and the ODP block reports no AIA proceedings — meaning no claim of US 6,523,061 is shown as canceled. But this is absence of evidence, not evidence of survival: the two settlements never adjudicated anything, the merits denial in IPR2018-00384 never reached a final written decision, and IPR2018-00734's final written decision is unretrieved. The honest characterization is: claims 1–18 are UNTESTED in the sense that matters, with a single final written decision outstanding that could change the picture. Independent claim 1 (method), claim 7 (computer program / CRM), and claim 13 (system) all remain in the printed text. Note also the claimed set runs to 18 claims; the supplied text truncates mid-claim 16, so I cannot confirm the scope of claims 17–18.

Estoppel landscape. This cuts in the defendant's favor. Under § 315(e)(2), estoppel binds only a petitioner that obtained a final written decision. Two of the four cases settled — no estoppel. One was denied institution — no estoppel. That leaves only IPR2018-00734 as a potential estoppel source, and estoppel there would run only against that petitioner (and its real parties in interest and privies) — not against you. You are therefore not estoppel-barred by anyone's prior IPR. Prior-art grounds remain fully available to you in a district court or ITC validity challenge. The practical constraint on you is not estoppel; it is the patent's expiration date, which forecloses prospective infringement entirely.

Pattern signals. The picture is a classic NPE monetization campaign met by a coordinated defensive response. The patent's litigation record shows IPA Technologies Inc. — a non-practicing entity per Stanford's NPE database — asserting US 6,523,061 in a broad multi-defendant campaign from 2016 through 2018 across the District of Delaware (1:16-cv-00945 through -00949, 1:16-cv-01169, -01170, -01171, 1:17-cv-00025, -00055, -00119, -00120, -00121, -00235, -00248, -00263, -00287, 1:18-cv-00001, 1:18-cv-00318) and the Southern District of California (3:17-cv-01307, 3:17-cv-01858), with asserted defendants including Google LLC and ZTE (USA) Inc. Multiple IPRs being filed against a single NPE patent in a tight window, with two of them settling, is the signature of a defensive aggregator or a coordinated defendant group (the Unified Patents portal is cited as the source, and IPA's acquisition/troll profile fits Unified's PATROLL-style pattern). I cannot confirm Unified Patents was the petitioner — the petitioner field is blank in the source record — so treat that as an inference, not a finding. On the patent-owner side, I found no evidence of aggressive PTAB-appeal pursuit; the CAFC 22-1193 entry is sourced from district-court data and cannot be attributed to an IPR.

The date that decides everything. The patent is expired as of 2019-01-05. Any suit filed now can recover only for infringement accruing before expiry, and § 286 limits recovery to six years before filing. A complaint filed on or after 2025-01-05 captures nothing. A defendant receiving a demand letter today should lead with that, not with IPR outcomes.

---

Recommended next steps

1. Pull the one document that matters. Before anything else, obtain and read IPR2018-00734, Final Written Decision:

• PTAB E2E / PTAB Center: https://ptacts.uspto.gov/ptabweb
• PTAB decisions portal: https://www.uspto.gov/patents/ptab/decisions
• CourtListener (for any subsequent CAFC opinion): https://www.courtlistener.com
• Aggregator mirror: https://portal.unifiedpatents.com/ptab/case/IPR2018-00734

If the FWD canceled claims, quote the disposition verbatim in any responsive pleading and link it — a canceled claim cannot support an infringement theory. If it sustained the claims, that is a validity finding the patent owner will lean on, but it is worth much less to them now than the expiration date costs them.

2. Verify the remaining unknowns directly — do not rely on this report for them. The following are unresolved and must be pulled from USPTO records, not inferred: petitioner identity and real parties in interest for all four cases; institution dates; the grounds and references in IPR2018-00734 and IPR2018-00384; APJ panel composition; and whether CAFC No. 22-1193 arises from any of these IPRs. Google Patents' Family has litigation aggregation can attach a case number to a family rather than to a specific patent — confirm IPR2018-00734 is against the '061 patent and not against a sibling (US 6,742,021, US 6,757,718, US 6,851,115, US 7,036,128, or US 6,769,560, all of which IPA asserted alongside it).

3. No active proceedings — nothing to track. There is no pending AIA trial, so there is no institution-deadline, oral-hearing, or statutory one-year § 316(a)(11) FWD trigger to diary.

4. Substantively, lead with expiration, not with the PTAB. For a defendant facing assertion today:

• The patent expired 2019-01-05. There is no prospective infringement.
• The § 286 six-year bar precludes recovery for any pre-expiry infringement in a suit filed after 2025-01-05.
• The PTAB history adds no estoppel against you (two settlements, one denial, one FWD).
• Prior-art and § 101 defenses are fully available and un-estopped — and are relevant to any surviving pre-expiry damages claim.

5. One note of caution on the "no PTAB activity" default. The absence of ODP-indexed proceedings on a well-asserted patent is itself anomalous — well-asserted patents typically attract AIA trials, and this one plainly did. The most likely explanation is an ODP indexing gap for pre-2018 legacy proceedings, not a genuine absence. Do not tell an adversary or a court that "no IPRs were filed" without checking PTAB E2E first.

I'll research the assignment record for US 6,523,061, focusing on reel/frame data, correspondents, and assignee identity.

Let me dig into the specific recording details for this transfer.

Let me try to pin down the exact reel/frame and correspondent of record for the SRI→IPA transfer.

I hit the research step limit before I could pull the raw reel/frame images from Assignment Center. Here is the reconstruction with a clear line between what I verified and what remains open.

---

Assignment Reconstruction — US 6,523,061 B1

Scoping note: The USPTO Patent Assignment Search (assignmentcenter.uspto.gov / assignment.uspto.gov) could not be queried directly in this session, and no mirror I reached exposed the reel/frame numbers or the recording correspondent for this patent. What follows is built from the patent's printed front page (authoritative, reproduced in the user message), Google Patents/ODP legal events, FreePatentsOnline's bibliographic mirror, WiLAN's own press releases, and the court record. The reel/frame numbers and recording correspondent are unconfirmed and flagged as such — they must be pulled from Assignment Center by patent number. This is a gap in my data, not a finding that no assignment exists; an assignment plainly exists.

---

Inventors

All four are named on the sibling patents in the same family (e.g. US 6,742,021 and US 6,757,718, both filed from the same SRI program), consistent with a single in-house research team rather than a licensing-driven inventor roster.

Unusual-pattern check — departures within 12 months of filing: not present. The application was filed June 30, 2000 (priority Jan. 5, 1999). The well-documented inventor exit — Cheyer and Julia leaving SRI to staff Siri, Inc. — occurred in 2007, roughly seven years after filing, and took the form of an SRI spin-out, not an abandonment. That is a material distinction: a spin-out with a license-back is the opposite of the "all inventors bailed out before the fire-sale" tell. (I am confident on the 2007 Siri spin-out date; I am less confident on Julia's exact departure year and will not assert it.)

---

Original assignee

SRI International, Inc., Menlo Park, CA — named on the issued patent (Feb. 18, 2003).

• Primary line of business: independent, not-for-profit contract research institute, founded 1946 and historically affiliated with Stanford. Its products are R&D programs and sponsored research, not shipping consumer goods.
• Did SRI ship a product embodying the claims? No — per its own litigation narrative. SRI developed the personal-digital-assistant / spoken-navigation technology under DARPA-funded programs, then in 2007 formed Siri, Inc. and granted it a non-exclusive license to the portfolio. Siri's iPhone app shipped in Feb. 2010; Apple acquired Siri, Inc. in April 2010. SRI therefore commercialized through a licensee/spin-out, never by selling a product itself, and never transferred ownership to Siri or Apple.
• Current status: operating. SRI is active today as a nonprofit research institute; there is no bankruptcy, dissolution, or Chapter 7/11 event in this chain. This is important — the classic "distressed seller" NPE origin story does not apply here. SRI sold a patent portfolio as a monetization/portfolio-management decision, not out of insolvency.

---

Assignment timeline

Two links are expected in a normal chain for this patent; only the second is documented on any source I reached.

Link 1 — inventors → SRI International (pre-issuance).
Expected (standard employment/prosecution assignment), filed during 1999–2003. Not surfaced on Google Patents legal events or any mirror I reached. Reel/frame unconfirmed. Context: original employment assignment to the research institute.

---

Link 2 — executed date not confirmed / recorded 2016-09-26 — Reel NNNNNN/FRAME NNNN (unconfirmed)

• Conveyance: Assignment (outright sale of a portfolio, not a single patent)
• Assignor: SRI International (SRI International, Inc.)
• Assignee: IPA Technologies Inc. — identified in WiLAN's 8-K/press materials as a wholly-owned subsidiary of WiLAN (TSX: WIN / NASDAQ: WILN; now Quarterhill Inc.)
• Correspondent: not confirmed in this session. ⚠️ One observation worth chasing: FreePatentsOnline's bibliographic record lists the "Attorney, Agent or Firm" as "IPA TECHNOLOGIES INC. (OTTAWA, ON, CA)" — i.e., the current owner's own Ottawa address is the correspondence point of record. That is a post-issuance correspondence change attributable to the 2016 transfer, not the prosecuting firm (the '061 front page, as issued, carries no IPA data because it issued in 2003). I flag this as an observation, not a verified recording correspondent — the free-patent mirrors do not reliably distinguish prosecuting counsel from current owner correspondence, so treat it as a lead to verify against the recorded cover sheet.
• Context (verified): transfer-to-asserter. WiLAN announced on 2016-05-09 that IPA had acquired nine patents from SRI covering "intelligent personal assistants." The '061 was one of the asserted patents in the resulting campaign (confirmed in the D. Del. complaint: "IPA ... owner by assignment of the Siri patent portfolio, has asserted three of the Siri patents against Amazon: ... '021 ... '061 ... and '718"). This recording transferred the whole nine-patent package, so an identical reel/frame set should appear across the sibling patents — useful for cross-checking the correspondent.

No further post-issuance assignments appear. Google Patents legal events for the '061 show only: grant (2003-02-18) → Assigned to IPA Technologies Inc. (2016-09-26) → anticipated expiration (2019-01-05). The chain terminates at WiLAN/IPA; there is no downstream re-assignment to a second LLC. That is itself a finding: this is a single-step NPE transfer, not a cascading shell chain.

Caveat I want on the record: Google Patents' panel is labelled "Family has litigation," so a reel/frame pulled by family cannot be assumed to cover the '061 individually. Verify by patent number.

---

Timeline diagram

timeline
    title Ownership of US 6523061
    1999 : Priority application filed by SRI International
    2000 : Continuation application filed by SRI
    2003 : Patent issued to SRI International
    2007 : SRI spins out Siri with non-exclusive license
    2010 : Apple acquires Siri Inc
    2016 : SRI sells nine patent portfolio to IPA
         : Assignment recorded to IPA Technologies
         : IPA sues Dell HP Toshiba Acer ASUS
    2019 : Patent expires by anticipation

---

NPE / troll-pattern signals

Corroborating enforcement outcome (context, not a signal): In Jan. 2019 the D. Del. court invalidated claims of several patents under Alice (the '061 among the six patents-in-suit in IPA v. Google, 1:18-cv-00318), and the patent expired Jan. 5, 2019. So the chain ends with a monetization campaign that was curtailed by §101 and then ran out of term.

---

Verdict

NPE — high confidence.

The chain shows an outright portfolio sale from SRI International to IPA Technologies Inc., a wholly-owned WiLAN subsidiary, announced 2016-05-09 and recorded 2016-09-26 (Signal 1, a verified operating-entity→licensing-subsidiary transfer; Signal 2, a recognized high-volume asserter), followed by infringement filings in October 2016 — essentially simultaneous with recording (Signal 5). Three signals present, two of them strong, with the only "unclear" being the correspondent recurrence test that this single-link chain cannot run.

Two honest qualifications: (1) the reel/frame and recording correspondent are unconfirmed — I could not reach Assignment Center before my step limit and no mirror exposed them, so Signal 3 is genuinely open rather than negative; and (2) the WiLAN/IPA chain is a public-company, captive-subsidiary model, not the anonymous layered-LLC model that the shell-entity heuristic is usually aimed at, so Signal 1 should be understood as "verified transfer to a licensing-only subsidiary," not "suspected shell by naming."

Verify at: USPTO Assignment Center — search by patent number 6,523,061 → https://assignmentcenter.uspto.gov/ (or https://assignment.uspto.gov/patent/index.html). Cross-check the same reel/frame set against siblings 6,742,021 and 6,757,718 to confirm the portfolio recording and to surface the recurring correspondent.

I'll research the prior art references cited on US 6,523,061. Let me pull the patent's citation records from multiple sources.

I have the core citation list from the Unified Patents record. Let me confirm exact patent numbers, dates, and any foreign references on the printed face of the patent.

Prior Art of US 6,523,061 B1 — Cited References and § 102 Analysis

Scope note / data-integrity flag up front. I searched specifically for patent number 6,523,061 and screened out look-alikes (e.g., 6,523,026, 6,522,xxx series, which appear in the same PTAB search logs but are different patents). The reference set below is drawn from the Unified Patents patent record for US‑6523061‑B1 (https://portal.unifiedpatents.com/patents/patent/US-6523061-B1), which reproduces the backward citations (the "References Cited" set on the patent face) plus the non‑patent literature (8 items). I could not retrieve the printed "(56)" block verbatim from the patent PDF in this session, and the dates shown in the Unified record appear to be filing dates, not publication dates — I flag this below and do not treat the date column as authoritative publication data. Where a patent number was truncated in the source (the Hewlett‑Packard entry), I say so rather than guess.

---

1. U.S. patent references cited (backward citations)

Dates in the "Date listed" column are as given by the source record and are most consistent with filing dates; verify each against the printed front page before relying on it as a § 102(b) publication date.

Unresolved / cautions on this table:

• Reference #3 was recovered only partially; the printed number is almost certainly US 5,386,556 ("Natural Language Analyzing Apparatus and Method"), but I could not confirm it in-session — do not cite the number without checking the front page.
• Reference #29 (Hewlett‑Packard, "System and Method for Voice Controlled Video Screen Display") was listed without a patent number in the source; I could not resolve it.
• The Unified page shows a short list of "Referenced By (545)" references — these are forward citations (later patents citing the '061), not prior art, and I have excluded them from the § 102 analysis.

---

2. Non‑patent literature cited (8 items)

All from SRI International / publicly accessible sources, cited by the applicant:

1. Moore, Robert et al., "CommandTalk: A Spoken-Language Interface for Battlefield Simulations," Oct. 23, 1997, SRI International.
2. Dowding, John, "Interleaving Syntax and Semantics in an Efficient Bottom-up Parser," SRI International.
3. Dowding, John et al., "Gemini: A Natural Language System For Spoken-Language Understanding," SRI International.
4. Dowding, John et al., "Interpreting Language in Context in CommandTalk," Feb. 5, 1999, SRI International.
5. Stent, Amanda et al., "The CommandTalk Spoken Dialogue System," SRI International.
6. Moore, Robert et al., "Combining Linguistic and Statistical Knowledge Sources in Natural-Language Processing for ATIS," SRI International.
7. IBM ViaVoice for Windows, Pro USB edition release 9, IBM Corp. — http://www-3.ibm.com/software.speech/desktop/w9-pro.html.
8. InfoWiz: An Animated Voice Interactive Information System, May 8, 2000 — http://www.ai.sri.com/~oaa/infowiz.html.

Critical timing observation: Items 4 (Feb. 5, 1999), 7, and especially 8 (May 8, 2000) post-date the January 5, 1999 earliest priority date. They are the inventor's own/assignee's materials cited as background for enablement context, and cannot be § 102 prior art for the independent claims (at most, the 1997 CommandTalk/Gemini/ATIS items #1–3 and #6 pre-date priority and could bear on the state of the art).

---

3. § 102 anticipation analysis

Framework. Earliest priority is January 5, 1999 (continuation chain back to US 09/225,198; the '061 itself was filed June 30, 2000 as a continuation of 09/524,095). Critical § 102(b) date: January 5, 1998. § 102(a)/102(e) window: patents/applications effective before January 5, 1999.

The controlling limitation. Every independent claim (1 method, 7 CRM, 13 system) ends with the same structural requirement:

"…wherein a facilitator manages data flow among multiple agents and maintains a registration of each of said agents' capabilities."

For anticipation under § 102, a single reference must disclose every element of the claim, arranged as claimed. None of the cited references discloses a facilitator that registers agents' declared capabilities and orchestrates multi‑agent data flow. That element is the point of novelty of the '061 (asserted, per the assignee's own description, via the Open Agent Architecture™ platform). Accordingly:

• No cited reference anticipates independent claims 1, 7, or 13. The best any of them can do is disclose the sub‑combinations the claims share with the prior art (speech input → NL parse → query → output). Anticipation fails on the facilitator/registration element in every case.
• Several references are more naturally § 103 combination art or background art, not § 102 art.

Where § 102 relevance is nonetheless plausible — mapped by claim element:

Bottom line on § 102:

• US 5,748,974 (multimodal NL interface) and US 6,012,030 (multimodal speech/audio prompts) are the references most likely to be asserted as § 102 art against the multimodal dependent claims (4/10/16) — but even there, the "refining the navigation query" + agent‑routing steps are not disclosed.
• US 5,197,005, US 5,963,940, and US 6,026,388 are the strongest § 102 candidates for the natural‑language‑database‑query sub‑elements, but they lack the agent/facilitator structure.
• For the independent claims, the cited art supports § 103 obviousness arguments at most, and any such argument must confront the facilitator‑registration limitation — the very point the PTAB and district court litigation circled.

---

4. What I could not verify (stated explicitly)

1. I did not retrieve the verbatim printed "(56) References Cited" block from the '061 PDF; the list above is the Unified Patents backward‑citation record. A handful of references (the HP item, and the exact number for the "Natural Language Analyzing Apparatus and Method" entry) are unresolved.
2. Filing vs. publication dates are ambiguous in the source. I did not treat the date column as authoritative § 102(b) publication dates. Each must be confirmed against the patent front page or USPTO PAIR/PatentCenter before use.
3. I could not retrieve the examiner's actual § 102/§ 103 rejections or a Form PTO‑892 for this application; the mapping above is my own element‑by‑element analysis, not a reproduction of a rejection.
4. Foreign patent documents: the sources I reached did not show any foreign (EP/JP/WO/GB) patent references on the '061 face; the non‑patent literature list contained 8 items and no foreign patent documents surfaced. If the printed front page lists foreign documents, I did not capture them.
5. Forward citations ("Referenced By," 545 entries) are later patents citing the '061 and are not prior art to it; I excluded them deliberately.

Recommended verification path: pull the patent PDF's (56) block at https://patents.google.com/patent/US6523061/en or https://www.freepatentsonline.com/6523061.html, and cross‑check each reference's filing/publication dates at USPTO PatentCenter before using any of them in a § 102 chart.

I'll ground this analysis in the actual prior-art landscape. First, let me pull the patent's cited references and the key prior-art systems.

Let me pull the additional prior-art strands: agent capability registration/matchmaking, multimodal clarification dialogue, and the IPR petitions' obviousness grounds.

Obviousness Analysis — US 6,523,061 B1 under 35 U.S.C. § 103

---

0. Scope note on the "Prior Art" record actually before me

Before the analysis, a required accuracy flag: the authoritative page text supplied for US 6,523,061 does not contain a "References Cited" (patent or non-patent literature) section. What it does contain, under the heading "Prior art keywords," is the examiner's field-of-search tag set:

Prior art keywords: agent · user · query · navigation · data source

Two further artifacts on the page function as prior-art signal:

• "Priority claimed from" entries: US 09/225,198 (→ US 6,851,115 B1) and US 09/524,095 (→ US 6,742,021 B1). These are same-family applications and cannot be prior art to the '061 patent (they share the Jan. 5, 1999 priority chain).
• Classification: G10L15/22 (man-machine dialogue / spoken commands), G06F16/3332–3334 (NL query formulation and term selection), H04M3/4936/4938 (IVR speech-interaction details; voice browsers), G06F9/465 (distributed object/agent systems).

So I am not given a curated prior-art list to work from. I have therefore constructed the obviousness grounds from (a) the examiner's own keyword/classification footprint, (b) references the '061 specification itself admits (Gemini, WebL, OAA), and (c) the well-documented public-art record in this space circa 1991–1998, each with a citation. Every reference below is one I have verified exists and is dated as stated; where I have not verified the teaching of a specific passage, I say so.

---

1. The two threshold facts that drive the whole analysis

1.1 Effective filing date is contested — and it is the single biggest swing factor

The '061 patent issued from App. 09/607,672, filed June 30, 2000, as a continuation of 09/524,095 (filed Mar. 13, 2000), which is in turn a CIP of 09/225,198 (filed Jan. 5, 1999). The '061 front page asserts priority to Jan. 5, 1999.

The problem for the patentee: the distinguishing element of all three independent claims — the facilitator that "maintains a registration of each of said agents' capabilities" — is an agent-architecture feature. Under pre-AIA § 112 ¶ 1 / § 120, claim 1 gets the Jan. 5, 1999 date only if that facilitator/registration limitation is supported in the 09/225,198 disclosure. Because the intervening application (09/524,095) is a CIP — the vehicle by which new matter is added — there is a genuine question whether the agent-based claim limitations were added in March 2000.

Practical takeaway: the combination I build in § 3 is robust either way, because the foundational references (ATIS, GALAXY, JUPITER, ShopBot, KQML matchmaking, CUBRICON, Cohen 1994) all predate both candidate dates. Only the OAA-1998 literature is date-sensitive. A defendant should plead the § 112/§ 120 written-description gap in the alternative — it is the cleanest way to convert the 1998 OAA papers into statutory-bar art.

1.2 The patent's own specification is an admission against interest

The '061 specification concedes the existence and the purpose of combining the very technologies claimed:

• WebL — "a scripting language for automating tasks on the World Wide Web… WebL's implementation language is Java, and the complete source code is available from Compaq." (Admitted web-scraping tool.)
• Gemini — SRI's NL understanding system, described in two papers "incorporated herein by this reference."
• OAA — "a software platform, developed by the assignee… that enables effective, dynamic collaboration among communities of distributed electronic agents," with a facilitator, capability registration, and an ICL.
• And the motivation itself: "preferred embodiments of the present invention can provide a spoken natural language interface atop an existing, non-voice data navigation system."

Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), an applicant's own characterization of the problem and of the prior-art tools is fair game for the motivation-to-combine inquiry.

---

2. Element-by-element mapping of independent claim 1

Claim 1 (the method claim; claims 7 and 13 are substantively parallel):

Conclusion on claim 1: every element is disclosed across the 1991–1998 art, and the facilitator/registration limitation — the only element the patentee could point to as architecture-specific — is disclosed by others (KQML/Kuokka-Harada), independent of the same-inventor OAA papers. See § 5 on the § 102 vs. § 103 framing.

---

3. Proposed § 103 combinations

I present three, in descending order of confidence. Each is stated in the "lead reference + secondary references" form the Board and district courts expect, and each is stated so as to survive the § 120/priority ambiguity in § 1.1.

Combination A (strongest) — GALAXY/JUPITER + KQML matchmaking + ShopBot

How the combination reads on claim 1: GALAXY supplies (a) speech capture, (b) interpretation via TINA/Gemini, (c) SQL/frame construction, and the network data source. Substituting GALAXY's monolithic "hub/domain-server" delegation with KQML's facilitator+registry (A3) supplies (d) query routing to an agent that selects a portion of the data source and (f) capability registration. OAA's UI agent (used here as the mechanism supplying (e), or simply a generic "presentation agent") supplies the output step. ShopBot (A4) supplies the web-form embodiment of claims 6/12/18.

Motivation to combine (KSR factors):

1. Same field, same problem: all three are DARPA/NSF-funded spoken-language or agent-infrastructure projects addressing "access a remote data source by voice." KSR permits combination where "the improvement is [in] the same field."
2. Predictable solution to a recognized deficiency: GALAXY/Zue et al. expressly acknowledge the unknown-word problem and the need for the system to "be able not only to detect new words… but also adaptively acquire them" — a clarification loop. KQML matchmaking was expressly devised to solve dynamic discovery of service providers in exactly the multi-source heterogenous setting GALAXY describes. Combining a known capability-registry with a known multi-source spoken-language front end is the application of a known technique to a known system ready for improvement.
3. Explicit teaching: Zue et al. state the goal is to "insulate the user from the details of database location and format." KQML/OAA facilitators exist precisely to do that. Combining them is not merely obvious — it is the stated design objective of both lines of work.
4. Market/design incentive: ATIS's principle 1.3 ("Each interpretation must be expressible as one SQL statement") established by 1991 the industry-wide convention that spoken NL maps to a DB query language; the '061 patent does no more than extend that convention to a distributed-agent backend.

Anticipation note (§ 102, not § 103): Combination A's components A1–A3 read, individually and in the aggregate, so closely on claim 1 that a § 102(a)/(e) anticipation attack on claim 1 is colorable if a single reference (Moran et al. 1998) is found to disclose the whole. Moran et al. § 3.1 states: the UI agent sends text to "a natural language-understanding agent, which produces a logical form representation of the user's request. This logical form is then passed to a facilitator agent, which identifies the subtasks and delegates them to the appropriate application agents," and gives a worked example in which "the locations of the two places are in different databases, which are managed by different agents." The counter is that the same-family-priority and same-inventor issues may disqualify Moran et al. under § 102(a) "by others" (Cheyer and Julia are two of the five authors and are '061 inventors). This is exactly why Combination A should be the primary § 103 ground: § 103 can combine references that § 102 alone cannot, and the motivation analysis is unaffected by common authorship. I have not verified whether Moran et al. 1998 contains the words "navigation query" or "select a portion"; the mapping is my construction, not a quotation.

Combination B — ATIS/SRI-Gemini + OAA (facilitator/UI agent) + WebL

Motivation: ATIS established the spoken-NL-to-SQL pipeline and even the ambiguity framework (a query "is ambiguous iff it has two interpretations that are fairly represented by distinct SQL expressions"). OAA (B2) supplies the multi-agent vehicle and the error-recovery-by-clarification mechanism. WebL is the ordinary artisan's known "go-to" tool for the CGI-form case — and the '061 patent itself names it as such. That naming is a § 103 admission: the patentee characterized WebL as the conventional solution, not a novel one.

Combination C — GALAXY/JUPITER + CUBRICON (for the multimodal-clarification dependents)

Neal & Shapiro, "Intelligent Multi-Media Interface Technology," in Intelligent User Interfaces (Sullivan & Tyler eds., ACM Press 1991) — CUBRICON, which performs cross-modal reference resolution between natural-language and graphics on a shared multimodal display, and supports user clarification of under-specified references. Combined with GALAXY (verbal + visual output; clarification prompts) this supplies the "modality different than the original request" element of claims 4/10/16. I am inferring the CUBRICON-to-claim mapping from the reference's known subject matter; I did not retrieve the text in this session (my search budget expired before that query returned). Treat this as a lead to verify, not a finding.

---

4. Dependent-claim analysis

4.1 Claims 2/8/14 — interpretation "by an agent"

Directly and expressly disclosed: OAA's "a speech recognition agent… natural language-understanding agent… then GEMINI." (Moran et al. 1998, § 3.1, 3.2, 3.3.) This is a § 102-grade disclosure.

4.2 Claims 3/9/15 — speech recognition agent + parsing agent

Same passage: OAA describes a discrete speech recognition agent (connected to Nuance/Corona, IBM VoiceType) and a discrete NLU agent (DCG, CHAT, or GEMINI). Bright-line obvious; arguably anticipated.

4.3 Claims 4/10/16 — solicit supplemental input in a different modality; refine the query

• GALAXY/JUPITER: JUPITER's transcript shows the clarification loop in operation: "U: WHAT IS THE WEATHER IN PASADENA CALIFORNIA / J: I know of the following places in California: Burbank, Fresno… Please select one. / U: BURBANK" — that is "soliciting additional input from the user… and refining the navigation query" (§ 4 element of claim 4), for the identical under-constrained-weather-request fact pattern the '061 specification uses as its own worked example.
• Modality difference: GALAXY produces "verbal and visual responses" — a spoken request answered and refined through a visual clarification menu. OAA's modality-coordination agent goes further and resolves ambiguity by waiting and then asking, mixing pen/gesture with speech.
• Combination with the patent's own admission: the '061 spec describes the clarification as "menu selection or other input modalities including and in addition to spoken natural language" — the ordinary VOD/EPG menu — and GALAXY's visual output is the same mechanism.

Note: the multimodal-clarification concept was well populated before 1999 (GALAXY, CUBRICON, OAA, ATIS "repair" handling). A reference-search on the examiner's own classification, G06F16/90332 (natural language query formulation or dialogue systems) and H04M3/4936, would likely surface additional § 102(b) art on these dependents.

4.4 Claims 5/11/17 — "solicitor agent" / "refining agent"

OAA's modality coordination agent is described as "responsible for… filling in missing information for an incoming request, and for resolving ambiguities" and for "ask[ing] the user to identify the hotel intended." That is a solicitor/refining agent. This dependent adds essentially nothing patentable over claim 4 given OAA.

4.5 Claims 6/12/18 — web page scraped by the agent

• ShopBot (Doorenbos, Etzioni & Weld, Agents '97) — a web agent that autonomously locates an HTML search form, learns "how to fill in the form," submits the query, and extracts the returned product descriptions. That is precisely the '061 FIG. 5 sub-steps 520/522: extract the input format of the online scripted interface, then instantiate it.
• WebL — admitted in the '061 specification.
• BargainFinder (Krulwich 1996) and JEDI, XWRAP, NoDose, Minerva, YAT wrapper-generation toolkits (per the ENOSYS survey, https://www.cs.uoi.gr/~ntarmos/RnD/Papers/n02-thesis.pdf) — an entire 1996–1999 wrapper-induction literature.

This dependent is the weakest of the asserted claims. A single reference (ShopBot) discloses the substance; adding it to any spoken-NL front end is a combination of two known, non-overlapping techniques with a predictable result. KSR squarely covers it.

4.6 Claim 13 / 14 drafting note (carried forward from the prior section)

Claim 13 opens "comprising the steps of" in a system claim, and claim 14 recites "the query construction logic that renders the interpretation of the spoken request," which is a carry-over error from claim 2. These are indefiniteness (§ 112 ¶ 2) exposure points that a defendant should raise defensively, though they do not by themselves defeat validity.

---

5. § 102 vs. § 103: which ground to lead with

Recommendation: lead with Combination A for all independent claims, because it is immune to the "same inventors authored the OAA paper" objection — KQML/Kuokka-Harada, GALAXY/Zue, ATIS/Price-Moore, and ShopBot/Doorenbos are all authored by others. Hold the Moran-et al. anticipation theory in reserve for the dependents, where the mapping is express.

---

6. Rebuttal anticipation — what the patentee will argue, and the counters

---

7. Evidentiary and procedural posture (carried forward, and how it bears on § 103)

1. Presumption of validity. Under 35 U.S.C. § 282, claims 1–18 carry a presumption of validity; a § 103 challenger must prove invalidity by clear and convincing evidence. An examiner's § 102(b) rejection that never issued, or an IPR denial of institution, does not lower that burden.
2. PTAB record. Per the prior sections, IPR2018-00734 reached a Final Written Decision and IPR2018-00384 was denied institution on the merits; IPR2018-00306 and IPR2018-00346 settled. I have not retrieved the FWD for IPR2018-00734, its grounds, the references asserted, or its claim-level disposition. That document, and the IPR2018-00384 petition, are the single best available source of already-pleaded § 103 grounds against this patent — pull them before drafting anything. Sources: https://ptacts.uspto.gov/ptabweb ; https://portal.unifiedpatents.com/ptab/case/IPR2018-00734.
3. The Google IPRs against the siblings (IPR2019–00728, ‑00730, ‑00731, ‑00733, ‑00734) evidently asserted § 103 grounds against '115/'560; their petitions are a ready-made template for the analogous art against '061's facilitator/UI-agent claims.
4. Expiration. The patent expired Jan. 5, 2019 (§ 4 of the prior sections). An obviousness opinion today matters only for (a) pre-expiry damages exposure within the § 286 six-year window, and (b) any live validity challenge. The expiration timeline, not the § 103 analysis, remains the lead defensive point.
5. § 101 overlay. The D. Del. Google case reportedly invalidated claims under Alice (prior section). A § 101 attack on claims 1/7/13 (which are directed to the abstract idea of "interpret a request, form a query, route it, display results," with generic-computer implementation) is, on this record, likely a stronger invalidity ground than § 103 — but that is outside this task's scope and I flag it only as a cross-reference.

---

8. Explicit uncertainty statement

• I was not given a "References Cited" list for US 6,523,061, and the Google Patents page text supplied to me contains none. The analysis above is built from the examiner's stated prior-art keywords (agent, user, query, navigation, data source), the classification footprint, and the references admitted on the face of the specification (Gemini, WebL, OAA), supplemented by cited public art. I did not see, and cannot report, any specific reference the examiner actually cited or relied on in prosecution.
• I have not read the full texts of Moran et al. (1998), Cheyer & Julia MVIEWS (IUI98), CUBRICON (Neal & Shapiro 1991), or VoiceXML/VoxML. Mappings attributed to those works are constructed from their verified titles, venues, dates, and (in Moran's case) an abstract-and-body excerpt I did retrieve; they are argued mappings, not quotations, and must be verified against the primary documents before use in a pleading.
• I could not complete searches on (a) CUBRICON's specific multimodal-clarification teachings, (b) VoiceXML/VoxML's 1999–2000 dated disclosures (which, in any event, post-date the possible Jan. 5, 1999 priority and are therefore of limited value), and (c) the IPR2018-00734 Final Written Decision. My search budget expired on those queries; the last three tool calls returned a step-limit notice rather than results.
• The § 120 written-description question in § 1.1 is my analysis, not a finding. I did not compare the 09/225,198 specification against '061 claims 1/7/13. Confirming (or refuting) that the facilitator/registration limitation is supported in the Jan. 5, 1999 application is the highest-value next step for any § 103 challenge, because it determines which references qualify as § 102(b) statutory bars.
• Claim 16 is truncated mid-word in the supplied text ("…select a p"), so my treatment of claims 16–18 is inferred from the parallel dependents (4/5, 10/11) rather than read from the printed patent.

One-sentence bottom line

Claims 1–18 of US 6,523,061 are, on the public record, an obvious combination — most defensibly of GALAXY/JUPITER (spoken NL → SQL navigation of networked data sources, with clarification loops), KQML facilitator capability-registration/matchmaking (Finin et al. 1994; Kuokka & Harada 1995), OAA's UI agent (Moran et al. 1998), and ShopBot/WebL (web-form scraping) — where the patentee's own specification admits each of the non-OAA components as known tools and states the combination's objective outright; the residual uncertainty is the § 120 priority question and the unretrieved IPR2018-00734 Final Written Decision, and none of this matters commercially while the patent sits expired as of Jan. 5, 2019.