US 5634051

I'll verify the current status of US 5634051 against live sources before summarizing.

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

Verification summary — US 5,634,051

I checked live sources (Google Patents, USPTO-side reproductions, litigation aggregators, CAFC June 2026 scheduled-cases listing) for this exact number. Searching for the literal string "5634051" in 2026 CAFC/PTAB docket material returned no matter involving this patent. What surfaced were unrelated items with coincidentally matching numbers (e.g., a Brazilian appellate case nº 5634051-87.2019.4.03.9999; an art-unit "5634051" string inside a PTAB petition exhibit). I do not have authoritative confirmation of any 2026 Federal Circuit appeal docketed under US 5,634,051.

---

Bibliographic data

Assignment history of record: security interests to Silicon Valley Bank (2000-02-01, 2004-02-23), Sopheon PLC (2000-06-30), merger into Sopheon Corporation (2001-03-28), releases in 2007-07-02 and 2014-03-12.

Litigation of record: Google Patents flags "Family has litigation," citing Texas Western District Court case 1:11-cv-00332 (Unified Patents data) and a Darts-IP family-litigation entry (family 22510042). Note that case numbering (2011 vintage) is far outside 2026.

---

Abstract (verbatim)

"An information storage, searching and retrieval system for large (gigabytes) domains of archived textual dam [sic]. The system includes multiple query generation processes, a search process, and a presentation of search results that is sorted by category or type and that may be customized based on the professional discipline (or analogous personal characteristic of the user), thereby reducing the amount of time and cost required to retrieve relevant results."

---

Independent claims — plain-language overview

Claim 1 — System (broadest storage/classify/summarize claim). A system that (a) stores a large domain of data held in many source records, where some source records themselves contain documents of multiple document types; (b) searches at least a substantial portion of that data with a query to find responsive documents of multiple types; and (c) categorizes the responsive documents by document type, including generating a summary showing how many responsive documents fall into each of various predetermined document-type categories.

Claim 10 — System (database-oriented variant). Same architecture, but the storage element is recited as storing data "contained in multiple document types" rather than in source records, and the search is over "at least a portion" of the data. Categorization and a count-by-category summary are still required.

Claim 14 — System (source-record-independent sorting). Adds the distinguishing limitation that categorization is done based on document type and independently of the source record from which the documents came, and the summary counts responsive documents falling within each of the document types. This is the claim that most directly captures the patent's stated advance over Dialog-style, source-organized systems.

Claim 17 — Method (core method). Stores the large multi-source domain electronically (some source records containing multiple document types); generates an electronically executable query; searches at least a portion of the data for multiple document types; then sorts the responsive documents and presents a summary of counts by document type, independently of the source record.

Claim 18 — Method (category-set selection). Adds administrative structure: define one or more sets of categories of document types, each category mapping to one or more document types; run the query and search; select one of the category sets for reporting; and sort/summarize the responsive documents by category using the selected set. (Dependent claims 19–21 add automatic selection based on a user code or personal characteristic, selection by the user's professional class, and searching substantially all of the data.)

Claim 22 — System (many-to-one category mapping). Recites multiple predetermined sets of categories of document types where at least one category in at least one set corresponds to more than one document type (e.g., product announcements + product reviews + product specifications → "Product Information"), plus summary generation of counts for the categories of one such set.

Dependent-claim architecture: 2–9 depend from 1 (category sets; automatic selection by user code/characteristic or professional class; user predetermination and customization of categories); 11–13 depend from 10; 15–16 depend from 14; 19–21 depend from 18.

---

Notable drafting anomaly (flagged literally, not corrected)

Claim 20 reads: "The method of claim 16 wherein the step of selecting one of the sets of categories is performed automatically based on a code identifying the professional class of the user…" — but claim 16 is a system claim dependent on claim 14, and the "selecting one of the sets of categories" step appears in method claim 18. Read literally, claim 20's dependency reference is to a system claim, which is an apparent error in the printed patent. I am reporting the text as issued.

Prosecution/related-document notes of interest

• The parent application 08/144,767 (1993-10-28) was abandoned; the granted patent issued from the 1996 continuation.
• WO 1995012173 A2 (Teltech Resource Network Corp., published 1995-05-04) is the PCT counterpart and appears in later third-party search reports as cited prior art (e.g., cited as category "A" against JP 2006-502510).
• The patent is cited by hundreds of later documents (233 cited-by entries listed), including Oracle's US 6,035,560 and US 6,460,034 and Sun's US 6,055,540.

---

Confidence and uncertainty

• High confidence on title, inventor, filing/issue dates, priority chain, abstract text, claim text, and the 1993 priority / 2013 lapse dates — these come directly from the full patent text and Google Patents bibliographic data.
• Moderate confidence on current assignee (Sopheon Corp.). The patent text and assignment records support the Teltech → Sopheon chain, and Sopheon's 2003 Form 20-F describes the patent as "Sopheon's (formerly Teltech's) 'Presentation of search results from large domains of archived textual data by category or type'… for which a US patent was granted in 1997." The FY2002 20-F also notes the US information-management business (including Teltech) was divested to FIND/SVP, Inc. effective 2003-07-01, so a chain-of-title to that business line is plausible but I have not verified a recorded assignment of this patent to FIND/SVP or any later owner.
• Low confidence / no evidence for any 2026 CAFC activity. Given the patent expired 2013-10-28, there is no live patent right to assert in 2026; any Federal Circuit involvement would have to be a residual appeal from much older litigation, and my searches found none. I could not access a full-text CAFC docket search directly, so I cannot state categorically that no such appeal exists.
• The recorded district-court litigation (1:11-cv-00332, W.D. Tex.) is from 2011 and predates expiration; I did not retrieve the complaint, parties, or outcome.

Known litigation involving U.S. Patent No. 5,634,051

Short answer: The public record for this patent is thin. I was able to confirm that litigation activity is referenced for the '051 patent family, but I was not able to fully verify parties, case numbers, and outcomes for each matter within my search budget. Below I separate what is confirmed from what is unverified, and I flag one search hit that is very likely a false positive.

⚠️ Methodology caveat up front: A large number of documents returned by the search simply cite "5,634,051" as prior art (e.g., IPR exhibit lists, Information Disclosure Statements such as the one filed in [Cisco Systems Inc.](/litigations/by-plaintiff/Cisco%20Systems%20Inc.) v. TracBeam LLC, IPR2018-01725, and the PCT search report for WO2001015010A3, which lists "US 5 634 051 A (THOMSON WILLIAM K) 27 May 1997"). A citation in an IDS or IPR petition is not litigation of the '051 patent. Likewise, the Brazilian state-court docket number "5634051‑87.2019.4.03.9999" (TRF‑3, Brazil) is a Brazilian case number, not this patent, and should not be conflated with it. Per your instruction I have not auto-corrected any of these identifiers.

---

1. Litigation affirmatively flagged on the face of the patent record

Sopheon / Teltech v. [[Unknown Defendant]](/litigations/by-defendant/%5BUnknown%20Defendant%5D) — U.S. District Court for the Western District of Texas

This appears in the Google Patents "Family has litigation" field as: https://portal.unifiedpatents.com/litigation/Texas%20Western%20District%20Court/case/1%3A11-cv-00332 (jurisdiction attributed to "District Court"). I was unable to open the Unified Patents case page or PACER within my search budget, so I cannot name the parties or state an outcome. The "1:11" prefix indicates it was filed in 2011 in the Austin Division of the Western District of Texas.

First worldwide family litigation (Darts‑IP family ID 22510042)
Google Patents also records a "First worldwide family litigation filed" entry pointing to https://patents.darts-ip.com/?family=22510042…. Darts‑IP is a commercial litigation database; the underlying document was not retrievable in my search, so I cannot independently confirm which case this corresponds to — it may well be the same W.D. Tex. matter above.

2. A related but very likely unrelated case (must be excluded unless verified)

Pherah, LLC v. Sopheon Corporation

Caution: In this case Sopheon was the defendant, not the plaintiff. Since Teltech/Sopheon has owned the '051 patent throughout, this is most likely an assertion of a different patent against Sopheon, not an assertion of the '051 patent. I could not verify the patent-in-suit, so I do not assert that 5,634,051 was involved. I list it only because a naive search of "Sopheon + patent litigation" surfaces it, and you should not mistake it for '051 litigation without checking the complaint (RPX Insight litigation ID txedce-169748-pherah-v-sopheon).

---

3. What I could NOT establish

• I found no verified assertion, complaint, IPR/PGR, ITC investigation, or CAFC appeal in which U.S. Patent No. 5,634,051 itself was the asserted patent.
• I found no confirmed outcome (judgment, settlement, dismissal, invalidity holding) for the W.D. Tex. 1:11-cv-00332 matter.
• I did not find any CAFC opinion, PTAB proceeding, or ITC Section 337 action naming '051.
• I did not exhaust the Unified Patents case-list, RPX Insight, CourtListener/RECAP, PACER, or Darts‑IP family 22510042.

4. Context that bears on enforcement

• Prosecution/ownership: The '051 patent issued May 27, 1997 from Application No. 08/585,383 (filed Jan. 11, 1996), a continuation of Ser. No. 08/144,767 (filed Oct. 28, 1993, now abandoned); priority date Oct. 28, 1993. Inventor: William K. Thomson. Original assignee Teltech Resource Network Corp.; later assignments to Silicon Valley Bank (security interests, 2000 and 2004), SOPHEON PLC (2000), and SOPHEON CORPORATION (merger, 2001). Google Patents lists current assignee as Sopheon Corp and status as Expired – Lifetime, with an anticipated expiration of Oct. 28, 2013.
• Enforcement implication: Because the patent expired (per the anticipated-expiration field) on Oct. 28, 2013, any litigation would have had to be filed before that date to recover damages, and the only live remedy thereafter would be for past infringement within the six-year lookback. This makes the 2011 W.D. Tex. case the most plausible enforcement event and tends to corroborate that later Sopheon-involved suits (e.g., Pherah) did not concern '051.

5. How to close the gaps (recommended verification)

1. Unified Patents portal: the exact URL stamped on the patent — https://portal.unifiedpatents.com/litigation/Texas%20Western%20District%20Court/case/1%3A11-cv-00332 — should yield the party names for 1:11-cv-00332.
2. CourtListener / RECAP and PACER (W.D. Tex.) for the full docket and disposition of 1:11-cv-00332.
3. Darts‑IP family 22510042 for the "first worldwide family litigation" detail and any foreign case.
4. RPX Insight record txedce-169748-pherah-v-sopheon and the underlying E.D. Tex. complaint to confirm the patent(s)-in-suit in Pherah.
5. CAFC docket search for "5,634,051" / "5634051" (I found none).

If you want, I can run these targeted lookups next and return a fully sourced case table with parties, filing dates, and dispositions.

Bottom line: One litigation matter is affirmatively associated with the '051 family in the patent record — a 2011 Western District of Texas case, No. 1:11-cv-00332 — but I could not verify its parties or outcome. I have no confirmed instance of a case in which 5,634,051 was the asserted patent, and I found no CAFC, PTAB, or ITC activity involving it. The Pherah v. Sopheon case is a probable false positive for '051 and should be excluded unless the complaint is checked.

I'll start by verifying the structured data against public sources, since the canonical block says no AIA proceedings are on file.

Let me run a few more targeted searches to be thorough before concluding.

Proceedings overview

Zero AIA trial proceedings are on file against US 5,634,051. The USPTO ODP structured block returns no proceedings, and my independent web checks (PTAB/P-TACTS, Darts-IP family data, Unified Patents litigation data) surfaced no IPR, PGR, or CBM naming this patent as the challenged patent. The bottom line for a defendant: there is no PTAB record to lean on — no claims have been canceled, no FWD exists, and no institution decision has narrowed the patent. Any defense must be built from scratch (and validated against the pre-AIA statutory limits described below), not from an existing PTAB win.

Verification caveat: I could not exhaustively page P-TACTS from this environment. The single recurring web hit for the string "5634051" in PTAB-hosted documents is a prior-art search string (a Westlaw/patent-database query listing many patent numbers) inside an unrelated petition's exhibit — not a challenge to this patent. I am not aware of any proceeding and found none, but treat that as a high-confidence negative rather than a certified null.

Proceedings by number

None. There are no IPR####-#####, PGR####-#####, or CBM####-##### dockets to report for US 5,634,051.

Why the absence is structural, not just statistical

A defendant should understand why this patent has never been through an AIA trial — because the same reasons constrain your own options:

• Filing date / priority: 1993-10-28 (continuation of App. No. 08/144,767, filed 1993-10-28; this application filed 1996-01-11; granted 1997-05-27). This is a pre-AIA patent.
• PGR is statutorily unavailable. Post-Grant Review reaches only patents having at least one claim with an effective filing date on or after 2013-03-16. This patent's 1993 priority date puts it entirely outside PGR.
• CBM is a closed door. The transitional Covered Business Method program sunset on 2018-09-16, and CBM was limited to "covered business method" patents (claims directed to a financial-services practice). Even if this patent arguably qualified, the program no longer exists.
• The patent is expired. Google Patents records an anticipated expiration of 2013-10-28 (20 years from the 1993 filing), and the Darts-IP/reassignment history reflects the last security-interest release in 2014. An expired patent carries no injunctive leverage, and post-expiration IPR practice is disfavored — the practical incentive to petition is near zero.
• The § 315(b) window has long closed for the known litigation. The structured data records a Texas Western District Court case, docket 1:11-cv-00332. Any defendant served in that 2011 action had one year from service to petition for IPR; that window lapsed over a decade ago. A defendant not served in that case would run its own clock from its own service date — but see the expiry problem above.

What this means practically: the only realistically available AIA vehicle today would be an IPR under 35 U.S.C. § 311 (§ 102 / § 103 on patents and printed publications only — no § 112, no public-use, no prior-product art). Given expiration, an IPR would be an exercise in cleaning up a paper record rather than neutralizing a live threat.

Strategic summary

Claim status — CANCELED / SUSTAINED / UNTESTED. All 22 claims of US 5,634,051 (independent claims 1, 10, 14, 17, 18, and 22, plus dependents) are UNTESTED at the PTAB — none canceled, none sustained, none construed by the Board. There is no IPR narrowing to point to and no surviving-claims list to hand a client, because no claim has ever been adjudicated in an AIA trial.

Estoppel landscape — this is largely a non-issue. Because there are no PTAB proceedings, there is no § 315(e)(1)/(2) estoppel and no petitioner/privity bar tied to this patent. Any prior-art ground you wish to raise — § 102 or § 103 on patents/printed publications, or a § 282 invalidity defense in district court (including § 112 and prior public use/sale) — remains available to you, subject only to the ordinary district-court rules and the patent owner's right to respond. Conversely, there is no FWD admission or cancellation for the patent owner to hide behind, and no Board claim construction to constrain how a court reads the claims.

Pattern signals.

• Repeat petitioners: None — no petitioner has ever filed against this patent, so no repeat-filer pattern exists.
• Patent-owner PTAB aggression: None observed. The patent owner never had occasion to defend an AIA trial or appeal an FWD against this patent, so there is no Federal Circuit appeal on record from any PTAB decision here. (Note the ownership lineage is active in other ways: original assignee Teltech Resource Network Corp, merger/reassignments to Sopheon PLC then Sopheon Corporation, with Silicon Valley Bank security interests recorded in 2000, 2004, and released/released-by-secured-party in 2007 and 2014.)
• Defensive aggregator: Unified Patents does not appear in the chain. The only reference to this patent in the Unified/Darts-IP litigation metadata is the Texas Western case 1:11-cv-00332; there is no Unified-filed challenge on file.

Recommended next steps

1. Do not represent to a court or a client that any claim of 5,634,051 has been invalidated. It has not. There is no FWD to quote, and I found no institution decision.
2. Rebuild the invalidity case affirmatively. With no § 315(e) estoppel and no Board claim construction, your best PTAB-side option (if genuinely needed) is an IPR on § 102/§ 103 patents-or-printed-publications art — but weigh it against the 2013-10-28 expiry, which undercuts any enforcement threat and therefore the value of an IPR.
3. Confirm the null result directly before relying on it. Pull the patent's case history in USPTO PTAB E2E/P-TACTS (search by patent number) at https://ptacts.uspto.gov/ptacts/ui/home and cross-check the patent's Google Patents "PTAB proceedings" tab at https://patents.google.com/patent/US5634051/en. Because I could not exhaustively page P-TACTS, a direct lookup is the prudent final confirmation step — this is exactly the case where the absence of proceedings is itself the finding, and it should be verified from the authoritative source.
4. Check the litigation record for non-AIA validity outcomes. The 2011 Texas Western case (1:11-cv-00332) may contain a district-court invalidity or validity ruling or a summary-judgment decision — that is a different estoppel/issue-preclusion analysis than PTAB estoppel, and it is worth running down on PACER/CourtListener even though it is not an AIA trial.

---

Constraint compliance / honesty notes: I have not invented any proceeding number, petitioner, panel, or decision — because there are none to report. Where I state "no Federal Circuit appeal," that flows directly from "no FWD exists," not from an exhaustive CAFC docket pull. The two hard data points I am anchoring on are (a) the USPTO ODP structured block returning zero AIA trials, and (b) the 2013-10-28 anticipated expiration recorded on Google Patents. Everything about statutory availability (PGR date cutoff, CBM sunset, § 315(b) one-year bar) reflects the governing law as applied to those facts.

I'll research this patent's ownership chain across multiple sources.

Let me dig into the litigation and assignment records specifically.

Ownership reconstruction — US 5,634,051 ("Information management system")

Sourcing note / limitation. This memo is built from the Google Patents legal‑events record for US5634051A (fetched 2026‑10‑01) cross‑referenced with Crunchbase, Sopheon plc RNS/20‑F disclosures, and trade press. The USPTO Assignment Center database (assignmentcenter.uspto.gov / assignment.uspto.gov) exposes reel/frame numbers and the correspondent of record, but those fields were not retrievable in this session; the Google Patents legal‑events feed exposes the conveyance, assignor, assignee and date only. I therefore do not report reel/frame or correspondent values — reporting them without the record in hand would mean fabricating them. Dates below are the dates carried on the legal‑events feed, which generally correspond to the recording date at USPTO rather than the execution date.

---

Inventors

Unusual-pattern check: Only one inventor is named, so the classic "all inventors departed the assignee within 12 months of filing" fire‑sale precursor cannot be assessed (n=1). The continuation was filed ~27 months after the parent, which is ordinary prosecution practice and not a distress signal. No inventor assignment quirks observed.

---

Original assignee

Teltech Resource Network Corporation — Minneapolis, Minnesota.

• Line of business: independent provider of business‑to‑business research and knowledge‑management consulting services; it ran a large archive of technical/business documents and an expert‑matching service — exactly the "large domain of archived textual data" the patent describes. So the patent is a genuine operating‑company invention, not a paper asset.
• Product embodying the claims: Teltech's research/portal service implemented the claimed "search the whole domain, then summarize hits by document‑type category." This is corroborated by Sopheon's later SEC Form 20‑F, which lists "Sopheon's (formerly Teltech's) 'Presentation of search results from large domains of archived textual data by category or type'" among its most significant IP assets (bamsec.com copy of Sopheon Form 20‑F 2002).
• Current status: Acquired / absorbed. Teltech shareholders approved the merger on 2000‑09‑08 in Minneapolis and the deal closed in September 2000 (~$35M, ~$15M cash plus Sopheon stock); Teltech CEO Andy Michuda became Sopheon's CEO and founder Joe Shuster joined the Sopheon plc board. Teltech no longer exists as a separate entity.

---

Assignment timeline

All entries are recorded events from the Google Patents legal‑events feed (Reel/frame not retrieved — see sourcing note). Correspondent not retrieved for any entry.

• 2000‑02‑01 (recorded) — Reel not retrieved

  • Conveyance: Security Interest ("see document for details")
  • Assignor: Teltech Resource Network Corporation
  • Assignee: Silicon Valley Bank
  • Correspondent: not retrieved.
  • Context: Securitization — a bank taking a lien on Teltech's IP (typical venture‑debt collateral).

• 2000‑06‑30 (recorded) — Reel not retrieved

  • Conveyance: Security Interest ("see document for details")
  • Assignor: Teltech Resource Network Corporation
  • Assignee: Sopheon PLC
  • Correspondent: not retrieved.
  • Context: Acquisition‑related financing/security granted to the buyer (Sopheon) during the run‑up to the merger.

• 2001‑03‑28 (recorded) — Reel not retrieved

  • Conveyance: Merger
  • Assignor: Teltech Resource Network Corporation
  • Assignee: Sopheon Corporation
  • Correspondent: not retrieved.
  • Context: Corporate merger / acquisition — the one true transfer of title in the chain; Teltech merged into the Sopheon group.

• 2004‑02‑23 (recorded) — Reel not retrieved

  • Conveyance: Security Interest ("see document for details")
  • Assignor: Sopheon Corporation
  • Assignee: Silicon Valley Bank
  • Correspondent: not retrieved.
  • Context: Securitization — post‑acquisition refinancing lien on the Sopheon US entity's IP.

• 2007‑07‑02 (recorded) — Reel not retrieved

  • Conveyance: Release
  • Assignor: Silicon Valley Bank
  • Assignee: Sopheon Corporation
  • Correspondent: not retrieved.
  • Context: Lien release — SVB discharging its security interest (internal housekeeping).

• 2014‑03‑12 (recorded) — Reel not retrieved

  • Conveyance: Release by Secured Party ("see document for details")
  • Assignor: Sopheon PLC
  • Assignee: Sopheon Corporation
  • Correspondent: not retrieved.
  • Context: Internal corporate lien release/clean‑up within the Sopheon group.

• Non‑assignment event: 2013‑10‑28 — anticipated expiration (patent term end); no transfer.

Current assignee of record: Sopheon Corporation (now part of the Wellspring group after Wellspring's February 2024 take‑private of Sopheon). The chain terminates in an operating company, not an NPE.

---

Timeline diagram

timeline
    title Ownership of US 5634051
    1993 : Priority application filed
    1996 : Continuation filed by Teltech
    1997 : Patent issued
    2000 : Bank security interest recorded
         : Sopheon security interest recorded
    2001 : Teltech merged into Sopheon Corp
    2004 : Bank security interest recorded
    2007 : Bank lien released
    2013 : Patent expired
    2014 : Secured party release recorded

---

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. Every named party is an operating enterprise or a bank: Teltech Resource Network Corporation, Sopheon PLC, Sopheon Corporation, and Silicon Valley Bank (secured party only). No entity with an "IP / Patents / Licensing / Holdings / Ventures" suffix and no single‑purpose LLC anywhere in the chain (events 2000‑02‑01 through 2014‑03‑12).

2. Known asserter in the chain — Not present. None of the assignees match any public NPE list (Acacia, Marathon, Intellectual Ventures, IPNav, Wi‑LAN, Conversant/Mosaid, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities, etc.). Sopheon was a publicly listed operating software vendor (LSE: SPE, AIM) throughout the relevant period.

3. Repeat correspondent across the chain — Unclear. The correspondent of record was not retrievable in this session, so I cannot test for a recurring recording attorney/firm. This signal is genuinely unassessed, not negative — flag for follow‑up via the Assignment Center.

4. Cascading transfers — Not present. There is one title‑transferring event (the 2001‑03‑28 merger). The remaining five recorded events are liens and releases spanning 2000–2014 — the opposite of a rapid LLC‑to‑LLC cascade in under 24 months.

5. Pre-litigation transfer — Not present / unclear. No assignment is dated within 6 months before any suit. The sole title transfer (2001‑03‑28) predates the only litigation entry (W.D. Tex. 1:11‑cv‑00332, per the Unified Patents litigation link on the Google Patents page) by a decade. I could not verify the parties or posture of that 2011 W.D. Tex. case, so I do not treat it as a confirmed assertion by the owner.

6. Bankruptcy fire-sale — Not present. No Chapter 7/11 is recorded for Teltech or Sopheon, and no assignment recites a bankruptcy sale. Note for completeness: Sopheon divested its US information‑management business to FIND/SVP, Inc. in July 2003 (~$5M) — a going‑concern sale, not a bankruptcy estate sale, and the '051 patent appears to have remained with Sopheon (Google Patents still lists Sopheon Corp as assignee, and the 2002 20‑F claimed the patent as core IP).

7. Privateering — Not present. No transfer of the patent from Sopheon to an external NPE, and no evidence of Sopheon funding third‑party assertion against competitors.

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. It terminates at an operating software company (Sopheon → Wellspring).

---

Verdict

Operating-company assertion (qualified).

Justification: The entire recorded chain — Teltech Resource Network Corp → (2001‑03‑28 merger) → Sopheon Corporation, with only bank/insider security interests and releases in between (2000‑02‑01, 2000‑06‑30, 2004‑02‑23, 2007‑07‑02, 2014‑03‑12) — shows the patent never left an operating‑company family, and Sopheon's own SEC 20‑F described this patent as one of its most significant IP assets used in its search/portal products. Because the "assertion" element rests only on the single, unverified W.D. Tex. 1:11‑cv‑00332 litigation entry (and I could not retrieve reel/frame or correspondent data to test for hidden NPE links), I assign this verdict with a caveat rather than high confidence. No NPE signal is present on the record.

Verification links:

• USPTO Assignment Center (search "5634051"): https://assignment.uspto.gov/patent/index.html#/patent/search?q=5634051
• Assignment Center home: https://assignmentcenter.uspto.gov/
• Google Patents legal events: https://patents.google.com/patent/US5634051/en
• Family litigation (per Google Patents/Unified): https://portal.unifiedpatents.com/litigation/Texas%20Western%20District%20Court/case/1%3A11-cv-00332

Recommended next step to close the two gaps: pull the six reel/frame records from Assignment Center to capture the correspondent of record, and confirm the parties/posture of W.D. Tex. 1:11‑cv‑00332. Those are the only two facts that could move this verdict.

I'll research US5634051's prior-art citations. Let me start with targeted searches.

Let me dig into the citation record specifically.

Let me look for the specific prior-art citations on record for this patent.

Prior-Art Analysis: US 5,634,051 A — "Information management system"

1. Exact patent identification (no similar-number substitution)

Searching the literal number 5634051 resolves to:

⚠️ Do not conflate this with look-alike numbers that surfaced incidentally in searching: US 6,034,551 (a different, unrelated patent), US 5,634,051 vs. US 5,634,849, etc. All statements below are keyed to 5,634,051 only.

---

2. Critical finding on the record — please read before relying on any list

The authoritative full text of US 5,634,051 (Google Patents, fetched 2026-10-01) contains NO "References Cited" / "Patent Citations" section. It moves directly from the "Country Status" block to the "Cited By (233)" table. In other words, the record supplied for this patent discloses only forward citations (later documents citing 5,634,051) — not the backward citations (prior art cited during prosecution) that a § 102 analysis requires.

I attempted to retrieve the examiner/applicant citation list for 5,634,051 from additional sources (Google Patents, USPTO report mirrors, Espacenet/EPO citation data, and WO-family search reports) before my search budget was exhausted. I could not verify a citations list, so I will not fabricate one. Under the operating rule to say so explicitly when I do not know something with high confidence: I do not have a verified list of the patents cited against US 5,634,051, and the fetched record does not contain one.

This is not unusual for 1990s US grants: the backward-citation metadata is frequently not imported into third-party patent databases even when it exists in the USPTO file wrapper.

---

3. What the patent's own disclosure actually cites (verified)

The only prior-art reference named inside the specification text is non-patent literature:

Pritchard-Schoch, "Natural Language Comes of Age," Online, pages 33–43, May 1993 — cited as describing Westlaw's "WIN" natural-language query system.

• Citation: Pritchard-Schoch, "Natural Language Comes of Age," Online (magazine), May 1993, pp. 33–43.
• Publication date: May 1993 (before the 1993-10-28 priority date — § 102(a)/(b) eligible).
• Description: A survey article on natural-language search systems, including Westlaw's WIN, used by the patentee to characterize the known natural-language query-generation art.
• § 102 relevance: Low for anticipation. It is discussed in the specification as admitted background art for the query-generation aspect only. The claims (see § 5) are directed to document-type categorization and per-category result summaries, not to query generation. This reference cannot anticipate claims 1–22 as a whole, and the specification does not incorporate it by reference.

---

4. Direction-of-citation: where 5,634,051 itself was used as art (verified)

This is the reverse of the requested analysis, but it is what the record actually supports and it is probative for claim-scope interpretation:

1. WO 2001/015010 A3 — Search tool, system and method (Medical Data Services GmbH). International search report lists US 5,634,051 A (Thomson), 27 May 1997, as category "A" (background art, not of particular relevance) against claims 1–50. → Treated as general state of the art, not anticipatory.

2. WO 2016/112109 A1 — Citation table lists US 5,634,051 A (Teltech Resource Network), 27.05.1997 as category "A" (background art).

3. US 9,124,932 (Starsight Telecast / Rovi — "Systems and methods for contextually linking television program information") lists 5,634,051 | May 27, 1997 | Thomson among its references (per Justia).

Interpretation: Across independent examiners, 5,634,051 has repeatedly been classified as mere background art ("A"), never as anticipatory or inventive-step-defeating art ("X"/"Y"). That is a strong signal about the breadth and specificity of its claims.

The 233-entry "Cited By" list is likewise not prior art to 5,634,051. Every entry there (e.g., US 5,745,899; US 5,884,305-adjacent DEC filings; US 6,018,055; US 6,055,540; US 6,460,034; Oracle/Sun/IBM concept- and category-hierarchy patents) post-dates the 1993-10-28 priority date and therefore cannot be § 102 art against it. Using that list as "prior art" would be an analytical error.

---

5. What the claims actually require (for scoping any § 102 analysis)

Governing law for this file is pre-AIA 35 U.S.C. §§ 102(a)/(b)/(e) and 103 (priority 1993-10-28; never in AIA transition). The 22 claims are drafted in means-plus-function form (§ 112 ¶ 6), so any anticipation analysis must map prior art to the corresponding disclosed structure, not merely to the recited function.

The independent-claim cores are:

• Claim 1 (system): store large multi-source-record domain where a source record contains documents of multiple types → search "at least a substantial portion" with one query → categorize responsive documents by document type → generate a summary of the number of responsive documents per predetermined category.
• Claim 10 (system): same, but domain is "multiple document types" (no source-record limitation).
• Claim 14 (system): categorization "independently of the source record" from which documents came; summary of counts per document type.
• Claim 17 (method): store → generate query → search → sort and summarize by type of document, independently of source record.
• Claim 18 (method): define sets of categories, each category = one or more document types → search → select one set → sort by document type and summarize counts per category in the selected set.
• Claim 22 (system): multiple sets of categories, at least one category maps to more than one document type, plus count summary.
• Dependent customization claims: 3, 9, 12, 13, 16, 19, 20 (auto-select category set based on a user code or personal characteristic; claim 9 / claim 20 specifically recite "professional class").
• Dependent user-selection claims: 4 (user selects the set), 5–8 (predetermined/customizable sets and categories).

Notable drafting defect to account for: claim 20 recites "The method of claim 16…", but claim 16 is a system claim. This is an apparent dependency/antecedency error in the printed patent. Per the strict no-auto-correct instruction, I am reporting it as printed, not correcting it; it materially complicates any anticipation analysis of claim 20 and would likely be challenged.

The practical consequence: the narrowest, most defensible independent claims are 14, 17, 18 and 22; claim 1 is the broadest.

---

6. Honest bottom line

• I cannot give you "each patent citation for 5634051" with citation-level confidence, because the authoritative record I was given and my searches did not surface a verified "References Cited" list. Any enumeration I invented would violate the no-fabrication rule.
• The only reference actually named in the patent text is the May 1993 Online article by Pritchard-Schoch (non-patent literature), directed to query generation — weak § 102 art against these document-categorization claims.
• 5,634,051's own citations were classified as background ("A") art in later PCT search reports, indicating its claims were regarded as narrow/specific rather than broad.

To complete a defensible § 102 chart, obtain these primary sources:

1. USPTO PatentCenter / file wrapper for application 08/585,383 — the "References Cited" (PTO-892) and any applicant IDS are the definitive citation list.
2. WO 1995/012173 A2 international search report (family member) — the ISA's cited documents are usually the same art considered in the US case.
3. The USPTO assignment/prosecution history for 08/144,767 (parent) and 08/585,383 — abandoned-parent files often contain examiner citations absent from the grant.
4. Westlaw/Lexis litigation file for Teltech Resource Network Corp. v. (Texas Western District Court, case 1:11-cv-00332) — invalidity contentions filed there will name the specific prior art actually asserted against these claims.

If you can supply the PTO-892 or the WO search report text, I will produce the full per-reference table you asked for — full citation, publication/filing date, brief description, and the specific claim(s) each reference potentially anticipates under pre-AIA § 102 — with the means-plus-function mapping for each independent claim (1, 10, 14, 17, 18, 22).

I'll ground this analysis in the actual prior-art record for this patent, since the fetched page omitted the front-page "References Cited" section. Let me pull the prosecution-era prior art and the litigation record.

Good — I now have the Dialog OneSearch/duplicate-removal lineage. Let me pull the patent's own front-page references and the litigation/PTAB record, plus the document-type/categorization prior art.

Obviousness Analysis — US 5,634,051 (35 U.S.C. § 103)

0. Evidentiary basis — and a gap I have to flag before you rely on this

The fetched page does not actually contain a "References Cited" (front-page prior-art) list. What the page supplies under "prior art" is only:

I therefore could not verify the examiner-cited U.S. patent references from this fetch, and I will not fabricate them. Per the operating rules, if I lack high confidence I will say so. This memo is built on (a) prior art admitted in the patent's own specification, (b) documentary prior art verified by live search in this session, and (c) well-known art flagged with a confidence marker. A definitive § 103 opinion requires the file wrapper (USPTO PatentCenter for US 08/585,383) and the WO 1995012173 A2 search report.

Also note: only one entry in the 233-item "Cited By" list carries a pre-1993 priority date — US 6,101,512 A (Enigma Information Systems, priority 1991-07-19), directed to generating representations for and randomly rendering electronic documents. It is forward-cited, was published in 2000, and its disclosure is not on point for document-type categorization. It is not a useful § 103 reference here.

---

1. Governing law and the critical date

• Critical date: 1993-10-28 (priority; the 1996-01-11 filing is a continuation of abandoned Ser. No. 08/144,767). This is a pre-AIA application — the AIA first-inventor-to-file provisions apply only to applications filed on or after 2013-03-16. So the analysis is under pre-AIA § 102/§ 103(a).
• Legal framework: Graham v. John Deere (scope/content, differences, PHOSITA level, secondary considerations) + KSR Int'l v. Teleflex, 550 U.S. 398 (2007) — "a court must ask whether the improvement is more than the predictable use of prior art elements according to their established functions." MPEP 2143 supplies the exemplary rationales (A)–(G).
• PHOSITA: a person with a bachelor's degree in CS/IS (or equivalent) plus ~2–4 years' experience designing or operating commercial online information-retrieval systems (Dialog, BRS, Orbit, Westlaw/Lexis), including Boolean query engineering, inverted-file indexing, multi-file searching, and bibliographic/abstracting record structures. This was a mature, crowded art by 1993.

---

2. Element decomposition of the independent claims

Key observation: every claimed element is generic information-retrieval machinery. Nothing in the independent claims recites the thesaurus, natural-language front end, the inverted-file structure, the gateway engines, or the column/clone architecture — all of which are described in the specification as optional or preferable. The claims are therefore much narrower in inventive ambition than the disclosure, and the obviousness attack lands squarely on the reporting layer.

---

3. Verified prior-art landscape

---

4. Obviousness combinations, claim by claim

Combination A — OneSearch + RD + RANK(DT) → claims 1, 10, 17

Why a POSITA would combine them. All three commands were designed to be used together on the same platform, in the same session, against the same result sets — this is not hindsight glue, it is the manufacturer's documented workflow. The Dialog literature instructs users to "select databases with DIALINDEX, search with OneSearch, then use the duplicate-removal feature" (JST summary). RANK is simply the platform's general statistical-analysis tool; applying it to the DT= field to get a document-type histogram is rationale (C)/(D) under MPEP 2143 — use of a known technique (field-ranking) to improve a known device (a multi-file result set) in the same way, ready for improvement. The motivation is stated verbatim in the patent's own background: users were "faced with the daunting task of sifting through large unorganized answer sets."

Anticipation-style closeness: RANK on a document-type field yields the claimed "summary of the number of documents responsive to the query which fall within various predetermined categories of document types" almost verbatim. The only arguable gap is that RANK reports per field value rather than per predetermined bucket, which Combination B closes.

Combination B — DIALINDEX category groupings + RANK(DT) → claim 22 (many-to-one mapping)

Claim 22 requires only that one category in one set correspond to more than one document type (the spec's "Product Information" = product announcements + product reviews + product specifications). DIALINDEX's file categories already establish that the art routinely built predetermined higher-level buckets from many lower-level members. Aggregating several DT= values into one reported bucket is the "simple substitution / predictable variation" rationale (B)/(F). Note that claim 22 does not require the bucket to be user-specific — it requires only plural sets and one many-to-one category. This is the claim most exposed on the face of the prior art, and it is also the broadest independent claim in the patent (it drops any customization requirement).

Combination C — user-profile selection + DIALINDEX/RANK → claims 3, 9, 12, 13, 16, 19, 20

The "automatic customization by user code or personal characteristic" element is the patent's principal point of novelty, but it maps onto well-established 1980s practice:

• Login ID / user administration database → selecting a stored profile (the patent's own reference to "user administration relational database 22" and "fights [rights] found in the user administration file" confirms that per-user records driving system behavior were routine, not novel).
• Selective Dissemination of Information (SDI) profiles — long-standing in Dialog/BRS/Orbit (stored search profiles with per-user output formats and file lists) — establish that the art customized retrieval and output parameters by user identity.
• SET FILES, SET DETAIL, and user-defined cost/format preferences establish that the reporting layer was user-configurable.

Combining a user profile look-up with a per-field histogram is rationale (C) — applying a known personalization technique (profile-driven output) to a known reporting technique (field ranking) to achieve the predictable result of a user-tailored summary. The result — "the same answer set summarized differently for two users" — is exactly the predictable outcome a POSITA would expect from a switchable profile key. There is no teaching-away and no unexpected result.

Combination D — Z39.50 / broadcast search + local doc-type index + count-by-type → claims 14, 17, 18

For the source-independence limitation (the claim the patent itself frames as its advance over Dialog), the case is strong: a broadcast/OneSearch result set is by construction a merged result list in which the source file is merely a record attribute, not the organizing axis. Segmenting a merged set by a record attribute (DT=) rather than by the file of origin is the ordinary, obvious way to slice a merged set — indeed RD's own per-file count display is the only thing the prior art does by source, and moving from a source-keyed to an attribute-keyed histogram is a pure change of sort key. Change of a grouping key with no change in function is the paradigm of rationale (B).

Additionally, the specification's own description of the art is fatal to any attempt to claim the multi-source feature as the point of novelty: US 6,105,512 A aside, the specification concedes that "systems of this type are known in the industry including, e.g., Westlaw's 'WIN' system," and that "most systems used today" already let a user select multiple sources.

---

5. Consolidated rationale table (MPEP 2143)

No teaching away exists in the art; the platforms were marketed on more complete multi-file retrieval, which increases (not decreases) the incentive to summarize.

---

6. Where the patent pushes back (and how strong that is)

Residual nonobviousness argument the patentee could make: none of the located references expressly discloses automatically keying a predetermined document-type category set to a stored user code for the purpose of summarizing a cross-domain search result. That is the narrowest defensible redoubt — and it is a claim-differentiation argument, not an argument that the broad claims are valid.

---

7. Secondary considerations

Nothing in the record supports a nexus. Sopheon's Form 20-F describes the asset as "[Sopheon's] (formerly Teltech's) 'Presentation of search results from large domains of archived textual data by category or type' … for which a US patent was granted in 1997" (FY2002 20-F), and Sopheon's information-management business was divested to FIND/SVP effective 2003-07-01. But (i) Sopheon's own filing states it has "no other dependence on patents, licenses … which is material to its business"; (ii) commercial success of the service (expert network, custom research, portals) is not shown to be attributable to the claimed summarization feature; and (iii) the patent was never litigated to a merits validity finding — the only recorded case, 1:11-cv-00332 (W.D. Tex.), is a 2011 Docket-level entry I was unable to retrieve the pleadings for. There is thus no objective evidence to weigh against the § 103 case.

---

8. Collateral matters worth recording

1. Claim 20's dependency is broken. As flagged in the earlier section of this analysis, claim 20 reads "The method of claim 16 wherein…" — but claim 16 is a system claim depending from claim 14, and the "selecting one of the sets of categories" step appears only in method claim 18. Read literally (and I am not auto-correcting it), claim 20 is defective. That defect (a) weakens any validity argument resting on claim 20 and (b) independently exposes it to a § 112(b) challenge.
2. Expiration arithmetic may be misreported by the source. The page states "anticipated expiration 2013-10-28" (20 years from the 1993 priority date). For a pre-AIA patent filed before 1995-06-08, the term is 17 years from issue or 20 years from the earliest effective filing date, whichever is later — which would be 2014-05-27. I flag this as a possible inaccuracy in the source rather than correcting it silently. Practical significance: an expired patent cannot be asserted but remains challengeable by ex parte reexamination at any time (35 U.S.C. § 302), and remains citable as prior art against later filings.
3. The specification is the strongest evidence against the claims. It concedes the multi-source search, the natural-language/thesaurus front end (WIN), the Boolean/inverted-file engine (Fulcrum, InfoPro, Folio), and the source-organized baseline (Dialog) as existing. What remains — re-keying the result summary from source to document type and counting — is a reporting-layer change.
4. The many "means for" recitations are functional-only. For § 103 purposes this is a double-edged sword: it makes the claims easy to read onto prior-art systems, but it also triggers § 112(f), and under Williamson v. Citrix (Fed. Cir. 2015) functional claim language without corresponding disclosed algorithm is presumptively indefinite. This is a § 112 issue, not § 103, but it materially affects how much narrowing weight the patentee can extract from the claims.

---

9. Confidence and what remains to be verified

Recommended next steps to make this opinion definitive: (1) pull the file wrapper for US 08/585,383 from USPTO PatentCenter to obtain the actual cited references and any § 103 rejections/arguments of record; (2) obtain pre-1993 Dialog bluesheets for the DT field and pre-1993 Chronolog/database documentation of RANK; (3) review the WO 1995012173 A2 international search report; (4) confirm whether any validity challenge was actually adjudicated in the W.D. Tex. case.

Bottom line: Independent claims 1, 10, 14, 17, 18, and 22 are, on the evidence available, more likely than not obvious over Dialog OneSearch + duplicate removal + RANK/document-type field, alone or in combination with DIALINDEX-style category groupings — each element being a known prior-art element arranged for its established function, with the motivation expressly supplied by the art itself and conceded in the patent's own background. The only claim family with a non-trivial nonobviousness argument is the automatic user-code/professional-class selection family (3, 9, 12, 13, 16, 19, 20), and even there the argument is one of degree, not of kind — and it is undermined by claim 20's broken dependency and the routine per-user profile machinery described in the patent itself.