US 5970473

I'll verify this patent against authoritative sources, including checking for any CAFC litigation activity.

Let me do one more targeted check on the patent's legal status and any litigation history.

US Patent 5,970,473 — Verification Summary

Search coverage note: I checked the USPTO/patent databases (Google Patents authoritative text, uspto.report, FPO) and searched the CAFC April 2026 argument calendar and 2026 opinions/orders for the exact number. No CAFC 2026 docket activity for 5,970,473 was found. That is consistent with the patent's status: it is recorded as Expired – Lifetime, with anticipated expiration on 2017-12-31 (20 years from filing), so it would not be the subject of live 2026 appellate litigation. I found no PTAB (IPR/PGR) proceeding or reexamination tied to this specific number. I cannot rule out district-court history that predates my search coverage, but I found none.

Caveat on your search framing: My searches returned results for other patents (e.g., an IPR involving U.S. 8,144,183 and various AT&T/TiVo/Broadband iTV matters). I have excluded those — none concern 5,970,473. Similarly, the "'473" references in some results refer to different patents (e.g., a TiVo/Seagate acoustic-noise patent), not this one.

Bibliographic Data (per the authoritative patent text)

Abstract (as issued)

"Catalog information stored in a database is accessible by remotely located customers in order to provide in-home shopping capabilities."

Plain-Language Overview of the Independent Claims

There are three independent claims — claim 1 (system), claim 11 (apparatus), and claim 12 (method). The rest depend on them.

Claim 1 — System. A communications system that gives users catalog information about goods/services and lets them buy. It requires:

• A database server that (a) stores users' shipping data and (b) is configured to determine the lowest cost of an item a user selects for purchase.
• Multiple user terminals, each with a unique terminal identifier. The server stores a table mapping each terminal's unique identifier to that user's shipping address — this is the core novelty (auto-lookup of shipping address by device ID).
• Each terminal has a touch-sensitive screen that displays catalog information and accepts touch input to select an item for purchase, plus a control unit that turns purchase selections into purchase data sent to the server.

Claim 11 — Apparatus. A catalog-services apparatus comprising: a display means; an input means for receiving order data; a control unit that generates purchase data from that order data; and a centralized database server in communication with the control unit that processes the order data based on the purchase data. (This is the broader, means-plus-function-style counterpart to claim 1.)

Claim 12 — Method. A method for providing in-home catalog services: configuring a user terminal with a unique terminal identifier; connecting it to a database server; storing a table relating the terminal's unique identifier to the user's shipping address; displaying catalog information on the terminal's display; receiving purchase orders via an input device; and determining the user's shipping address from that stored identifier-to-address table. (Note: claim 12 does not itself recite the "lowest cost" limitation — that appears in dependent claim 16.)

Representative dependent claims: user-profile-based catalog display (2, 3, 17); password gating of purchase or display (4, 5, 14); purchase-history records (6); pending-order status requests forwarded to the vendor (7, 8); returning the stored shipping address for user confirmation (9, 18); and shipping preferences — carrier, insurance, delivery times, extended warranties (10).

Plain-Language Takeaway

The patent describes an AT&T videophone/DSL network architecture (ISD → FMP → NSP) where a home videophone's unique identifier/IP address is used to automatically look up the customer's saved shipping address, so a purchase can be completed with just a touch on the screen — no re-typing of shipping details and, because the device is in the home, reduced credit-card fraud. The distinguishing claim element versus ordinary e-commerce is the device-identifier-to-shipping-address table plus, in claim 1, automatic lowest-price determination for a selected item.

Uncertainty

• Assignee: Google Patents lists original assignee AT&T Corp. and current assignee AT&T Inc. I have not independently confirmed the current chain of title (reassignments after the 1998 assignment record), so treat "current assignee" as a listing assumption.
• Litigation/CAFC: I found no 2026 CAFC docket entry for this number; given the 2017 expiration, any enforcement activity, if it existed, would predate that. I cannot treat this as an exhaustive history of all district-court or pre-2018 proceedings.
• Claim construction: No court or PTAB interpretation of these claims was located; the plain-language claim readings above are my own, drawn from the specification and claim text as published.

I'll research litigation involving US patent 5970473.

Initial searches didn't surface litigation. Let me dig into specific patent litigation databases.

Let me search more targeted sources for litigation on this specific patent.

Summary of Findings

Bottom line: I found no evidence of any litigation — no district court case, no ITC Section 337 investigation, and no PTAB proceeding — involving U.S. Patent No. 5,970,473 ("Video communication device providing in-home catalog services," Gerszberg et al., AT&T Corp., filed Dec. 31, 1997, issued Oct. 19, 1999).

I was unable to complete an exhaustive multi-database verification (I hit my tool-call limit), so treat the above as "nothing located" rather than a certified negative. Details and caveats below.

What I searched

• General web queries for US 5970473 litigation, 5970473 patent infringement lawsuit, "5970473" patent case, "5,970,473" infringement (both comma-formatted and unformatted, to catch either convention).
• Unified Patents portal pages (portal.unifiedpatents.com/patents/patent/US-5970473-A appeared only as a citation embedded inside other patents' pages, e.g., US-7,623,436 and US-7,359,871 — not as a litigation entry).
• The Google Patents page for US5970473A provided in the record, which includes a "Cited By (168)" list and a "Family" section but no litigation/PTAB tab content.

Results: no cases to report

Important distinctions (so you don't mis-cite a similar number)

1. Forward citations ≠ litigation. The 168 "Cited By" entries on Google Patents are later patents/prosecution citations that reference the '473 patent as prior art. A concrete example surfaced in my search: PCT application PCT/US 03/30766 (JP 2006-501725) cites "US 5 970 473 A (Gerszberg)" as an "A" category reference against claim 14-17/34-35/46-47. That is patent-office citation, not a lawsuit.
2. Same-family continuations ≠ litigation. The "Cited By" list includes AT&T-family continuations by the same inventors, e.g., US 2005/0254484 A1 (Barzegar, "Multifunction interface facility…") and US 2006/0159116 A1 (Gerszberg, "Facility management platform for a hybrid coaxial/twisted pair local loop network service architecture"). These are related applications, not adverse proceedings.
3. Beware other "'473" patents. During my search I hit several unrelated "'473" matters that are easy to conflate: US 6,587,473 (IWS, Wi-Fi/Ethernet access point case, E.D. Tex., 2013 letters referenced); the Seagate-related "'473" hard-drive patent reexamination (acoustic noise/"seek" limitation, cited in a courtlistener PDF); and other unrelated '473 patents (e.g., Wilus/Samsung, Motorola/Apple). None of these is US 5,970,473.

Context bearing on enforceability/enforcement history

• Term/status: Google Patents lists "Anticipated expiration 2017-12-31" and legal status "Expired – Lifetime." A patent filed Dec. 31, 1997 without term extension would have expired around Dec. 31, 2017, so any infringement window is long closed. (The 2017 date shown is the PTO's anticipated expiration; I did not confirm whether any PTA/terminal disclaimer changed it.)
• Owner: Originally AT&T Corp.; current listed assignee AT&T Inc. This patent is one of the large AT&T "HFTP / videophone" family (roughly 45 concurrently filed applications listed in the specification, including sibling applications like "Restaurant Ordering Via VideoPhone," "Ticket Ordering Via VideoPhone," and "A Communication Server Apparatus For Interactive Commercial Service").
• Subject matter: Claim 1 and claim 12 (the two independent claims) are directed to a catalog database server with a terminal-identifier-to-shipping-address table and a touch-sensitive ordering terminal. Its G06Q 30/06 classification and business-method character make it the type of patent that could have drawn Covered Business Method (CBM) review during the 2012–2018 window — but I located no CBM, IPR, PGR, or district court docket naming it.

Recommended definitive checks (before you rely on this)

Because absence of evidence here is not proof, I suggest confirming against sources I could not fully query:

1. Unified Patents litigation portal — https://portal.unifiedpatents.com/patents/patent/US-5970473-A (patent page) and https://portal.unifiedpatents.com/litigation/caselist (case list).
2. RPX Insight / Docket Alarm / Lex Machina patent case lookup by patent number.
3. PACER / CourtListener RECAP free-text search for "5,970,473."
4. USPTO PTAB (PTAB E2E / P-TACTS) and USITC EDIS (Section 337) by patent number.
5. USPTO Patent Assignment records for any recorded security interest, license, or transfer that might hint at enforcement activity.

If those return nothing (as my searches did), the correct statement for your filing is: no litigation involving US 5,970,473 is known as of April 26, 2026.

Note on confidence: I am reporting a search-grounded negative result. I did not have complete access to the paid litigation databases, and I did not independently verify the anticipated-expiration date or whether any unrecorded licensing/enforcement activity occurred. I did not find any search result that contradicts this negative finding.

I'll verify the structured "no proceedings" default against live sources before drafting.

Let me run targeted checks for the two proceeding types this patent's profile most plausibly attracted — CBM (it's a 1999 e-commerce patent) and any Unified Patents filing.

Proceedings overview

Total AIA trial proceedings on file for US 5,970,473: zero (0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denied) — the USPTO Open Data Portal (ODP) API returns no IPR, PGR, or CBM proceeding naming this patent, and there is no "PTAB proceedings on file" entry to enumerate. The bottom-line defensive posture is therefore not "the patent has survived two IPRs and is hardened," and it is also not "claims 1–5 have been canceled." It is: the patent's 18 issued claims are entirely untested at the Board, and the reason is almost certainly economic, not tactical — US 5,970,473 expired on 2017-12-31, so there is no live right to exclude left to attack. A defendant receiving a demand letter today should treat the absence of PTAB history as a signal to check the demand's standing/value rather than as evidence of claim strength.

Verification note: I searched Google Patents, uspto.report, PTAB/PTAB E2E-linked petition documents, PTAB decision collections, and general PTAB docket reporting for this number, by patent number and by title. My searches surfaced several near-misses that are different patents and must not be confused with this one — e.g., the Virentem/Enounce IPRs on U.S. 7,683,903 and others (D. Del. C.A. No. 18-917-MN), the ingress/audio-video IPRs on U.S. 8,144,183 (IPR2022-00814), and various Unified Patents filings on unrelated numbers. None concerns 5,970,473. I also note the Google Patents "Cited By" list for this patent contains 168 third-party citations but is not a PTAB docket. I cannot rule out a pre-2018, never-indexed, or fully-settled-before-FWD proceeding, but I found no evidence of one.

---

Proceedings

None to enumerate. No {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} entry can be constructed without fabricating a proceeding number, which the task constraints forbid. I am not creating placeholder sections.

---

Strategic summary

Claim status. All 18 issued claims of US 5,970,473 are UNTESTED at the PTAB. There are no canceled claims (no FWD has ever issued) and no sustained claims (nothing has ever been adjudicated). That means claims 1–18 remain presumptively valid as issued — including the three independents: claim 1 (system: database server storing shipping data + "lowest cost" determination + table mapping unique terminal identifiers to shipping addresses + touch-sensitive-screen terminals), claim 11 (apparatus, means-plus-function style), and claim 12 (method: unique terminal identifier → stored identifier-to-address table → determining shipping address). The dependent claims on password gating (4, 5, 14), purchase-history records (6), pending-order status requests to the vendor (7, 8), shipping-address confirmation (9, 18), and shipping preferences (10) are likewise unadjudicated. Anyone telling you a PTAB panel has already killed or blessed any of these claims is misreading the record.

The real gating issue is the expiry, not estoppel. Because US 5,970,473 is recorded as Expired – Lifetime with an anticipated expiration of 2017-12-31 (20 years from the 1997-12-31 filing), the § 315(e)(2) estoppel landscape is academic: estoppel analysis presupposes a petitioner. The practical constraints are statutory instead —

• Under 35 U.S.C. § 286, a plaintiff can recover damages only for infringement within six years before suit. With the patent expired on 2017-12-31, any damages claim reaches only pre-expiration conduct, and a suit would have had to be filed by roughly 2023-12-31 to capture the full lookback. A demand letter dated in 2026 asserting this patent is, on its face, asserting a patent whose exclusionary term ended nearly nine years ago.
• Expired patents are not categorically immune from IPR, but they are poor IPR targets: the patent owner cannot amend claims (no § 316(d) relief available), and claim construction follows Phillips rather than the BRI standard. Petitioners generally do not spend six-figure budgets attacking a patent with no forward-looking exclusion right — which is the most parsimonious explanation for the empty docket.

CBM is the conspicuous non-event. This is precisely the profile that the Transitional Program for Covered Business Method Patents (in force 2012-09-16 to 2020-09-15) was built for: a 1997-filed AT&T patent whose claims are directed to data processing in the practice/administration of commercial transactions (ordering and paying for goods and services), with § 101 abstract-idea vulnerability. The patent was still unexpired for most of the CBM window. No CBM petition was filed. That is a genuine pattern signal: NPEs and their targets simply never litigated this patent in the PTAB era. Combined with the previous section's finding of no CAFC 2026 activity, the picture is a patent that sat out the entire AIA trial regime.

No petitioner repetition, no PO appeals, no aggregator. There is no serial-petitioner pattern (nothing to repeat — zero filings), no patent-owner appeal to the Federal Circuit arising out of a PTAB proceeding, and no Unified Patents (or comparable defensive aggregator) challenge in the chain based on my search of Unified's public patent/exhibit portals. The prior-generated sections' conclusion of no CAFC docket activity for this number is consistent with, and corroborated by, the zero-proceeding finding here; I flag no contradiction between them.

---

Recommended next steps

1. If you have received a demand letter asserting US 5,970,473: the first response item is not a prior-art search — it is a date check. Confirm with the patent owner (or via the USPTO PatentCenter/Global Dossier record linked at https://patents.google.com/patent/US5970473/en) that the patent expired 2017-12-31. Any infringement theory requires pre-expiration conduct within the § 286 six-year window; a 2026 demand cannot support forward-looking injunctive relief and will be damages-limited at best. Silence on expiration in the letter is itself informative.

2. If you face actual litigation (which would be unusual given the expiry), the invalidity record is a blank slate. There is no IPR estoppel under § 315(e)(2), because there is no petitioner — so every § 102/§ 103 ground, every printed publication, and every system-art theory is available in district court without estoppel exposure, and § 101 subject-matter challenges to the "device-identifier → shipping-address table" and "lowest cost" concepts are entirely un-preempted by any prior Board ruling. Expect the patent owner to resist any IPR as a waste given the expiry; expect a court to weigh § 286 and laches-type equities heavily.

3. Do not over-read the empty docket. Zero PTAB proceedings here does not mean "the claims are strong." It almost certainly means "the claims are stale." The correct inference is economic: well-asserted patents attract IPRs; expired, unasserted ones do not. If a counterparty is nonetheless citing claim 1's "lowest cost" element or claim 12's identifier-to-address table as a blocking right in 2026, ask them to identify the specific pre-2018 accused product and the calculation of pre-expiration damages. The absence of any Board history plus the 2017-12-31 expiration makes that a very short conversation.

4. If you have reason to believe a proceeding exists that I did not find — e.g., a 2013–2017 CBM that settled before institution and therefore never generated a public FWD — treat my "zero" as a finding bounded by my search coverage, not as a certification. The canonical source to re-verify is the USPTO PTAB E2E / ODP system (https://ptacts.uspto.gov) searched by patent number, plus CourtListener (https://www.courtlistener.com) for any CAFC appeal that would necessarily follow an FWD. Flag any hit you get back to me and I will re-run the analysis against the actual papers rather than against the empty record.

Confidence statement: High confidence that no AIA trial proceeding has reached a Final Written Decision on US 5,970,473. High confidence that the patent is expired as of 2017-12-31. Moderate-to-high confidence that no petition of any kind (IPR/PGR/CBM) was ever filed against it — supported by the ODP structured block plus independent searching, but not mathematically provable from public sources for the 2012–2018 window.

I'll reconstruct the assignment record for US 5,970,473. Let me pull the authoritative sources.

Ownership Chain Reconstruction — US Patent 5,970,473

Data-access caveat up front (read this before the tables): I was able to retrieve the assignment events for this patent (from the authoritative Google Patents legal-events record, which mirrors USPTO Assignment Center data), but I was not able to retrieve the reel/frame numbers, the correspondent attorney of record, or the execution dates on the underlying assignment documents. The Assignment Center and assignment.uspto.gov search pages are JavaScript-driven and did not surface machine-readable text in my searches. I am therefore not reporting reel/frame values — fabricating them would be worse than omitting them. Every reel/frame slot below is marked not retrieved, and the two facts I do assert (recorded date, parties) come from the patent text itself, which is authoritative.

---

Inventors

Unusual-pattern check — negative. All three inventors are AT&T personnel, and they assigned to AT&T Corp. on record. There is no evidence of any inventor departing the assignee within 12 months of filing, and no evidence of inventor-held (unassigned) residual rights. The application was filed under assignment (37 CFR 1.46 by the assignee of the entire interest), consistent with a normal corporate R&D filing, not a portfolio being staged for sale. Note that this patent is one of 45 concurrently-filed applications (all filed 1997-12-31) listed in the specification's incorporation-by-reference block — a coordinated, single-day corporate filing campaign, which is characteristic of an operating company's platform build-out, not an NPE accumulation.

---

Original assignee

AT&T Corp. (New York corporation), as named on the issued patent and on the recorded assignment dated 1998-04-29.

• Primary line of business: Interexchange/long-distance telecommunications carrier and, by 1997, a local-access and data-services entrant. The patent's own background section (column 1) frames the invention as an AT&T strategy to "penetrate the local telephone company's territory" and to "differentiate their services" — i.e., the assignee is a classic operating carrier.
• Did they ship a product embodying the claims? Partially / unclear. AT&T Corp. was unquestionably an operating company shipping telecom services. However, the specific commercial embodiment described here — the HFTP local loop with a home videophone (130), ISD (22), FMP (32), and NSP (36) driving an on-screen catalog with device-ID-based shipping-address lookup — reads as a 1997 architecture vision rather than a mass-market AT&T product. I have no evidence that AT&T commercialized the claimed videophone catalog service specifically. Treat "ships products embodying the claims" as not established for the claimed subject matter, while "operating company" is clearly established.
• Current status: Operating, as a corporate successor. AT&T Corp. was acquired by SBC Communications Inc. in 2005; SBC then adopted the AT&T Inc. name. AT&T Inc. remains an active operating carrier. Important: the "current assignee — AT&T Inc." entry on Google Patents is a corporate-successor inference by Google, not necessarily a recorded USPTO assignment for this patent. I found no recorded change-of-name or merger assignment for US 5,970,473 in the machine-readable record.

---

Assignment timeline

The recorded chain for this patent is a single link. Google Patents' legal-events record for US 5,970,473 shows exactly one reassignment entry, plus the grant/publication/expiration events. No post-issuance transfers appear.

• Executed not retrieved / recorded 1998-04-29 — Reel NNNNNN/NNNN (not retrieved)
  • Conveyance: Assignment of assignors' interest ("ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: GERSZBERG, IRWIN; MARTIN, JEFFREY S.; WALKER, HOPETON S. (as listed individually on the record)
  • Assignee: AT&T CORP.
  • Correspondent: not retrieved. (Flag: the firm/attorney that recorded the three inventors' assignment to AT&T Corp. is unknown to me. The only correspondent clue in the patent is the outside-counsel docket prefix "3493.xxxxx" on two concurrently-filed sibling applications — e.g., "VideoPhone Mail Machine (Attorney Docket No. 3493.73170)" — indicating a single outside firm ran the whole 45-application batch. That is a lead, not a finding: I cannot name the firm or tie it to the recording, and a single appearance would not be an NPE signal anyway.)
  • Context: Acquisition — initial corporate assignment of employee inventions to the employer at filing; routine, not a sale.

Subsequent recorded events (not assignments):

No other recorded assignments exist in the accessible record. There is no transfer to any LLC, no security agreement, no merger recording, and no termination/release. In other words, no post-issuance chain exists to analyze.

Cross-reference results (and what I could not verify):

• Google Patents legal events (authoritative text provided): consistent with the above; no change-of-name to AT&T Inc. and no IP-holding-entity transfer recorded for this patent.
• USPTO PEDS / ODP: not separately retrieved; no contrary indication found.
• SEC filings: No 10-K/8-K identified that specifically names this patent in a divestiture or sale.
• RPX / Unified Patents asserter directories: searches surfaced "AT&T Corp" only as an assignee of record on unrelated patents (e.g., Unified's page for US 7,623,436) and surfaced AT&T as a defendant in NPE suits (e.g., KAIFI LLC v. AT&T Inc., E.D. Tex. 2:19-cv-00138). Neither concerns US 5,970,473. I found no suit naming this patent and no party in this chain on any NPE list.

---

Timeline diagram

timeline
    title Ownership of US 5970473
    1997 : Filed 31 Dec by three AT&T inventors
    1998 : Assignment recorded to AT&T Corp
    1999 : Patent issued 19 Oct
    2005 : SBC acquires AT&T Corp
         : AT&T Inc becomes successor name
    2017 : Patent term expires

---

NPE / troll-pattern signals

---

Verdict

Insufficient data — specifically, "only the original assignment" under the rubric's own definition, and that is the substantive finding: no NPE pattern is present.

Justification: The entire recorded chain is a single conveyance — the three AT&T inventors' assignment to AT&T Corp., recorded 1998-04-29. There is no post-issuance assignment of any kind (no shell LLC, no security interest, no merger recording, no aggregator), so there is no transfer chain in which an NPE pattern could appear, and the "current assignee AT&T Inc." listing is Google's corporate-successor inference rather than a recorded assignment I could verify. All eight signals score not present on the available evidence, with signal 3 marked insufficient-data only because the correspondent was not retrievable and a one-link chain cannot exhibit recurrence. Accordingly this reads as a normal operating-company-owned patent (original assignee retained it through expiration in 2017) — the inverse of a troll chain — but because I could not confirm reel/frame values or the correspondent, I stop short of any affirmative ownership characterization beyond the recorded 1998-04-29 assignment.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search "5970473") · mirror — https://assignment.uspto.gov/patent/index.html · patent text — https://patents.google.com/patent/US5970473/en

---

Contradiction / gap flags for the record

1. No reel/frame obtained. Neither the 1998-04-29 AT&T Corp. assignment nor any hypothetical successor recording has a reel/frame in my retrieved data. Anyone relying on this for a title opinion must pull the reel/frame directly from Assignment Center.
2. "Current assignee AT&T Inc." is unsourced as a recorded assignment. It conflicts with the absence of any change-of-name/merger event in the legal-events list. Treat it as an inferred corporate successor (SBC → AT&T Inc., 2005), not a USPTO recording. This nuance is consistent with — and sharpens — the caveat already carried in the previously generated summary.
3. Bulk AT&T portfolio transfers (unverified for this patent). AT&T Corp. patents were, in some eras, bulk-assigned to an AT&T intellectual-property holding entity. I could not verify any such recording for US 5,970,473, and I did not locate a reel/frame. If a future reviewer finds one, it would be a change-of-name / internal-reorg conveyance within the AT&T family — still not an NPE signal — but it would need to be added to the timeline above.

I'll research the prior art cited in US 5,970,473. Let me run several targeted searches.

Let me try to retrieve the patent's front-page "References Cited" (backward citations) directly.

Prior-Art Analysis — US 5,970,473

A required threshold caveat before any of the tables below, because it bears directly on the deliverable you asked for:

The authoritative patent text I was given (fetched from https://patents.google.com/patent/US5970473/en) contains the "Cited By (168)" list — i.e., forward citations — but it does not contain the patent's own front-page "(56) References Cited" list, which is the set of documents that actually constitute prior art against US 5,970,473. My follow-up database searches (patentimages/issued-PDF front pages, OCR of the (56) block, USPTO/USPTO-report mirrors) did not return that face-of-patent reference list before I exhausted my search budget. I therefore cannot truthfully enumerate the applicant/examiner-cited references with the specificity your request asks for, and I will not fabricate them. Below I (1) confirm I have the correct record, (2) explain precisely what citation data I do have and why it is not §102 art for this patent, (3) give the pre-AIA §102 framework and the critical date, (4) do an element-by-element anticipation analysis so you know exactly which claims are exposed, and (5) list candidate prior art explicitly labelled as unverified — not as the cited list — plus the exact sources where the real (56) list can be pulled.

---

1. Confirmation the correct patent was returned

The searches returned the record matching the authoritative text exactly, so there is no number-scrambling in what I report:

Near-number disambiguation (per your "do not return results for similar numbers" rule). One search hit displayed the adjacent entry "5,970,173 A — 10/1999 — Lee et al." immediately beside "5,970,473 A — 10/1999 — Gersberg[z] et al." on the same OCR page. US 5,970,473 and US 5,970,173 are different patents; only the '473 (Gerszberg) is this record. Likewise excluded: US 6,587,473, the Seagate-related "'473," and the TiVo "'473" matters already flagged in the prior sections.

---

2. Why the "citations" I obtained are not §102 prior art for the '473

US 5,970,473 has both a filing date and priority date of 1997-12-31 (no earlier priority claim appears in the record). Therefore, under pre-AIA 35 U.S.C. §102 (which governs this 1997 filing):

• §102(b) statutory-bar critical date = 1996-12-31. Any patent or printed publication dated on or before Dec. 31, 1996 — and any public use/on-sale activity before that date — is a statutory bar that cannot be sworn behind.
• §102(a)/(e) window = 1996-12-31 through 1997-12-31. Documents in this window can still be prior art, with §102(e) covering U.S. patents/applications by another that were filed before 1997-12-31.
• §102(e) is the only route by which a later-published U.S. patent can be prior art — and only if its underlying application was filed before 1997-12-31.

Now apply that to the citations I actually retrieved:

2.1 The "Cited By (168)" list is forward art — it cannot anticipate the '473

Every reference in that list cites US 5,970,473 as prior art, which means each post-dates the '473 (its publication is 1998–2006+ in the sampled entries: e.g., US 6,154,771 (priority 1998-06-01), US 6,351,279 (priority 1998-12-18), WO 00/51050 (priority 1999-02-26), US 6,850,901 (priority 1999-12-17)). A reference that cites a patent cannot be §102 prior art to it. So the 168-entry "Cited By" table tells you about the '473's influence, not its validity. The one entry whose displayed priority (1997-12-29, US 6,215,482, Samsung, "Memory swapping method in a web video phone") superficially looks earlier is a citing document that published 2001-04-10; it cannot be a §102(a)/(b) reference, and I did not verify any application-filing date that could support a §102(e) theory — flagging rather than asserting it.

2.2 The AT&T-family continuations in that list are not prior art either

The forward list includes same-inventor/same-family applications, e.g. US 2005/0254484 A1 (Barzegar, "Multifunction interface facility…," priority listed 1997-12-31) and US 2006/0159116 A1 (Gerszberg, "Facility management platform for a hybrid coaxial/twisted pair local loop network service architecture," priority listed 1997-12-31). These are common-ownership / same-day family members, not "by another," and cannot be §102(a)/(e) art against the '473.

2.3 The one confirmed third-party citation of the '473 is also forward, not backward

The prior litigation section records that PCT/US 03/30766 (published as JP 2006-501725) cites "US 5 970 473 A (Gerszberg)" as an "A"-category reference against another party's claims 14-17 / 34-35 / 46-47. That is the '473 being used as prior art against someone else — the opposite direction from what this task needs, and it is dated 2003+, so it cannot be prior art to the '473 itself.

Net effect: none of the citation data I retrieved is a §102 reference against US 5,970,473. The genuine prior-art set is the (56) block on the issued front page, which I could not pull.

---

3. What the authoritative text does tell us about the art considered

The Google Patents record surfaces "Prior art keywords": user; services; purchase; database server; customer. These are algorithmically extracted concept terms derived from the patent's own text — not a citation list — but they are a useful indicator of the art units/fields the examiner was working in (electronic catalog ordering and database-server-based purchasing). The specification's Background cites no specific prior-art patents; it discusses only general industry practice (DSL splitters, DSLAM switches, ETHERNET-over-DSL, local-exchange vs. interexchange-carrier access). Any anticipatory art therefore comes from the front-page (56) block, not from the narrative.

---

4. Element-by-element anticipation map (what a §102 reference must show)

Because §102 requires a single reference disclosing all claimed elements arranged as claimed, here is the exposure analysis by independent claim:

Claim 1 (system) requires, in combination: (i) a database server storing catalog info and users' shipping data and configured to determine the lowest cost of a selected item; (ii) a plurality of user terminals, each having a unique terminal identifier; (iii) a stored table relating terminal identifiers → shipping addresses; (iv) a touch-sensitive screen displaying catalog info and accepting touch selection; and (v) a control unit generating purchase data sent to the server.
→ No single printed reference located discloses element (iii), the terminal-ID-to-shipping-address table, nor the (i) "lowest cost" determination. Claim 1 is the hardest to anticipate. Likely §102 challenge, if any: combination/§103, not §102.

Claim 11 (apparatus) is the broadest and most §102-exposed claim. It requires only: display means; input means receiving order data; a control unit generating purchase data; and a centralized database server in communication with the control unit that processes the order data. It does not recite the unique-identifier table, the touch screen, or lowest-cost determination. A generic pre-1997 electronic-catalogue/point-of-sale ordering terminal with a central catalogue server could, on its face, read on claim 11. This is where a §102 reference is most likely to land, and it is the claim to attack first.

Claim 12 (method) requires: configuring a terminal with a unique terminal identifier; connecting it to a database server; storing a table relating the terminal identifier → the user's shipping address; displaying catalog info; receiving purchase orders via an input device; and determining the shipping address from that stored table.
→ Same defensive strength as claim 1. The identifier→address lookup step is the distinctive limitation; a reference that merely shows "customer enters address" or "account has an address" would not anticipate. Notably, claim 12 does not itself recite "lowest cost" — that limitation lives only in dependent claim 16 — so a §102(e) reference need not show price optimization to anticipate claim 12 as an independent claim, only the identifier-to-address determination.

Dependent claims (2, 3, 17 profiles; 4, 5, 14 password gating; 6 purchase history; 7, 8 status request/forwarding to vendor; 9, 18 address confirmation; 10 shipping preferences) each add narrow features and are correspondingly harder to anticipate individually.

---

5. Candidate prior-art documents a §102 search must target — UNVERIFIED as to whether they appear on the '473 face

Read this label carefully: the references below are candidates identified from the general art landscape of pre-1997 catalogue/ordering and kiosk systems. I could not confirm that any of them is listed in the (56) block of US 5,970,473. Treat every row as "must be verified against the issued front page before citation." I am not asserting these were cited by the examiner.

Why the §102 exposure is concentrated on claim 11: claim 11 lacks the two features that give the '473 its novelty — the terminal-identifier→shipping-address table and the lowest-cost determination (the latter is in claim 1 only; both are absent from claim 11 and the table alone distinguishes claims 1/12). Consequently:

• §102 anticipation is realistically possible only against claim 11, and only by a reference showing a catalogue terminal + central order-processing server without needing any identifier-to-address mapping.
• Claims 1 and 12 survive §102 on the record I have, absent an undiscovered reference that teaches the identifier-to-shipping-address table (and, for claim 1, the lowest-cost query).

---

6. Where the actual face-of-patent (56) reference list can be obtained

To complete the deliverable you asked for, pull the issued front page (which carries "(56) References Cited") from:

1. USPTO PatentCenter — application US 09/001,421, "References Cited" / IFW front page.
2. The issued-PDF front page via patentimages.storage.googleapis.com (search the specific US5970473 PDF) — the (56) block also lists Foreign Patent Documents and Other Publications.
3. Espacenet / Global Dossier for US 5,970,473 — "Cited documents."
4. The examiner's PTO-892 / Notice of References Cited in the file wrapper, which is the definitive list of what was actually considered.

---

7. Bottom line

• Verified: searches returned the correct patent (US 5,970,473, app. 09/001,421, AT&T, filed/priority 1997-12-31, issued 1999-10-19); near-number US 5,970,173 excluded.
• Not obtainable in this session: the front-page (56) "References Cited" — i.e., the real §102 prior-art set. I will not invent it.
• What the citations I did retrieve are: the "Cited By (168)" forward list and one third-party PCT/JP citation — none of which is §102 prior art against the '473, because each post-dates or was filed after the '473's 1997-12-31 date.
• Pre-AIA §102 framework: §102(b) critical date 1996-12-31; §102(e) window to 1997-12-31.
• Anticipation exposure: concentrated on claim 11 (broad apparatus — no ID/address table, no lowest-cost element). Claims 1 and 12 appear to withstand §102 on the present record because the unique-terminal-identifier→shipping-address table is not taught by the ordinary catalogue/ordering art; any validity attack on them would more plausibly be §103.
• Confidence: high on the bibliographic/date framework and the forward-vs-backward citation distinction; explicitly low/unknown on the identity of the examiner-cited references, which is the one thing this task specifically requested and the one thing I could not verify.

I'll ground this § 103 analysis in actual prior-art references rather than my training memory. Let me first check what the patent's own prior-art (backward citation) record shows, then verify candidate references.

I have strong candidate references. Let me verify the specific identifiers and dates so I don't mis-cite.

Obviousness Analysis — US 5,970,473 under 35 U.S.C. § 103

0. Scope caveat (read first)

The page I was given does not contain a backward-citation "Prior Art" section. The fetched Google Patents text for US5970473A contains only:

• a "Prior art keywords" line (user, services, purchase, database server, customer), and
• a "Cited By (168)" table, which is a list of later documents that cite the '473 — forward citations, which are not prior art.

So I cannot "use the results from the Prior Art section" in the literal sense; there is no such section on this page. Everything below is built from (a) the patent's own specification as quoted in the record, (b) the "Cited By" table as a source of leads that I then had to check for dates (several were same-family AT&T continuations with the 1997-12-31 priority, e.g. US 2005/0254484 A1 and US 2006/0159116 A1 — not prior art), and (c) independently retrieved prior art, with verification status marked throughout. I hit my tool-call limit mid-verification, so § 7 lists exactly what I could and could not confirm.

Governing law: The application was filed 1997-12-31 and issued 1999-10-19, so pre-AIA §§ 102/103 apply (no AIA § 102(a)(1)/(2) "effective filing date" regime). This matters, because the strongest reference below is prior art only through pre-AIA § 102(e).

---

1. Person of ordinary skill in the art (POSITA)

A POSITA here would have a B.S. in electrical engineering or computer science and roughly 2–3 years of experience in telecommunications or client/server data systems, or equivalent. That person would be familiar with: 1990s central-office/subscriber-loop network architecture; client-server databases and relational tables; GUI terminals including touch screens as used in POS/kiosk systems; call-handling telephony services such as ANI/caller ID and 911/vanity-number call routing; and, by 1995–1997, commercial electronic shopping systems (videotex, CompuServe/Prodigy/AOL shopping, Web storefronts, IVR order lines, mail-order catalog systems).

Critical framing point for everything below: the claims are generic. Claims 1, 11, and 12 recite a "database server," "user terminals," a "touch-sensitive screen," a "control unit," and a "centralized database server." They recite no ISD 22, no FMP 32, no NSP 36, no DSL/TVRC modem, no hybrid fiber-twisted pair. The genuinely novel part of the '473 disclosure — the AT&T videophone/DSL access architecture — is described but not claimed. The claims are therefore assessed as what they are: an electronic catalog/ordering system plus an identifier-to-shipping-address lookup, on generic hardware.

---

2. Claim 1 — element-by-element mapping and the primary combination

Claim 1 requires: (i) a database server processing catalog information; (ii) the server storing users' shipping data; (iii) the server being configured to determine a lowest cost of an item selected for purchase; (iv) a plurality of user terminals each with a unique terminal identifier; (v) the server storing a table relating terminal identifiers to shipping addresses; (vi) each terminal having a touch-sensitive screen displaying catalog information and accepting touch selection of an item to purchase; and (vii) a control unit generating purchase data sent to the server.

2a. Primary reference — US 5,710,887 (Chelliah et al., BroadVision) — VERIFIED

• Identity/dates verified (Espacenet bibliographic record; the patent's own front page reproduced as an exhibit in CBM2013-00001, Ex. 1006): US 5,710,887, "Computer system and method for electronic commerce," Appl. No. 520,627, filed 1995-08-29, issued 1998-01-20, assignee BroadVision, Los Altos, CA.
• Why the 1998 issue date does not disqualify it: under pre-AIA § 102(e), a U.S. patent is prior art as of its U.S. filing date. Chelliah was filed 1995-08-29 — more than two years before the '473's 1997-12-31 filing — so it is § 102(e) prior art against the '473 notwithstanding that it issued after the '473 was filed. It is "by another" (BroadVision, not AT&T), and pre-AIA § 103(c) common-ownership disqualification therefore does not apply. A § 102(e) reference can be combined with other art for § 103.
• What it discloses, mapped to claim 1: a system facilitating commercial transactions "between a plurality of customers and at least one supplier … over a computer driven network"; a customer site with "an associated display and an input device"; an item database "associated with a supplier and including information on presented items"; a pricing subsystem that "receives information from the item database to determine the cost associated with a presented item"; a customer information database "for storing information relating to a customer"; a per-customer customer monitoring object that responds to enquiries, receives a selection of a presented item, determines and presents cost, and "pass[es] a delivery initiation communication to initiate the delivery of the item"; and order fulfillment initiation means.

That is elements (i), (iv) (multiple customer sites + a customer information database), (vi) (display + input selection), and (vii) (the control-object equivalent) — plus delivery initiation, i.e., the shipping step. Its claims 3, 11, 12, and 20 add order-fulfillment initiation, security-code validation before payment, order confirmation to the customer, and an observation subsystem that stores customer interaction information.

2b. Secondary reference for the core novelty — Shaffer et al., "Automatic information and routing system for telephonic services" family — PARTIALLY VERIFIED

• Verified content (from the specification text of US 6,381,324 B1 and US 8,085,924 B2, and from an IDS page in US 6,385,312 referencing U.S. App. No. 08/598,392, filed June 19, 1997): the system captures a caller-provided or network-provided telephone number, indexes it through a "Master Telephone Number to Spatial Key" table, and retrieves the caller's CASS-certified USPS address, which the VRU "speaks … back to the caller for confirmation."
• It further discloses "eliminating servicing locations based on days and hours of operation and/or services offered," including the express example: "having the caller select a pickup or delivery option, (for pizza, for example) and eliminating servicing locations from the list that are not currently open or do not offer the desired service." It also expressly frames the mechanism as anti-fraud: "If the addresses are different, there is a high potential that the caller is trying to make a fraudulent order and the operator would ask additional questions."
• Why it matters: this single reference supplies elements (ii)/(v) — the identifier→stored-address table and its use to derive the shipping address — element (iii)'s geographic filtering analog (which vendors service this address), and dependent claims 9/18 (address read back for confirmation), and an explicit motivation (fraud reduction) that mirrors the '473's stated benefit.
• Date caveat: the parent application 08/598,392 is dated 1997-06-19 (before 1997-12-31), but I was not able to verify within budget the precise continuation chain by which US 6,381,324 / 6,385,312 / 8,085,924 claim benefit of it, nor their issue dates. A formal § 102(e) date must be confirmed in PatentCenter before relying on this reference in a filing.

2c. Secondary reference for the touch screen — Mueller et al., US 5,353,219 — NUMBER/TITLE VERIFIED FROM A SECONDARY SOURCE; DATE NOT VERIFIED

"Suggestive Selling in a Customer Self-Ordering System": "A retail store based, touch screen system is used for direct entry by customers of orders. A suggestive selling subroutine displays a screen suggesting items … For example, on a fast food restaurant touch screen order entry system…" (quoted in the background of the later Katz patent US 6,055,513). A 5,3xx,xxx number is consistent with pre-1997 issuance, but I did not confirm its issue date.

2d. The combination and the motivation

Combination 1 (claims 1 and 12 core): Chelliah '887 + Shaffer (identifier→address). Rationale under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) and MPEP 2144:

1. Known technique used to improve a similar device in the same way (MPEP 2144.03 / KSR factor (C)). Chelliah already taught the full browse→select→price→purchase→deliver loop and already stored customer data in a customer information database. Shaffer taught that a live communications session carries an inherent, machine-readable customer identifier (the calling number) that can be mapped through a stored table to a postal address — an application that grew directly out of 911/ANI address-lookup practice. Wiring Shaffer's lookup table to Chelliah's customer information database is the textbook "known technique to improve a known device" combination.
2. Explicit problem statement and articulated benefit (the strongest motivation here). The '473 specification itself supplies the motivation: "the customer does not need to, for example, provide a delivery address to every vendor"; "there is no need for the customer to enter a delivery address each time a purchase is made"; and, on fraud, "since access to the videophone in a home or a business is required to make purchases, the likelihood of fraudulent use of credit and charge cards is decreased." Shaffer independently recites the same two benefits — relieving the customer of re-entering address data and detecting "a fraudulent order" when the retrieved address mismatches. Two references converging on the same stated advantages is about as clean a motivation-to-combine showing as exists.
3. Simple substitution / predictable result (KSR factor (B)). Replacing Chelliah's manually-entered or per-store customer data with an automatically-resolved identifier→address table is substitution of one known data-acquisition mechanism for another, with no change in the underlying commercial process and no unexpected result.
4. Market forces (KSR factor (F)). The 1994–95 trade press cited on Chelliah's own face — Home PC, Dec. 1, 1994 ("Shopping at Home: One Modern Line, No Waiting"); Advertising Age, Dec. 19, 1994 ("Cashing In: The Rush is on to Buy and Sell on the Internet"); Response TV, July 1995 ("Cyberselling: Shopping Comes to the Web") — establishes that online retailing and its friction points (checkout data entry) were a hot commercial problem area well before the '473 filing. Reducing checkout friction and reducing chargeback fraud were recognized design incentives.

Combination 2 (the touch-sensitive screen): Chelliah + Mueller '219. Chelliah's "display and an input device" is functional; Mueller '219 discloses the specific touch-screen self-ordering terminal, in a retail environment, at least two years before the '473 filing. Applying a touch-screen order-entry convention to a home ordering terminal is a simple substitution of a known element to obtain a predictable result (KSR factor (B)); the '473 specification does nothing with the touch screen beyond using it as an input device and for "soft keys." No teaching away was identified.

---

3. Claim 1's "lowest cost" limitation — the one element I could not fully ground

Caveat: I did not verify a pre-1998 "best price / lowest cost" reference. My search for a price-comparison patent returned only a later document (US 6,405,175, ~2002, a product-and-price database that "each include[s] a lowest price for the product and an online store … selling the product at that price") — which postdates the '473 and is therefore useless as prior art. I flag this rather than assert a reference number I could not confirm (I explicitly declined to assert US 5,664,115, whose number/title/dates I could not verify, and likewise the 5,319,542 / 4,992,940 leads).

What I can say on the record:

• Chelliah already discloses a multi-supplier mall — "a plurality of customers and at least one supplier," with each store carrying "a particular" configuration and its own pricing, and a pricing subsystem that "determine[s] the cost associated with a presented item." Extending a per-supplier pricing subsystem to return the minimum across suppliers is a comparison of stored numbers — a predictable algorithmic variation requiring no new hardware or new data (KSR factor (A); cf. In re Grams line of reasoning that a result computed from known data by known steps is not a patentable advance). "Best price" guarantees and comparison shopping were ordinary commercial practice by the mid-1990s.
• Practical recommendation: to make the § 103 record airtight on this element, add a dedicated pre-1998 reference teaching price comparison across multiple vendors (an electronic "comparison shopping" service, a reverse-auction system, or an interactive-TV shopping service that lists multiple retailers' prices). This is the element most exposed to a patent owner arguing "no reference teaches minimizing across vendors."

---

4. Claim 11 (apparatus) and claim 12 (method)

Claim 11 is a means-plus-function-style echo of claim 1: display means, input means, a control unit generating purchase data, and a centralized database server processing order data based on the purchase data. Chelliah's own claims are drafted in precisely this means-plus-function form ("means for causing at least one supplier to be represented on the display for selection …"; "means for effecting presentation of items …"; "customer information database …" ; "means for … passing a delivery initiation communication"). A single reference therefore maps claim 11 nearly element-for-element; claim 11 is obvious over Chelliah alone, and a fortiori over Chelliah + Shaffer. If claim 11 is construed under § 112(f), the corresponding structure disclosed in the '473 (generic controller 170, memory 173, NSP server 183) is itself generic and would have been obvious in view of Chelliah's equivalent generic computer hardware.

Claim 12 is the method counterpart and is broader in one respect: it does not include the "lowest cost" limitation (that first appears in dependent claim 16). Claim 12 requires only: configure a terminal with a unique identifier → connect to a database server → store an identifier-to-shipping-address table → display catalog information → receive orders via an input device → determine the shipping address from the stored table. The Chelliah + Shaffer combination maps onto claim 12 directly, with Shaffer supplying the "unique terminal identifier" (captured telephone number), the "table," and the "determining" step, and Chelliah supplying catalog display, ordering via input device, and server-side processing. Claim 12 should be the primary § 103 target because it is the broadest independent claim that captures the alleged novelty.

---

5. Dependent claims — mapping and secondary references

Aside (§ 112, not § 103): claim 3 recites "wherein user profile comprises …" but depends from claim 1, which does not recite any "user profile" — a lack of antecedent basis that claim 2 (from which claim 3's term evidently migrated) would have supplied. Worth noting if claim 3 becomes the subject of argument.

---

6. The KSR/MPEP rationale ledger (what a petition or invalidity contention would say)

Defensive notes a patent owner would raise, and the responses: (1) § 102(e) disqualification via § 103(c)? No — Chelliah is BroadVision-owned, not AT&T-owned, so pre-AIA § 103(c) does not remove it. (2) Swear behind? The '473's only priority date on the record is 1997-12-31 (the page shows "Applications Claiming Priority (1)" = US 09/001,421 and no earlier priority), so overcoming Chelliah (1995) or Shaffer (mid-1997) would require proving an invention date earlier than those, which is unlikely on this record. (3) Secondary considerations? None is evident on the face of the patent; if asserted (commercial success of AT&T videophone services), the nexus would be to the unclaimed network architecture, not to the claimed generic system.

---

7. Verification status — what I confirmed and what I did not

Verified:

• US 5,710,887 (Chelliah et al.), "Computer system and method for electronic commerce," Appl. No. 520,627, filed 1995-08-29, issued 1998-01-20, BroadVision — via Espacenet bibliographic data and the patent's full front page reproduced as Ex. 1006 in CBM2013-00001; its abstract, claims, and face-cited trade publications.
• Shaffer-family content: caller-number→USPS-address retrieval with spoken confirmation, service-area elimination (express pizza example), and the fraud motivation — via the specification text of US 6,381,324 B1 and US 8,085,924 B2 and an IDS page in US 6,385,312 naming App. No. 08/598,392, filed 1997-06-19.
• JP 特開平11-098253 (JP App. No. Hei 9-256308, filed 1997-09-22, published 1999-04-09): a telephone order system that reads the originating telephone number, keys a customer database to retrieve name and address, and performs shipping — via J-GLOBAL.

Could not verify (do not cite without checking):

• Issue dates of US 5,353,219, US 5,576,951, US 5,583,763, US 5,615,342, US 5,625,776, US 5,053,957, US 5,481,094, US 4,996,642, US 4,870,579 — numbers/titles are quoted from another patent's background, not from the documents themselves.
• The exact § 102(e) benefit chain and issue dates of US 6,381,324 / 6,385,312 / 8,085,924 back to App. No. 08/598,392.
• The filing date of US 6,055,513 / US 8,290,829 (Katz), which I would want for the "extended warranty/upsell" and ANI-identification points.
• Any pre-1998 "lowest price across vendors" reference (my search returned only US 6,405,175, ~2002 — post-dates the '473).
• I deliberately did not assert US 5,664,115, US 5,319,542, or US 4,992,940 as references because I could not confirm their identifiers, titles, or dates.

Important non-prior-art exclusion: The JP Hei 9-256308 application, although filed 1997-09-22 (before the '473 filing), published 1999-04-09 — after 1997-12-31. As a foreign application publication it is not § 102(e) art, and it is not § 102(a)/(b) art either, because it did not become public before the '473's filing date. I cite it only as evidence that the identifier→address technique was being independently developed in the same window — not as an available § 103 ground. Likewise, the concurrently-filed AT&T applications incorporated by reference by the '473 (including "Restaurant Ordering Via VideoPhone" and "Ticket Ordering Via VideoPhone") are same-day, common-ownership materials: not "prior art by another," and pre-AIA § 103(c) would in any event disqualify them for § 103 purposes.

---

8. Bottom line

On the record I have, claims 1, 11, and 12 are obvious under § 103 over US 5,710,887 (Chelliah) in view of Shaffer et al.'s identifier→address telephone-routing teachings, further in view of a touch-screen self-ordering reference (e.g., US 5,353,219) — with the "lowest cost" element of claim 1 and the shipping-preferences element of claim 10 being the two limitations needing an additional reference to make the record airtight. The motivation is unusually clean because the two lead references recite the same benefits the '473 asserts (no re-entry of shipping data; reduced order fraud), and because the '473's actual architectural novelty — the ISD/FMP/NSP DSL loop — is described but not claimed, so it cannot rescue the generic claims.

I flag one contradiction/gap against the task framing: I found no "Prior Art" (backward-citation) section on the fetched page, only a forward "Cited By (168)" table; treating that table as prior art would be a legal error, since it consists of later documents (several of them same-family AT&T continuations with the 1997-12-31 priority). No contradiction was found with the previously generated sections — this analysis is consistent with their conclusion that the patent is expired (2017-12-31) and has no PTAB or litigation history.

Confidence: High that Chelliah '887 is available § 102(e) art and maps to the bulk of claims 1/11/12. High that Shaffer-family teachings supply the identifier→address table, the confirmation step (claims 9/18), and the "which vendors serve this address" filtering (the FIG. 5B pizza example). Moderate on the § 102(e) chain for the Shaffer family and on all dates I could not verify. Low on any specific pre-1998 "lowest cost" reference — that limitation is my recommended next acquisition.