US 6941273

I'll search for current information on this specific patent, including any 2026 CAFC docket activity.

Let me check specifically for any 2026 litigation or CAFC docket activity involving this patent or its owner.

US Patent 6,941,273 — Analyst Summary

Scope note: I searched for this exact patent number plus 2026 CAFC docket activity. I found no CAFC 2026 docket entry, appeal, or pending litigation tied to US 6,941,273 in the sources I could reach. I also could not load USPTO Patent Center / PTAB dockets directly. So I cannot confirm or deny 2026 court activity — treat "no 2026 CAFC docket found" as a search limitation, not a verified negative. Everything below is grounded in the patent text you supplied (Google Patents, US6941273B1) plus uspto.report and corroborating pages.

Bibliographic data

Chain of title (from USPTO assignment records via Google Patents): inventors → Logic Tree Corporation (recorded 2008-12-03); security agreements to MMV Finance Inc. (2008) and Square 1 Bank (2010, 2013); assignment to Logic Tree, LLC (2013-04-04); later Square 1 Bank → Logic Tree, LLC (2015); Pacific Western Bank security interest (2016). Because the patent expired 2018-10-07, it is in the public domain.

Uncertainty flags:

• Google Patents shows the priority date as 1998-10-07; a Unified Patents listing shows 1998-10-06. The authoritative full text says 1998-10-07 for both priority and filing. Minor, unexplained discrepancy.
• Current-assignee attribution (Logic Tree LLC vs. Logic Tree Corp) is Google's own disclaimer-covered data; verify at the USPTO Assignment database if it matters.

Abstract (verbatim gist)

A voice-enabled system for online shopping provides a voice/telephony interface as well as a text/graphic interface for shopping over the Internet using a browser or a telephone. Customers access an online shop, search for items, select items, and pay by credit card over a phone line or the Internet. A telephony-Internet interface (TII) converts spoken queries into electronic commands for transmission to an online shop or database; markup-language pages returned to callers are parsed to extract selected information, which is reported to callers via audio messaging.

Plain-language overview of the independent claims

Claim 1 — Method of interacting with data applications via spoken queries. A user speaks a query; it is converted to text; an electronic request using that text is sent to a processing device tied to the data application, which searches its memory and returns results as a markup-language page. A multi-part session is established so a user can return to where a prior session ended if interrupted — where "interruption" is defined as a complete termination between the data applications and the user's telephone or computer terminal. The returned page contains audio tags in combination with non-audio tags identifying the information. The system receives results throughout all portions of the multi-part session, parses out the audio-tag-identified portion, generates an audio message for it, and concurrently generates the markup page for selective use. A first portion of the session lets the user order an item from a Web site via computer terminal or telephone; a second portion lets the user directly access information from the first portion via computer terminal or telephone.

Claim 9 — System for browsing and purchasing over the Internet. A telephony-Internet interface connects users via either a graphical interface device (display + input) or a telephony device — simultaneously or sequentially. A memory stores records for online-source items; a processing device accepts electronic and/or spoken queries. The TII converts spoken queries to electronic commands, a multi-part session is established (same interruption/return definition), returned markup pages carry audio and non-audio tags, and the TII parses using the audio tags to produce an audio message while concurrently generating the markup pages. Same first-portion/second-portion and "complete termination" limitations as claim 1.

Claim 10 — Interface connecting telephony users to online information sources. Three modules: a telephone interface module (handles multiple users, does speech recognition and speech↔text conversion), a data presentation module (converts commands into an online communication protocol and parses markup pages), and an online interface module (connects to the sources and retrieves information). Parsed pages contain audio tags plus non-audio tags; extracted information goes to audio users by audio messaging (using the audio tags) and to graphical users by text/graphics. The data presentation module sets up first and second sessions communicated over a computer and a telephone; same multi-part-session, first/second portion, and complete-termination limitations.

Claim 11 — Voice-enabled electronic shopping system. A telephony interface module connects users to online shops over analog, digital, or wireless links (user computer with microphone, telephone, or other telecom device), processes calls, converts spoken queries to commands, and manages multiple user connections. A data presentation module converts commands to HTTP-type electronic commands; a communication module connects to the Internet and manages multiple web-site connections. The online shop's processing device retrieves records and returns markup pages having audio tags in combination with non-audio tags. The DPM parses and produces an audio message for playback. The TIM, DPM, and Internet interface module establish their respective connections independently of each other and relate user connections to web-site connections. Same session/first-portion/second-portion and complete-termination limitations (the text of this claim contains typographical errors, e.g., "mufti-part" and "user commuter").

Claim 12 — Method (variant of claim 1). Receives a spoken query, generates an electronic request in response, sends it to a processing device that searches memory and returns a markup page with audio and non-audio tags. Establishes a multi-part session (same interruption/return definition), parses to the audio-tag-identified portion, generates an audio message, and concurrently generates the markup page. Includes the first-portion/second-portion and complete-termination limitations. Note this claim is textually garbled in the record for the second-portion clause ("via a computer and"), and it does not recite a separate "converting to text" step.

Claim 13 — Telephony-data application interface device. Comprises a telephony-interface module receiving spoken queries, a data presentation module generating an electronic request in response, and a data application interface module connecting the DPM to the data applications. The request goes to a processing device that searches memory and returns a markup page with audio tags plus non-audio tags. The DPM receives the results, parses the audio-tag-identified portion, and generates an audio message. The telephony interface module establishes first and second sessions for computer and telephone communication; the data application interface module establishes the multi-part session. Same first/second portion and complete-termination limitations. Claim 14 (dependent on 13) adds an Internet interface module for multiple web-site connections, multiple-user call handling, and the independent/related connection-management feature.

Technical architecture described (for context)

The specification describes three subsystems: (1) a voice-optimized database/catalogue using "Audio Vectors" (AV) — numeric phoneme-based vectors with a vector-distance search and an algorithm for unequal-length phoneme strings; (2) a voice-enabled electronic shopping cart with transaction, communications, payment, and account modules plus an audio interface directives module that inserts hidden tags into HTML-type pages; and (3) the Telephony-Internet Interface (TII) with a Telephone Interface Module (TIM), Data Presentations Module (DPM), and Internet Interface Module (IIM). Related subject matter is cross-referenced to co-pending application Ser. No. 09/167,486 (which issued as US 6,377,927, "Voice-optimized database system and method of using same"). Notable cited prior art includes GB 2317070, US 5,745,681 (stateless web shopping cart), US 5,884,262, US 5,953,392, and US 6,269,336 (voice browser).

Bottom line: This is a 1998-filed, 2005-issued, now-expired patent (anticipated expiration 2018-10-07) covering speech-driven web/telephony access to online shops and databases, audio-tagged markup parsing, and persistent multi-part sessions. I found no verifiable 2026 CAFC docket activity for US 6,941,273, and I'd recommend direct checks of the CAFC docket and USPTO PTAB/Patent Center before relying on that negative.

Sources: https://patents.google.com/patent/US6941273/en ; https://uspto.report/patent/grant/6941273 ; https://portal.unifiedpatents.com/patents/patent/US-[10096320](/patent/10096320)-B1 (assignee listing)

I'll research litigation involving US Patent 6941273. Let me search multiple sources.

Let me search more targeted litigation sources.

Litigation Search: US Patent 6,941,273

Bottom line: I found no litigation involving US Patent 6,941,273.

What I searched

I ran multiple queries targeting the specific number and its assignee, including:

• US patent 6941273 litigation
• "6941273" patent lawsuit
• Logic Tree LLC patent litigation 6941273
• "6941273" OR "6,941,273" patent infringement case PACER
• Logic Tree LLC v. patent case number district court

The only hits for the literal string "6941273" were unrelated to this patent (a Brazilian company registration "RNK REPRESENTACOES LTDA" and an ELKO electrical connector product datasheet). All patent-related hits were the Google Patents page for the patent itself and its full-text mirrors — none were docket, complaint, or case-list records.

Why this is a plausible result

US 6,941,273:

• Is titled "Telephony-data application interface apparatus and method for multi-modal access to data applications"
• Lists inventors Masoud Loghmani and Fred F. Korangy
• Was filed 1998-10-07 and granted 2005-09-06
• Is recorded as "Expired – Lifetime" (anticipated expiration noted as 2018-10-07)

The patent's recorded history in the Google Patents "reassignment" log shows only security interests and assignments, not litigation:

• 2008-11-13/14 — MMV FINANCE INC. security agreements (assignor Logictree Corporation)
• 2008-12-03 — assignment to Logic Tree Corporation (from Korangy and Loghmani)
• 2010-07-20 and 2013-01-29 — SQUARE 1 BANK security interests
• 2013-04-04 — assignment to Logic Tree, LLC
• 2015-02-13 — assignment of Square 1 Bank interest to Logic Tree, LLC
• 2016-05-04 — PACIFIC WESTERN BANK security interest

These are financing/ownership instruments, not infringement suits.

Important caveats

1. Confidence level: moderate. The web searches I was able to complete did not directly return a Unified Patents litigation caselist page, Docket Navigator, or PACER docket for this patent. My searches were also truncated by tool limits before I could exhaust litigation-specific databases (Unified Patents, Justia Dockets, CourtListener, Stanford NPE database).
2. Family context (search-behavior note, not an assertion about '273 itself): The '273 patent is part of an extended family — including US 7,283,973, US 6,377,927, and published app US 2006/0026206A1 — through inventor Loghmani / Logic Tree. If you are investigating assertion activity by this patent holder, cases filed on family members would be the place to look next; but per your instruction I have not reported those as if they were litigation on '273.
3. A negative search result is not proof of no litigation. Cases can be sealed, settled quietly, filed in state court, or indexed only in paywalled dockets.

Recommended next steps to confirm

• Unified Patents litigation portal: https://portal.unifiedpatents.com/litigation/caselist — search "6941273" and "Logic Tree"
• CourtListener / RECAP: https://www.courtlistener.com — search patent number and "Logic Tree LLC"
• USPTO Patent Center "Litigation" tab for application 09/167,485
• PACER docket search across districts where Logic Tree, LLC is a party
• Docket Navigator (subscription) for a definitive NPE-assertion report

If you can supply a case number or court, I can pull the specific complaint, filing date, and status details.

I'll verify the structured block's default against live sources before writing anything up.

Let me dig into the patent owner's litigation/PTAB history and any Unified Patents activity.

PTAB / AIA Trial Landscape — US 6,941,273 ("Telephony-data application interface apparatus and method for multi-modal access to data applications")

Proceedings overview

Total AIA trial proceedings on file: 0. The canonical ODP "PTAB proceedings on file" block returns no AIA trial proceedings for US 6,941,273 — no IPRs, no PGRs, and no CBM reviews — and my live searches surfaced no petition, institution decision, FWD, or appeal directed at this patent. The defensive posture is therefore "the patent was never tested at the PTAB" rather than either of the two scenarios in the brief: there is no claim-cancellation win to lean on, but equally there is no Board-sanctioned validity holding to overcome. Notably, the structured record shows the patent reached its anticipated expiration on 2018-10-07 and is marked Expired – Lifetime, which materially changes what a defendant would even want from an IPR.

(Caveat on method: I searched for petitions by patent number, by patent owner, and by subject matter. Search results for "Logic Tree" are contaminated by Logan Tree, LP — a distinct, unrelated owner (Brann step-count/activity patents, e.g. IPR2022-00037 on US 6,059,576) that is not the owner of this patent. None of those proceedings involve 6,941,273. My patent-number searches for the '273 also returned only unrelated patents (e.g. US 5,333,273, Samsung v. Wistron), which I have excluded rather than mis-reported.)

Proceedings

None to report. Per the constraint against fabricating proceeding numbers, I am not listing any IPR20xx-##### sections — there are no proceedings to populate them. Summary of what is and is not on file:

Strategic summary

Claim status. Because no AIA trial ever reached a Final Written Decision, no claim of 6,941,273 has been canceled by the PTAB. Every claim — independent claims 1, 9, 10, 11, 12 and 13, and dependents 2–8 and 14 — is untested at the Board, not "sustained." That distinction matters: an untested claim carries no estoppel benefit for a defendant and no res judicata effect, but it also means a defendant cannot shortcut invalidity by pointing to a prior cancellation. Any challenge must be built from scratch on the face of the claims and the specification.

Estoppel landscape. Because no proceeding was instituted, § 315(e)(2) estoppel does not attach to anyone — there is no petitioner, no privy, and no real-party-in-interest whose grounds were "raised or reasonably could have been raised." For a defendant being asserted against today, the entire prior-art field remains open: § 102 and § 103 combinations, § 112 written-description/enablement attacks on the "multi-part session" and "audio tags in combination with non-audio tags" limitations, and § 101 eligibility arguments. Conversely, there is no downside to filing, since a fresh petitioner faces no estoppel and no prior adverse panel finding on these claims.

Pattern signals. No petitioner has ever filed against this patent — not once, let alone a serial filer. There is no defensive aggregator (Unified Patents or similar) in the chain; I found no evidence of any such filing and I am not going to infer one. The assignment history in the structured record is a pure financing/lender chain (MMV Finance security agreements 2008, Square 1 Bank 2010/2013, Pacific Western Bank 2016) ending in Logic Tree, LLC — i.e., the patent was collateralized repeatedly, which is consistent with an active monetization effort rather than a defensive-holding posture. Note however that the patent expired in 2018, which typically extinguishes the economic motive for both assertion and for third-party IPR challenges.

One observation flagged, not asserted. The claim text as published on Google Patents contains a recurring "multi-part session … allows a user to return to where a previous session was terminated if the previous session was interrupted" limitation in independent claims 1, 9, 10, 11, 12 and 13. That language does not read like a 1998 drafting artifact, which hints at a possible later amendment (certificate of correction or ex parte reexamination certificate). I could not verify any reexamination or certificate-of-correction record, and I am not claiming one exists — but before relying on the published claim text, a defendant should pull the USPTO PatentCenter "Transactions" / "Reexamination Certificate" tab to confirm they are reading the operative post-issuance claim set. This is a documents-check, not a finding.

Recommended next steps

• Do not expect an IPR-based defense to already exist. There is nothing to cite. If a demand letter arrives asserting 6,941,273, the invalidity case starts at zero. The absence of any petition in 20+ years — on a patent whose family was apparently monetized (see the collateral-assignment chain) — is itself a weak signal that the claims were not an attractive IPR target, but it is not evidence of strength, because the patent expired 2018-10-07 and any post-expiration assertion is limited to pre-expiration damages.
• Confirm the operative claim set before anything else. Check PatentCenter for a reexamination certificate or certificate of correction, given the anomalous "multi-part session" language. Also confirm the expiration calculation and any terminal disclaimers, and check the two family members for cross-cutting estoppel-relevant findings: US 6,377,927 ("Voice-optimized database system and method of using same," same inventors/priority date), US 7,283,973 (CIP; US 10/282,239), and US 7,203,142 ("System for providing multi-phased, multi-modal access to content through voice and data devices"). I did not verify whether any IPR was filed against those siblings — a quick PTAB E2E search on each is the highest-value next click, since a sibling FWD construing shared claim language would be persuasive authority even without estoppel.
• If you file, treat the § 112 and § 101 angles as co-equal with § 102/§ 103. The claims mix functional "generating … concurrently with said audio message" steps with structural modules, and the specification's audio-vector material is largely directed to a different patent in the family (the '927). A § 103 attack grounded on the 1996–1998 voice-browser art cited on the face (US 5,953,392 Netphonic; US 5,884,262 Bell Atlantic; US 5,761,280 8×8; US 6,269,336 Motorola voice browser; GB 2 317 070 IBM voice processing/Internet) plus a conventional web shopping-cart reference (US 5,745,681 Sun, stateless shopping cart) is the natural starting combination.
• Trial-stage milestones: not applicable. No institution decision deadline, oral hearing, or statutory 1-year FWD date is running, because no proceeding exists. If you do file now, the normal clock would begin at the institution decision (FWD due within 12 months of institution under § 316(a)(11)), subject to the Board's current extension practices.
• Verify independently. Canonical records: the PTAB E2E / PTAB Center portal and the USPTO PTAB Decisions页 are the authoritative sources; the Google Patents page for US6941273B1 is where the claim text and assignment history quoted above come from. My conclusion is that both are silent on AIA trials for this patent — if your own E2E query returns a proceeding number, treat that as controlling over this write-up and re-scope accordingly.

I'll research the assignment chain and assignment-record details. Let me run several searches in parallel.

Let me dig into the assignment records, the entity history, and any assertion activity.

Data-access note (read first)

I was unable to retrieve the underlying USPTO Assignment Center records (reel/frame numbers and correspondents of record) for this patent. Direct queries against the assignment databases returned no usable records in my available searches, and the Google Patents legal-events log — the authoritative source supplied in this analysis — records dates, conveyance type, assignor and assignee only. It does not expose reel/frame or correspondent.

Accordingly:

• Every date, conveyance type, assignor and assignee below is sourced from the Google Patents reassignment log for US 6,941,273 (https://patents.google.com/patent/US6941273/en), which mirrors the USPTO assignment record set.
• No reel/frame number or correspondent attorney/firm is stated anywhere below, because I could not obtain one. Any reel/frame in this report would be fabricated, so there are none. The "repeat correspondent" signal is therefore graded unclear, not absent — this is a data gap, not a clean finding.
• Verify at: https://assignmentcenter.uspto.gov/ (search patent 6941273) and the legacy index https://assignment.uspto.gov/patent/index.html.

Inventors

Pattern flagged — 10-year assignment gap. The inventors' assignment to LOGICTREE CORPORATION was not recorded until 2008-12-03, roughly 122 months after the 1998-10-07 filing date. This is the inverse of the "inventors depart within 12 months" pattern, but it is equally notable: the patent sat with unrecorded individual ownership through issuance (2005-09-06) and for three further years. Practical effect: any party searching the record between 1998 and 2008 would find no recorded corporate owner.

Entity-timing wrinkle. Korangy's own published biography states he "founded LogicTree Corp." in 1999 (https://imiranian.com/fred-korangy/), yet the corporate assignee recorded in 2008 is LOGICTREE CORPORATION in connection with a patent filed 1998-10-07. If the 1999 incorporation date is correct, the 1998 applications were filed pre-incorporation and the corporate chain necessarily rests on later confirmatory paperwork — consistent with the recorded 2008 instrument. This is unresolved on the public record.

Co-filed sibling. US 6,377,927 ("Voice-optimized database system and method of using same"), same 1998-10-07 priority date, issued 2002-04-23 to Masoud Loghmani individually — i.e. the sibling patent's recorded owner is the inventor personally, not the corporation. Both are Loghmani-family assets.

Original assignee

Original assignee on the issued patent: none recorded — "Individual." The patent issued 2005-09-06 with the inventors as the unrecorded beneficial owners.

First corporate owner: LOGICTREE CORPORATION (recorded 2008-12-03).

• Line of business: interactive voice response (IVR) software for the transit and 511 markets — multi-modal communications platforms delivering location-based real-time information via voice, web, SMS, social, email and mobile. HQ Franklin, MA. PitchBook lists it as founded 1995, formerly known as Kasteris (https://pitchbook.com/profiles/company/54058-69).
• Did it ship a product embodying the claims? Yes, on the available evidence. This is an operating IVR vendor with deployed government/transit solutions, not a licensing vehicle. Korangy's biography claims deployments at state and federal agencies in Florida, Virginia, New York and Arizona and at the Pentagon, and >100% annual revenue growth through the decade.
• Status: acquired/merged. PitchBook records a merger/acquisition with Gold Systems executed 2014-06-26; pitchbook lists 4 employees and 7 financing rounds. Korangy departed as CEO in 2007.
• Do not treat "Logic Tree, LLC" as a new entity. The 2013-04-04 recording transfers from LOGICTREE CORPORATION to LOGIC TREE, LLC — facially a corporate form conversion of the same operating business (confirmed by the LLC appearing on the 2013-01-29 Square 1 Bank security agreement before the confirming assignment was recorded).

Contradictions in the public profile (flagging, not resolving). (a) Founding year 1995 (PitchBook) vs. 1999 (Korangy bio) vs. a 1998-10-07 priority filing. (b) Capital raised $2M (PitchBook) vs $14M by 2007 (Korangy bio). These figures are not reconcilable from the sources I could reach and should not be relied on.

Assignment timeline

All entries are recorded dates from the Google Patents reassignment log. Execution dates were not retrievable. Reel/frame: not retrieved (see data-access note).

• Executed: unknown / recorded 2008-11-13 — Reel not retrieved

  • Conveyance: Security Agreement
  • Assignor: Logictree Corporation
  • Assignee: MMV FINANCE INC.
  • Correspondent: not retrieved
  • Context: securitization — lender takes a security interest in the patent as collateral.

• Executed: unknown / recorded 2008-11-14 — Reel not retrieved

  • Conveyance: Security Agreement
  • Assignor: Logictree Corporation
  • Assignee: MMV FINANCE INC.
  • Correspondent: not retrieved
  • Context: securitization — second MMV recording on the next calendar day against the same assignor/assignee; likely a companion instrument or a corrective/duplicate filing. This should be checked at reel/frame level, because two same-party security recordings 24 hours apart is the type of entry that sometimes masks a later correction or release.

• Executed: unknown / recorded 2008-12-03 — Reel not retrieved

  • Conveyance: Assignment of Assignors' Interest
  • Assignor: KORANGY, FRED F. and LOGHMANI, MASOUD (individuals)
  • Assignee: LOGIC TREE CORPORATION
  • Correspondent: not retrieved
  • Context: confirmatory/clean-up assignment — perfects the inventors' transfer to the corporation ten years post-filing, recorded three weeks after the MMV security agreements it had to support. Sequence matters: the lender's collateral was recorded before the grantor's own title was recorded.

• Executed: unknown / recorded 2010-07-20 — Reel not retrieved

  • Conveyance: Security Interest
  • Assignor: Logictree Corporation
  • Assignee: SQUARE 1 BANK
  • Correspondent: not retrieved
  • Context: securitization — refinancing/replacement of the MMV facility with a new venture-debt lender.

• Executed: unknown / recorded 2013-01-29 — Reel not retrieved

  • Conveyance: Security Agreement
  • Assignor: LOGIC TREE, LLC (note: the LLC, not the Corporation)
  • Assignee: SQUARE 1 BANK
  • Correspondent: not retrieved
  • Context: securitization — the successor LLC re-pledges the asset to the existing lender; recorded 65 days before the assignment that put record title in the LLC.

• Executed: unknown / recorded 2013-04-04 — Reel not retrieved

  • Conveyance: Assignment of Assignors' Interest
  • Assignor: LOGICTREE CORPORATION
  • Assignee: LOGIC TREE, LLC
  • Correspondent: not retrieved
  • Context: internal reorg / corporate conversion — same business, new corporate form; consistent with an operating-company restructuring, not a sale to a third party.

• Executed: unknown / recorded 2015-02-13 — Reel not retrieved

  • Conveyance: Assignment of Assignors' Interest (recorded with assignor = the secured party)
  • Assignor: SQUARE 1 BANK
  • Assignee: LOGIC TREE, LLC
  • Correspondent: not retrieved
  • Context: release / termination of security interest — the secured party dropping off title and the asset returning to the operating company upon payoff or facility wind-down. The recitation "Assignment of Assignors' Interest" on a lender-to-borrower entry is the standard way a security-interest termination surfaces in this log.

• Executed: unknown / recorded 2016-05-04 — Reel not retrieved

  • Conveyance: Security Interest
  • Assignor: LOGIC TREE, LLC
  • Assignee: PACIFIC WESTERN BANK
  • Correspondent: not retrieved
  • Context: securitization — fourth lender in eight years takes a security interest in the same asset; consistent with a working-capital/venture-debt facility at an operating company.

• 2018-10-07 — Anticipated expiration. Status: Expired – Lifetime. No assignment is recorded after the 2016 security interest; the patent is now off the assertion table regardless of chain.

Family/continuation events (not assignments, but they explain the chain): priority to US 10/282,239 (CIP → US 7,283,973, granted 2007-10-16, assignee Logic Tree Corporation) on 2002-10-29, and to US 11/143,941 (continuation → US 2006/0026206 A1, abandoned) on 2005-06-03. If you are tracking this family, the same financing chain should be reconstructed separately for US 7,283,973 and for Loghmani's US 6,948,129 and US 2007/0174386 A1.

Timeline diagram

timeline
    title Ownership of US 6941273
    1998 : Filed by Loghmani and Korangy as individuals
    2005 : Patent granted to inventors of record
    2008 : Inventors assign to Logictree Corporation
         : MMV Finance security agreement recorded
         : Second MMV Finance security agreement recorded
    2010 : Square 1 Bank security interest recorded
    2013 : Square 1 Bank security agreement from Logic Tree LLC
         : Logictree Corporation assigns to Logic Tree LLC
    2014 : LogicTree merges with Gold Systems
    2015 : Square 1 Bank interest released back to Logic Tree LLC
    2016 : Pacific Western Bank security interest
    2018 : Patent expires

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT.
The only ownership transfer to a new entity is 2013-04-04 (LOGICTREE CORPORATION → LOGIC TREE, LLC). The evidence contradicts a shell characterization: the assignee LLC existed and was pledging assets to Square 1 Bank on 2013-01-29, before the confirming assignment recorded 2013-04-04; the corporate family is publicly described as an IVR software vendor in Franklin, MA with government/transit deployments, a 2014 merger with Gold Systems, and ~4 employees (https://pitchbook.com/profiles/company/54058-69). This reads as a corporate form conversion, not a transfer to a licensing-only vehicle. There is no recorded "IP / Patents / Licensing / Ventures" assignee anywhere in the chain, and no registered-agent-service address is visible in the accessible record.

2. Known asserter in the chain — NOT PRESENT.
Neither LOGICTREE CORPORATION, LOGIC TREE, LLC, MMV FINANCE INC., SQUARE 1 BANK, nor PACIFIC WESTERN BANK appears in the asserter lists I compared against (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities). The banks in the chain are lenders recording collateral, not owners acquiring the asset for assertion. Caveat: I could not complete a full RPX/Unified Patents directory sweep within available tool limits.

3. Repeat correspondent across the chain — UNCLEAR (data gap, not a clean negative).
Correspondents of record were not retrievable, so I cannot test whether one attorney or firm ran this chain. Two indirect observations, neither of which is a finding: (a) the chain spans four different lenders (MMV 2008, Square 1 2010/2013, Pacific Western 2016), which typically means lender-side counsel changed each time rather than one recurring NPE lawyer; (b) a "single repeat attorney across every recording" pattern is exactly what would be invisible in the data I have. Do not treat this as cleared.

4. Cascading transfers — WEAKLY PRESENT, non-troll flavor.
Two clusters of same-year, same-party filings: 2008-11-13 and 2008-11-14 (MMV, 24 hours apart) and 2013-01-29 and 2013-04-04 (Square 1 / LLC, 65 days apart). Both clusters are lender/restructuring related, and the entities do not share an anonymous registered-agent address. I do not read this as an NPE cascade, but the 2008 pair should be reconciled at reel level.

5. Pre-litigation transfer — NOT ASSESSABLE.
No litigation naming this patent was found (consistent with the earlier litigation section of this analysis — no contradiction). With no suit date, there is nothing to measure the 2013, 2015 or 2016 recordings against. Note the last recorded instrument is a security interest from 2016-05-04, and the patent expired 2018-10-07, so any assertion window closed.

6. Bankruptcy fire-sale — NOT PRESENT.
No Chapter 7/11, no receiver, no court-supervised patent sale appears in the record. The 2015-02-13 Square 1 Bank entry is a security-interest termination back to the operating company, not a distressed liquidation. (Square 1 Bank itself failed in July 2015, but after that February recording, and the bank's failure is not an event in this patent's chain.)

7. Privateering — NOT PRESENT.
No transfer from Logic Tree to an NPE asserting on its behalf; the ownership never leaves the Logic Tree corporate family.

8. Defensive aggregator — NOT PRESENT.
The chain does not terminate at RPX, AST, LOT, Unified Patents or OIN. It terminates at the operating company, then expires.

Verdict

Insufficient data — with the affirmative finding that no NPE signal is present in the recorded chain.

Justification: the recorded sequence is inventors → Logictree Corporation (2008-12-03) → Logic Tree, LLC (2013-04-04), punctuated only by lender security instruments (MMV 2008-11-13/14; Square 1 Bank 2010-07-20 and 2013-01-29; Pacific Western Bank 2016-05-04) and a security-interest release (Square 1 Bank → Logic Tree, LLC, 2015-02-13). No assignee matches a published asserter list, no licensing-only shell appears, no bankruptcy or court-ordered sale appears, and the patent expired 2018-10-07 with no litigation found. That chain is the signature of a venture-debt-financed operating software vendor, not a troll. The verdict is "insufficient data" rather than a clean exculpation solely because two inputs are missing: reel/frame and correspondent of record were not retrievable, so the recurrence-of-correspondent test (the single most probative NPE tell) could not be run — and the 2008-11-13/14 duplicate MMV pair remains unreconciled.

Two items worth a targeted follow-up if you have Assignment Center access: (i) the correspondents on the 2008-12-03 inventor-to-corporation assignment and the 2013-04-04 corporation-to-LLC assignment — if the same attorney appears on both, and on the sibling US 7,283,973 chain, that is a portfolio-management footprint worth mapping; (ii) the reel/frame split of the 2008-11-13 and 2008-11-14 MMV recordings.

Verify: https://assignmentcenter.uspto.gov/ · https://assignment.uspto.gov/patent/index.html · https://patents.google.com/patent/US6941273/en

I'll verify the patent record and key cited references via live search, then map each citation to the claims.

Prior-Art Analysis — US Patent 6,941,273

Patent under review: US 6,941,273 B1 — "Telephony-data application interface apparatus and method for multi-modal access to data applications"
Inventors: Masoud Loghmani; Fred F. Korangy
App. No.: 09/167,485 · Filed: 1998-10-07 · Granted: 2005-09-06 · Status: Expired–Lifetime
Source of record: https://patents.google.com/patent/US6941273/en (fetched 2026-09-24)

Note on the earlier section: the previously generated Litigation summary concluded "no litigation found" and recommended checking Unified Patents / CourtListener / PACER. That negative result is not contradicted by anything in this prior-art review. My USPTO-oriented searches surfaced only unrelated PTAB records (e.g., an IPR concerning US 9,998,607 and a Steagall/Baran/Bernstein petition) — none naming US 6,941,273. No correction needed; the two sections are consistent.

1. Method and scope

• I searched for the literal identifier 6941273 and confirmed the record at the Google Patents full-text mirror above (abstract, description, claims 1–14, and the "References Cited" tables).
• I then took each patent reference cited on the face of the '273 patent (24 references total — the table the patent labels "Patent Citations (24)"; Google's page shows 23 in one roster plus US 5,867,816 A) and analyzed it against the '273 claim set.
• Statutory framework: the '273 is a pre-AIA patent (filed 1998-10-07, before the AIA's 2013 first-inventor-to-file change), so 35 U.S.C. § 102(a), (b), and (e) govern. I assume, absent contrary evidence, an invention date no earlier than the 1998-10-07 filing date. Critical dates:
  • § 102(b): reference patented/published more than one year before filing → before 1997-10-07
  • § 102(a): reference patented/published before the invention date → before ~1998-10-07
  • § 102(e): US patent granted (or US application publication) on an application filed in the US before 1998-10-07

Important interpretation rule applied: "Anticipation" under § 102 requires a single reference disclosing every limitation of a claim. Where a reference supplies only some elements, I say so and identify it as § 103 (obviousness) evidence rather than anticipation. I do not auto-correct any number or date; where a cited reference's dates look inconsistent with prior-art status, I flag it explicitly.

2. The claim set, in shorthand

The feature that recurs in every independent claim (1, 9, 10, 11, 12, 13) is the "multi-part session … allowing a user to return to where a previous session was terminated if the previous session was interrupted," with interruption defined as "a complete termination between said data applications and at least one of said telephone and said computer terminal." Keep this in view — it is the limitation that most defeats clean anticipation by the cited art.

3. Summary table — every reference cited on the face of US 6,941,273

4. Reference-by-reference analysis

A. The "telephone ↔ Internet ↔ speech markup" cluster (most material)

US 5,953,392 A — Rhie et al., "Method and apparatus for telephonically accessing and navigating the internet"

• Full citation: US 5,953,392 A; filed 1996-03-01; granted 1999-09-14; Netphonic Communications, Inc. (now Intellectual Ventures I LLC via continuations US 6,366,650 / US 7,231,020 / US 8,600,016).
• § 102(e) (application filed 1996-03-01, before 1998-10-07).
• Description: A user dials in by telephone; the system retrieves web pages, converts page content from text to speech, audibly signals hyperlink selections, and lets the user choose links via DTMF; the next page is fetched and delivered by voice/fax/e-mail. This is a telephone↔internet browsing architecture with server-side HTML retrieval, page parsing, TTS, and DTMF control.
• Claims potentially affected: This is the single closest reference to the preamble + core of claims 1, 10, 12, 13 and to dependent claims 2, 3. It discloses: telephony interface ☑, retrieving markup pages ☑, converting to speech ☑, DTMF/voice menu selection ☑, parsing pages to locate links/info ☑. It does not clearly disclose the '273's audio-tag-in-combination-with-non-audio-tags markup scheme, nor the multi-part session with return-after-interruption limitation. → Potential § 102 support for the environment of claims 1/10/12/13 and for claim 2's DTMF/HTTP steps; strongest as a § 103 primary reference rather than full anticipation of the independent claims.
• Source: https://patents.google.com/patent/[US5953392A](/patent/US5953392A)/en

US 6,240,448 B1 — Innelinski (Imielinski) et al., "Method and system for audio access to information in a wide area computer network"

• Full citation: US 6,240,448 B1; inventors Tomasz Imielinski, Aashu Virmani; priority US provisional 60/009,153 filed 1995-12-22; PCT/US96/20409 filed 1996-12-20; § 371 date 1999-05-07; granted 2001-05-29; Rutgers, The State University of New Jersey.
• § 102(e) (international/US filing well before 1998-10-07).
• Description: "Audio-enabled pages" (ATML) served by an audio web server; a caller reaches a server (e.g., 1-800-AWEB), is offered a menu of options, uses touch-tone (and optionally speech) to traverse links, the server fetches the page via a network protocol (HTTP/URL) and converts text to audio via TTS, recursively treating links.
• Claims potentially affected: Directly reads on dependent claim 2 (pre-recorded/list of options + DTMF + HTTP-type fetch) and on much of the environment of claims 1, 10, 12, 13; also claim 3 (home page fetch; input supplied to pages) and claim 5 (TTS + recorded audio). Critically, the "audio-enabled pages" scheme is a parallel mark-up designed for audio, which is not the same as the '273's claim to audio tags placed in ordinary markup pages alongside non-audio tags. → § 103 primary reference; supports claims 2, 3, 5, and the general (but not the dual-tag / multi-part-session) limitations.
• Source: https://patents.google.com/patent/US6240448

US 6,269,336 B1 — Ladd et al., "Voice browser for interactive services and methods thereof"

• Full citation: US 6,269,336 B1; filed 1998-10-02 (US 09/165,487), priority provisionals 60/094,032 and 60/094,131 filed 1998-07-24; granted 2001-07-31; Motorola, Inc.
• § 102(e) — its application was filed five days before the '273's 1998-10-07 filing (and its provisionals predate the '273 filing by ~2.5 months).
• Description: A markup-language voice browser for telephone/internet information retrieval. Discloses a markup document with dialog/step/prompt/input elements identified by markup tags, a voice browser with network fetcher + parser + interpreter + state machine, TTS output, dynamic grammar generation from a pronunciation dictionary and phonetic rules, and — notably — a BARGEIN attribute allowing the service to be interrupted by user input, with help/error/cancel elements and next-step navigation. Supported data types include forms posted to a server (HTTP get/post).
• Claims potentially affected: Very relevant to claim 1/12 parsing of a tagged markup document + audio playback; to claim 4 (parser/interpreter extracting data); claim 5 (TTS + stored audio); claim 8 (interrupting with an audio command); and the general architecture of claims 10/13. The interruption here is a prompt barge-in within a dialog, which is not the '273's "return to where the previous session terminated after a complete termination." → Strong § 103 reference (arguably the best "tagged-markup voice browser" art); does not, on its face, disclose the multi-part-session/interruption-return limitation.
• Source: https://uspto.report/patent/grant/[6269336](/patent/6269336) and https://patents.google.com/patent/[US6269336B1](/patent/US6269336B1)

US 5,884,262 A — "Computer network audio access and conversion system"

• Full citation: US 5,884,262 A; filed 1996-03-28 (per Google front page; § 102(e) date 1996-03-27/28); granted 1999-03-16; Bell Atlantic Network Services, Inc.
• § 102(e).
• Description: Audio access to, and conversion of, computer-network content (i.e., retrieving web content and rendering it as audio to a caller). Core to the "convert markup pages to audio for telephone users" concept.
• Claims potentially affected: environment of claims 1, 10, 12, 13; claim 5 (conversion of extracted data to speech). → § 103 reference; no showing of the dual-tag markup or multi-part session.

US 6,101,473 A — "Using speech recognition to access the internet, including access via a telephone"

• Full citation: US 6,101,473 A; filed 1997-08-08; granted 2000-08-08; Board of Trustees, Leland Stanford Jr. University.
• § 102(e).
• Description: Using speech recognition to search/access internet resources, incl. telephone access.
• Claims potentially affected: environment/core of claims 1, 12, 13 (spoken query → internet search). → § 103 reference; does not supply the markup dual-tag or multi-part-session elements.

US 5,915,001 A — "System and method for providing and using universally accessible voice and speech data files"

• Full citation: US 5,915,001 A; filed 1996-11-14; granted 1999-06-22; Vois Corporation (also listed as Uppaluru reference in US 6,269,336).
• § 102(e).
• Description: Telephone-accessible voice/speech data files for universal access to web-type content.
• Claims potentially affected: environment of claims 1, 10, 12, 13. → § 103 reference.

US 6,335,928 B1 — "Method and apparatus for accessing and interacting an internet web page using a telecommunications device"

• Full citation: US 6,335,928 B1; filed 1997-06-06; granted 2002-01-01; Lucent Technologies, Inc.
• § 102(e) (filed 1997-06-06 < 1998-10-07).
• Description: A telecommunications device accessing and interacting with an internet web page (browse + submit).
• Claims potentially affected: environment of claims 1, 10, 12, 13; and the "interaction" (input back to a page) aspect of claim 3. → § 103 reference.

GB 2317070 A — IBM, "Voice processing/internet system"

• Full citation: GB 2317070 A; priority 1996-09-07; published 1998-03-11; IBM.
• § 102(a) (foreign printed publication before the invention date).
• Description: Voice processing combined with internet access.
• Claims potentially affected: environment/§ 103 for claims 1, 10, 12, 13. As a British application publication it is a printed-publication reference (not a US 102(e) reference). No showing of the markup dual-tag or multi-part session. → § 103 reference.

US 5,761,280 A — 8×8, Inc., "Telephone web browser arrangement and method"

• Full citation: US 5,761,280 A; filed 1996-09-04; granted 1998-06-02; 8×8, Inc.
• § 102(a).
• Description: Arrangement/method for a telephone to browse the web.
• Claims potentially affected: environment of claims 1, 10, 12, 13. → § 103 reference.

US 2001/0033564 A1 — Hickman, "Method and apparatus for accessing a wide area network"

• Full citation: US 2001/0033564 A1; priority/filing 1996-05-01; published 2001-10-25 (publication post-dates '273 filing, but the application's US filing date controls the § 102(e) date).
• Description: Accessing a wide-area network (internet access architecture).
• Claims potentially affected: environment of claims 1, 10, 12, 13. → § 103 reference.

B. Electronic-shopping / shopping-cart references (claims 6, 9, 11)

US 5,715,314 A — "Network sales system" (Open Market, filed 1994-10-24, granted 1998-02-03; § 102(a)). Discloses network-based selling, order/payment flow and cart-like tracking. Bears on claim 6 (cart tracking + purchase) and the "purchase items" aspect of claims 9, 11, but discloses no voice/telephony/markup/audio elements. → § 103 support for the shopping features; cannot anticipate any independent claim.

US 5,745,681 A — "Stateless shopping cart for the web" (Sun Microsystems, filed 1996-01-11, granted 1998-04-28; § 102(a)). A web shopping cart that maintains/logs state across pages without a persistent server session. Bears on claim 6 and the cart/session mechanics of claims 9, 11. Note an interesting tension: "stateless" cart is closer to the problem the '273 complains about (cart data removed after inactivity) than to the '273's multi-part-session solution. → § 103 support for the shopping-cart element only.

US 5,745,754 A — IBM, "Sub-agent for fulfilling requests of a web browser using an intelligent agent and providing a report" (filed 1995-06-07, granted 1998-04-28; § 102(a)). Web-browser proxy/agent that fetches and reports page content — relevant to the "parsing a markup page to extract selected information" element of claim 4 and the DPM-style mediation of claim 10. → § 103 reference; no speech/audio.

US 5,291,583 A — Racal-Datacom, "Automatic storage of persistent ASN.1 objects in a relational schema" (filed 1990-12-14, granted 1994-03-01; § 102(b), the only reference in the set that is more than one year before filing). Database-persistence teaching; bears only on the generic "memory device … storing records" element of claims 9, 11 and on claim 6's stored transaction data. → Background/§ 103 support; discloses nothing of the inventive combination.

C. Speech-recognition / phoneme references (claim 7)

All of the following pre-date 1998-10-07 and are directed to speech-recognition internals. They supply the "convert a spoken query to phonemes / phoneme-space processing" concepts that underlie claim 7, but none discloses the '273's specific step of searching a database by stored "audio vectors" with a distance/nearest-match algorithm:

• Claims potentially affected: claim 7 (phoneme conversion; audio-based search) and the AVV/pronunciation-variant discussions in the specification. Individually these are § 103 support for the phoneme-conversion half of claim 7; the "searching said database using audio vectors based on said phonemes" limitation is not squarely anticipated because these references address recognition/verification, not a database whose records are indexed by stored audio vectors and searched by vector distance.

D. IVR/menu and bridge references (claims 2, 10)

US 5,802,526 A — Microsoft, "System and method for graphically displaying and navigating through an interactive voice response menu" (filed 1995-11-15, granted 1998-09-01; § 102(a)). IVR menu presentation/navigation. Bears on claim 2 (pre-recorded menu; DTMF/voice selection). → § 103 support for the menu limitations; no internet/markup element.

US 6,134,235 A — AT&T, "POTS/packet bridge" (filed 1997-10-08, granted 2000-10-17; § 102(e) — filed one day over a year before '273). Bridges ordinary telephone service and a packet network. Bears on the telephony↔data-network bridging of claim 10 (TIM/DPM/IIM architecture). → § 103 reference; not a voice-browser/markup reference.

5. ⚠ Anomalies and contradictions to flag

1. US 6,167,376 A (Ditzik), filed 1998-12-21 appears in the '273 "References Cited" table, but its filing date post-dates the '273's 1998-10-07 filing by ~2.5 months. As a US patent it therefore cannot be § 102(a) or § 102(e) prior art against the '273 absent an earlier priority/benefit claim not reflected in the data provided. Its presence likely reflects citation as "art of interest" or a citation added late in the ~7-year prosecution, rather than a formal prior-art rejection. I am not auto-correcting this; I am reporting the date as recorded and flagging the inconsistency. (No earlier provisional/benefit claim for Ditzik '376 is shown in the record I retrieved.)
2. US 6,240,448 B1 and US 6,269,336 B1 have publication/grant dates after 1998-10-07 but qualify as prior art only via § 102(e) (US filings before that date — for '448 the 1995 provisional / 1996 PCT; for '336 the 1998-10-02 US filing and 1998-07-24 provisionals). If a court fixed an invention date earlier than these § 102(e) dates, their status would change; I assume the filing date governs.
3. Two rosters, one reference apart: Google lists a "Citations (23)" roster and a "Patent Citations (24)" roster; the difference is US 5,867,816 A, which appears only in the 24-item table. I have analyzed all 24.
4. The claim-differentiating limitation. Nothing in the 24 cited references discloses, in a single reference, the "multi-part session allowing a user to return to where a previous session was terminated if the previous session was interrupted … [where] the interruption … comprises a complete termination between said data applications and at least one of said telephone and said computer terminal." The two references with any notion of interruption (US 6,269,336, barge-in) and any notion of session persistence (US 5,745,681, cart state) address different things. This limitation — and the specific "audio tags in combination with non-audio tags" markup claim — appear to be what distinguished the '273 over the cited art.

6. Bottom line by claim

• Independent claims 1, 9, 10, 11, 12, 13 — no single cited reference anticipates. The art cluster {US 5,953,392; US 6,240,448; US 5,884,262; US 6,269,336; US 5,915,001; US 6,101,473; US 6,335,928; GB 2317070 A; US 5,761,280} collectively discloses telephone↔internet access, markup retrieval, parsing, TTS, and menu/DTMF control — i.e., strong § 103 material — but not the claimed dual audio/non-audio tag scheme or the multi-part session/interruption-return element.
• Claim 6 (cart/purchase): supported by US 5,715,314 (network sales) + US 5,745,681 (web cart); § 103 only.
• Claim 7 (phonemes/audio vectors): supported by the speech cluster (US 5,799,278; 5,806,030; 5,806,033; 5,809,462; 5,867,816; 5,991,712; 5,913,196) for the phoneme-conversion half; § 103 only — the audio-vector database search step is not anticipated.
• Claims 2, 3, 4, 5, 8: strongest partial disclosures in US 6,240,448, US 5,953,392, US 6,269,336, US 5,884,262, US 5,802,526 (menus, HTTP fetch, TTS+recorded audio, tag parsing, audio commands); § 103 evidence. Because they depend from the independent claims, they cannot be anticipated alone.
• Claim 14: supported by the multi-module architectures of US 5,953,392 / US 6,240,448 / US 6,269,336; § 103.

7. Caveats

1. My assessment uses the claim text and reference data on the '273's face and the Google Patents full text. I did not pull the complete prosecution history (file wrapper) for application 09/167,485, which would show exactly which references the examiner applied and under which subsection, and whether any claims were allowed over specific art. That record is the definitive source for "what anticipated/rejected what."
2. Reference characterizations are drawn from abstracts/front-page data and the patent text's own framing; a limitation-by-limitation reading of each reference's full specification (element-by-element claim charts) would be required to convert any § 103 observation into a formal rejection or invalidity contention.
3. Dates are stated as recorded in the patent's citation tables and corroborated where possible via Google Patents / uspto.report; I did not independently verify every priority claim (e.g., provisional benefit) for each reference.
4. For a definitive anticipation/invalidity conclusion, cross-check via USPTO Patent Center (App. 09/167,485 "Documents" tab), Unified Patents / Docket Navigator PTAB & litigation, and a full-text prior-art search of the 1996–1998 window (including non-patent literature such as Hemphill et al., "Surfing the Web by Voice," ACM 1995, and the Netphonic PCT WO 97/32427 / EP 1 224 793 family).

I'll verify a few key reference disclosures before building the § 103 combinations, so the motivation-to-combine reasoning is grounded rather than assumed.

Obviousness Analysis Under 35 U.S.C. § 103 — US Patent 6,941,273

"Telephony-data application interface apparatus and method for multi-modal access to data applications"
App. 09/167,485 · Filed 1998-10-07 · Granted 2005-09-06 · Inventors Loghmani & Korangy · Status Expired – Lifetime

0. Scope, assumptions, and continuity

Builds on, does not repeat: This analysis uses the reference set and reference characterizations developed in the earlier Prior art section of this page (all 24 citations on the face of the '273, with § 102(b)/(a)/(e) status assigned), and the claim construction shorthand in that section's § 2. It also carries forward two prior findings rather than re-deriving them:

• Litigation / PTAB sections are consistent with this analysis — no asserted claim construction exists, and no PTAB FWD has construed any term. Every construction below is therefore proposed under Phillips, not adopted.
• The claim-text anomaly flagged in the PTAB section persists and materially affects this analysis. The "multi-part session … return to where a previous session was terminated if the previous session was interrupted … complete termination" language appears in every independent claim (1, 9, 10, 11, 12, 13). If the operative claim set differs from the Google Patents text (e.g., a certificate of correction or reexamination certificate not visible in the supplied record), Combination 4 below must be re-run. I am not asserting such a certificate exists.

Two corroborating search findings added since the Prior art section (used below with appropriate weight):

Statutory framework. Pre-AIA (filed 1998-10-07). No claim of this patent was ever tested at the PTAB (per the PTAB section), so there is no § 315(e) estoppel and no prior Board construction constraining the analysis. The § 103(a) question is simply: would the subject matter as a whole have been obvious to a person having ordinary skill in the art at the time of the invention? Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), the analysis may combine multiple references where a POSITA would have had a reason to combine them with a reasonable expectation of success, and the combination need not have been the subject of explicit teaching, suggestion, or motivation in the references.

POSITA definition proposed. As of October 1998: a bachelor's degree in electrical engineering, computer science, or equivalent, plus two to three years of experience in telephony/IVR systems or Internet application development, or a master's degree with roughly one year of such experience. The POSITA is familiar with (i) telephony signaling and DTMF, (ii) HTML and CGI-style server scripting, (iii) TTS/ASR engines and voice-response platforms, and (iv) electronic commerce/shopping-cart architecture. The '273's own specification confirms this level of skill — it treats TTS/ASR as "commercially available software" and the browser/cart architecture as conventional.

A caution on § 102 vs. § 103. The Prior art section concluded — correctly — that no single cited reference anticipates the independent claims, principally because of the dual audio/non-audio tag scheme and the multi-part-session/interruption-return element. That conclusion does not weaken the § 103 case; it defines it. Each combination below is framed to close precisely one of those two gaps.

1. Key terms driving the analysis

2. The general motivation landscape (KSR "design incentives / market forces")

Before reaching specific pairings, three system-level motivations were so well documented by 1998 that they supply the "reason to combine" for nearly every ground below:

1. The explicit access-gap problem. Imielinski's own specification (priority 1995-12-22) frames it: "the majority of the world's population does not have access to a computer," and a user away from a PC "is not in the position to access the Internet." Imielinski's stated object is precisely to deliver Web resources over an ordinary telephone. A POSITA reading this has an express, reference-sourced reason to marry telephony to Web content retrieval.
2. The 1995–1998 "voice Web" literature and product wave. Hemphill & Thrift (ACM Multimedia 1995) demonstrated voice surfing, speakable links, and on-the-fly pronunciation generation — i.e., the spoken-command-over-HTML concept — before the '273's filing date. Netphonic's WebOnCall (US 5,953,392, filed 1996-03-01) was a commercial implementation. The field was crowded and converging; predictable-innovation logic under KSR applies strongly.
3. The 1998 e-commerce imperative. Open Market's network-sales system (US 5,715,314, granted 1998-02-03) and Sun's web shopping cart (US 5,745,681, granted 1998-04-28) established that merchants wanted carts extended to every available channel. The '273's own background concedes the art's awareness of the cart-persistence deficiency ("shopping cart information … removed … after a prolonged period of inactivity"), which is itself a reference-grade admission of the problem the multi-part-session limitation purports to solve.

Cross-corroboration (weight carefully). Imielinski (US 6,240,448) is used repeatedly as IPR grounds art — e.g., as Ex-1005 in the IPR petition against US 8,391,298 (Ground 2: Chang + Byrne + Imielinski), and 8×8's US 5,761,280 appears as Ex-1052 in IPR2022-00766 (Samsung v. Smart Mobile Technologies). These are petition allegations, not Board findings, so they prove only that skilled drafters treat these references as disclosing audio-enabled Web pages and telephone Web browsing — which is exactly the "recognized body of art" point I am using them for.

3. The principal § 103 combinations

3.1 Combination 1 — The core ground: Rhie (US 5,953,392) + Ladd (US 6,269,336) + Imielinski (US 6,240,448)

What each supplies

Why combine them (motivation). All three are in the same field (telephony access to networked information), were cited by the examiner on the face of the '273, and address the same problem. Ladd's contribution is the one the '273 needs most: a markup document whose elements are tagged, some of which are spoken and some of which are functional/non-audio (the HIDDEN input "used to create variables and assign values"). Imielinski supplies the authored audio-tag vocabulary and the recursive link traversal; Rhie supplies the operational telephony/DTMF + recorded-voice/TTS pipeline. Combining three references that each solve a sub-part of one problem, all in the same art, is the paradigm KSR case: "a combination of familiar elements according to known methods [that] does no more than yield predictable results."

Reasonable expectation of success. High. There is no new hardware; each reference is a working system, and the integration (parse page → extract tagged item → speak it) is a server-side scripting exercise within the POSITA's skill.

3.2 Combination 2 — The shopping grounds (claims 6, 9, 11): Combination 1 + Open Market (US 5,715,314) + Sun (US 5,745,681) [+ IBM GB 2 317 070 A]

Mapping. Open Market '314 supplies the network sales flow and payment-order transaction; Sun '681 supplies the Web shopping cart with its explicit state-maintenance problem; GB 2317070 (IBM, "Voice processing/internet system," published 1998-03-11) supplies the voice-processing-plus-Internet bridge at the system-architecture level for claim 11. Claim 6's "voice-enabled electronic shopping cart" that tracks, stores, and later purchases is the mere substitution of an audio input channel into '314+'681's cart — a predictable use of an old element for its known function.

Motivation. (i) Explicit merchant demand to reach non-PC buyers (Imielinski's access-gap rationale applied to commerce); (ii) the '273's own background identifies Web carts and their retention deficiency as the prior art — an admission of the starting point; (iii) claim 11's "HTTP-type electronic command" is satisfied verbatim by Ladd's FORM … METHOD (get|post) posting and Rhie's HTTP request. There is no teaching away; if anything, the references reward unification.

Claim 11's independent-management limitation ("TIM, DPM, and Internet interface module … independently of each other") is a server load-balancing/architecture limitation supported by Ladd's separate VRU client/server structure and by Rhie's caller-server model. § 103 treatment: arranging known functional blocks so that connection handling scales independently is a predictable engineering choice with a known benefit (also the subject of Ladd's VRU/markup-server split).

3.3 Combination 3 — Claim 7 (phonemes / audio vectors): Imielinski '448 + Hemphill 1995 + Ericsson (US 5,809,462; US 5,867,816) + Matsushita (US 5,806,030)

Claim 7 requires (a) converting a spoken query to phonemes, and (b) searching said database using audio vectors based on said phonemes.

• (a) is squarely met by the cited speech-recognition cluster (Ericsson neural-network phoneme recognition; Matsushita feature/clustering; IBM HMM '278; Telia '033) — as the Prior art section already found, these support the phoneme-conversion half of claim 7.
• (b) is the harder half, but it is not novel in kind: Imielinski discloses a "speech index … generated of spoken words to provide search capabilities in a similar manner as WWW indexes" organized as a tree of audio enabled pages, with per-page "context" sets that limit the recognition vocabulary; Hemphill 1995 discloses dynamic grammars, a "speakable hotlist" associating utterances to URLs, and on-the-fly pronunciation generation.
• Motivation: the POSITA has a strong, reference-expressed reason to index DB content by sound rather than spelling — Imielinski states that tables/numeric data "are difficult to convert to speech" and that textual/lexical matching of phone-recited names is error-prone; Hemphill's stated motivation is that "names of hotlist entries … do not always associate well with the corresponding page." Both expressly motivate phonetic association. Combining those motivations with a nearest-neighbor/feature-distance algorithm — the standard tool in the cited clustering and HMM art — yields "search the database using audio vectors based on phonemes."

Honest weakness: I have no prior-art reference in the cited set that stores a vector per database record and searches by vector distance. Imielinski indexes rather than distance-searches; the speech cluster distances feature vectors rather than database items. Claim 7 is therefore best attacked as § 112(a)/(b) + § 103 over Imielinski/Hemphill plus the clustering art, and a defendant should not claim clean anticipation. Separately — and this is a § 103 not a § 102 point — the same subject matter is described in the same-day sibling US 6,377,927 (issued 2002-04-23). Because both applications bear the same 1998-10-07 filing date, '927 likely falls outside § 102(e) against the '273 absent proof its filing was earlier in time; I flag it as a file-wrapper-dependent lead, not a conclusion.

3.4 Combination 4 — The distinguishing limitation: Uppaluru/Vois (US 5,915,001 and family, incl. US 6,400,806 B1) + Sun '681 + Ladd '336

This is the combination on which any real invalidity case turns, because the multi-part session / complete-termination / resume element is the only claim language the Prior art section found nowhere in a single reference.

Motivation to combine (the key paragraph). A POSITA confronted with the '273's acknowledged problem — a caller who hangs up mid-transaction loses the session — would look to the two persistence mechanisms already commonplace in the Web art: (i) server-side stored profiles addressed by URL (Vois) and (ii) hidden server-side or page-carried state (Ladd's HIDDEN, and cookies/state objects per Sun '681). Both are "known techniques to improve similar devices in the same way"; using them to let a telephone shopper resume is the predictable application of a known solution to a recognized problem — the factual pattern KSR labels obvious. The result is "a multi-part session … allowing a user to return to where a previous session was terminated."

Where a patentee would push back. The claim's clause — "the interruption … comprises a complete termination between said data applications and … said telephone …" — arguably requires that the data application itself (not merely the caller's connection) accommodate the gap, and the patentee would argue the cited art persists identity/cart, not a session with the data application. That is a genuine § 112 indefiniteness argument in the patentee's favor, but it also dooms any narrow construction: read narrowly, the limitation is a routine timeout/session-recovery feature; read broadly, it is met by Vois profiles. Either way it is vulnerable (broad → obvious; narrow → indefinite). A defendant should press both horns.

3.5 Combination 5 — The "concurrent markup + audio" and dual-user limitation (claims 1, 9, 13): Rhie '392 (or Ladd '336) + 8×8 (US 5,761,280)

Claim 1 requires generating the markup page concurrently with the audio message for selective use, and claim 9 claims users on a graphical interface device and a telephony device, simultaneously or sequentially.

Motivation. 8×8 '761,280 discloses a telephone plus a display circuit/display in the same interface, with a Web browser in memory — i.e., the same user simultaneously using a voice channel and a visual browser. Rhie supplies the audio rendering of the same Web content. Because Rhie and 8×8 both retrieve and render the same HTML pages, serving one page to a browser and an audio rendering of that page to a phone is not a new architecture but the presentation of the same data twice. Claim 9's "simultaneous and sequential" language tracks the ordinary multimodal use case (talk on the phone while looking at the screen).

Note on claim 10's contrast. Claim 10 (interface with TIM + DPM + online interface module) is the most structurally clean of the independents — it is essentially Ladd's VRU-client/voice-browser/markup-server split with Rhie's telephony layer, and is correspondingly the easiest to map element-for-element. A defendant should expect claim 10 to be the weakest claim on the merits.

4. Claim-by-claim § 103 summary

Observation on claim 12. Claim 12 omits claim 1's "converting said spoken query to text" step yet retains "generating an electronic request using at least one electronic command in response to said spoken query." That asymmetry is a § 112 written-description problem (the spec's whole architecture converts speech to text at the TIM), and it also broadens claim 12 so that Imielinski's direct-speech ATML path reads on it more cleanly. Claim 12 is likely the most exposed independent claim.

5. The articulated motivations to combine (consolidated)

A defendant needs these stated affirmatively; each is drawn from the references themselves or the '273's own admissions:

1. Same field, same problem — every primary reference is telephony-to-networked-data; examiner-cited overlap (§ 103 rationalization (A)).
2. Explicit access-gap teaching — Imielinski: most people lack computers; solve it via the telephone.
3. Predictable substitution of a known input channel — DTMF/speech for keyboard in a known Web shopping flow (KSR "familiar elements / known method / predictable result").
4. Design incentive from the cart-persistence problem — admitted in the '273's background; solved by known state-persistence tools (Ladd HIDDEN; Vois URL-addressed profiles; Sun cart state).
5. Interoperability/standards pressure — Ladd and Imielinski both show the field standardizing on tagged markup documents served over HTTP as the common substrate; a POSITA would reuse that substrate rather than invent one.
6. Performance/scaling benefit — splitting call handling, command translation, and network I/O into independently scalable modules (claim 11/14) is a known architecture with a known benefit (Ladd's VRU/markup server split).
7. Contemporaneous NPL — Hemphill 1995 removes any argument that voice-over-Web was non-obvious as of 1998.

6. Where the § 103 case is weak — counterarguments to anticipate

1. The multi-part-session limitation (§ 3.4) is the patentee's best defense. There is no single reference disclosing complete termination of the connection to the data application followed by resumption at the prior point. The combination must be argued, and the patentee will attack the "reasonable expectation of success" prong by arguing that persisting a shopping session (not merely a profile) across a full teardown in 1998 was non-routine. Rebuttal: Sun '681 and the general state-object/cookie art make session persistence routine; and the '273 spec's own account shows the solution is server-side record retention, not new technology.
2. § 112 ambiguity as a shield. If the patentee narrows "multi-part session" to avoid Vois/Sun, indefiniteness rises; if it broadens, obviousness rises. The defendant should force the election early (in an IPR, via the Board's Phillips construction; in litigation, via a § 112 motion).
3. Claim 7's audio-vector database search is not anticipated by anything in the cited set (§ 3.3). A § 102 theory should not be pleaded on claim 7.
4. Reference-dating. Two § 102(e) dependencies (Imielinski via its 1995 provisional/1996 PCT; Ladd via its 1998-07-24 provisionals and 1998-10-02 filing) assume the '273's effective date is its 1998-10-07 filing date. If the patentee proves an earlier conception/reduction to practice, Imielinski's status as § 102(e) art may be affected (Ladd's July 1998 provisional and Imielinski's 1995 priority are probably still early enough, but Ladd's § 102(e) date is only five days before filing — a thin margin worth verifying against the provisionals' support for the asserted disclosure).
5. The US 6,167,376 (Ditzik) anomaly flagged in the Prior art section stands: a post-filing reference cannot be § 102(a)/(e) art against the '273. Do not rely on it.
6. Claim text uncertainty (§ 0). If a reexamination certificate changed the claims, re-scope.

7. Graham factors and secondary considerations

There is no evidence in the record of objective indicia, and several factors run against the patentee:

• No commercial success nexus shown — the assignee Logic Tree was an operating IVR vendor (per the NPE-signals section), but there is no evidence tying any product's success to these claims as opposed to the family's other assets. Commercial success without a nexus is entitled to no weight.
• No industry praise, no licensing of the '273 specifically, no copying evidence, no unexpected results in the supplied record.
• Long-felt need cuts for the defendant — the need (telephone access to Web commerce; cart persistence) is admitted in the '273's own background and expressly articulated in Imielinski (1995) and Hemphill (1995) years before filing.
• The crowded, fast-moving 1995–1998 art (Rhie, Imielinski, Ladd, 8×8, Netphonic's WebOnCall, Hemphill) is itself evidence that the claimed combination was within ordinary skill, not beyond it.
• Expired status shifts the practical question from validity to pre-2018-10-07 damages, which reduces both parties' incentive to litigate.

8. Alternative / co-equal invalidity theories (not § 103, flagged for completeness)

Because the § 103 case is strongest on the structural claims (2–6, 8, 10, 14) and weakest on claim 7 and the session clause, a full challenge should run these in parallel:

• § 112(a)/(b): the functional, result-oriented language of the independent claims ("establishing a multi-part session," "concurrently generating," "directly access information from the first portion") and claim 12's missing text-conversion step.
• § 112(b): typographical/antecedent defects in the published claims (claim 11 "mufti-part"/"user commuter"; claim 12's truncated second-portion clause; "telephony — interface module" in claim 13) — assuming, per § 0, that these appear in the operative claim set.
• § 101: if the claims survive § 103, the "concurrently generating … markup page … for selective use" and session-management steps are arguably mental-process/data-gathering-adjacent; weak, and post-Alice practice in these telephony-server cases has generally favored eligibility, so treat as secondary.
• Prior-art search extension beyond the face citations: Hemphill & Thrift 1995 (NPL, § 102(b)); the Vois family US 6,400,806 B1 (voice Web pages with "specially tagged set of key words and touch tone sequences" and URL-addressable personal profiles); and IPR-tested art like the Ladd continuations (e.g., US 6,473,617, US 6,493,671, US 6,535,359) that elaborate the tagged-audio markup model.

(I have not verified the exact filing/priority dates of US 6,400,806 B1 or the Ladd continuations beyond what the search results showed; confirm before relying on any of them as § 102/§ 103 art.)

9. Bottom line

• The independent claims are obvious under § 103, most persuasively through the Rhie '392 + Ladd '336 + Imielinski '448 core combination, extended for the shopping claims with Open Market '314 + Sun '681, and for the session clause with Vois '001 / Ladd HIDDEN inputs / Sun '681. The references are examiner-cited, co-field, and mutually reinforcing; the motivations are express and the expectation of success is high.
• The strongest grounds are claims 2, 3, 4, 5, 6, 8, 10, and 14 — several of which are map-the-elements exercises rather than arguments.
• The two hard limitations are (i) the multi-part session / complete-termination / resume clause, and (ii) claim 7's audio-vector database search. These require combination arguments and should be briefed with care; neither is foreseeably anticipatable by a single reference.
• No tactical constraint comes from the PTAB or the courts — per the PTAB and Litigation sections of this page, no AIA trial exists and no litigation was found, so there is no estoppel, no prior construction, and no adverse validity finding to work around.
• This is a paper invalidity case, not an enforcement-exposure case: the patent expired 2018-10-07 and is off the assertion table.

Verify before relying: (a) the operative claim set in USPTO Patent Center for application 09/167,485 (Transactions / Reexamination Certificate tab) given the anomalous claim text; (b) the file wrapper to confirm which references the examiner applied and under which subsection; (c) the provisional-support dates for Ladd's § 102(e) margin (1998-07-24 provisionals); (d) any earlier conception evidence for the '273; and (e) the filing dates of US 6,400,806 and the Ladd continuations if added to a ground.

Sources: US6941273B1 full text — https://patents.google.com/patent/[US6941273](/patent/US6941273)/en · US 5,953,392 — https://patents.google.com/patent/US5953392A/en (RPX Insight: https://insight.rpxcorp.com/patent/US5953392A) · US 6,240,448 — https://patents.google.com/patent/US6240448 (spec PDF: https://patentimages.storage.googleapis.com/1c/24/65/2cfe516cc27734/US6240448.pdf) · US 6,269,336 — https://uspto.report/patent/grant/6269336 (spec PDF: https://patentimages.storage.googleapis.com/f4/e2/54/5d5858c4ddfb4e/US6269336.pdf) · US 5,761,280 (8×8) — Docket Alarm Ex-1052 PDF in IPR2022-00766 · Hemphill & Thrift, "Surfing the Web by Voice," ACM Multimedia 1995, pp. 215–222 — https://dl.acm.org/doi/fullHtml/10.1145/[217279](/patent/217279).[215269](/patent/215269) · Vois-family voice web disclosure — US 6,400,806 PDF · IPR Ground 2 using Imielinski as Ex-1005 (US 8,391,298) — PTAB petition documents at ptacts.uspto.gov.