US 6526411

I'll verify the patent details against live sources before summarizing.

Let me make one more targeted check for any 2026 appellate activity.

US Patent 6,526,411 — Verification Summary

Note on the identifier: I interpreted the number literally as US 6,526,411 (the spaced/zero-padded form of "6526411"). All results below are for that exact grant. I found no CAFC 2026 docket naming this patent — see "Litigation / appellate status" below.

Bibliographic data

Related provisionals incorporated by reference: 60/165,726 and 60/165,727 (both filed 1999-11-15) and 60/166,039 (1999-11-17).

Google Patents also lists a large litigation tag set (Texas E.D., Texas W.D., Delaware, Colorado, N.Y.S.D., Illinois N.D., Texas N.D.) from the 2019–2021 assertion campaign. Source: https://patents.google.com/patent/US6526411/en

Chain of title (from the assignment record)

Individual (Ward) → Relatable, LLC (2003-03-05) → Rosenstrauss Applications V, LLC (2009-06-04) → Callahan Cellular L.L.C. (2016-01-14, by merger) → Intellectual Ventures Assets 120 LLC (2018-12-18, nunc pro tunc) → AVInnov LLC (2019-02-06) → Cedar Lane Technologies Inc. (2019-10-17). This matches the RPX account of the IV→AVInnov→Cedar Lane pipeline (https://litigation.rpxcorp.com/news/59211), which describes Relatable as a Virginia entity founded in 1999 by Sean Ward, and notes an IV affiliate acquired the '411 in May 2009.

Abstract (verbatim)

"Method and system provided for creating a dynamic playlist including meta-data having potential association with a respective content item configured to be played on a content player, and having dynamic addition of subtraction of playlist items. The system maintains a database of linkages between elements associated with content items as well as weighted linkages between elements and respective properties. The system is a hybrid content based and collaborative filtering system, wherein the insertion of a new item into the database results in the new item sharing preference weights and number of preferences associated with items pre-existing in the database. Thus, an initial input query list of items potentially results in the return of many content, called a 'dynamic playlist', has a high correlation with the user's preference or with whatever other basis was used to frame the input list, and individual content items on the dynamic playlist may not have been previously experienced by the user."

Plain-language overview of the four independent claims

The patent has 24 claims total; the independents are 1, 6, 12, and 22.

Claim 1 — System. A computer system for building a dynamic playlist with four cooperative components plus storage: (a) storage holding an elements table and a pairs table; (b) a component that accepts a "meta-category" — defined as a set of one or more criteria, each criterion capable of being associated with a content item (e.g., "this artist," "this genre"); (c) a component that queries one or more content providers and gets back a result set of metadata matching any criterion, where that metadata is enough to actually acquire the content items; (d) a component that inserts that result set into the dynamic playlist; and (e) a component that seeds the next meta-category with the result set. In short: a playlist is expressed as abstract tags/criteria rather than concrete tracks, and the system resolves each tag against live content sources and folds the results back in as a seed for the next tag.

Claim 6 — Basic method. The method counterpart of claim 1, without collaborative filtering: accept a meta-category, retrieve a matching metadata result set from one or more content providers, insert it into the dynamic playlist, then seed the next meta-category with that result set and loop through retrieve/insert/seed until all meta-categories are processed. This is the "basic mode" of FIG. 3 (steps S1–S7).

Claim 12 — Recommendation method. Same accept/retrieve/insert/seed-and-repeat loop as claim 6, but with a filtering step in between: the retrieved result set is passed through a "collaborative filtering query algorithm" to produce a filtered result set, which is what gets inserted. This is the "recommendation mode" of FIG. 4, where a sort server ranks/culls results at each step. Dependent claims under it add: user play-pattern data including manual intervention during playback (claim 13), explicit rating data indicating preference or distaste (claim 15), a first reordering algorithm chosen from ranking / random removal / keep-top-N / pairing sort (claims 16–17), a second reordering algorithm such as random reordering (claims 18–19), and the pairing-sort and hash-clustering sub-steps (claims 20–21).

Claim 22 — Recommendation method with source and sort limitations. The narrowest independent claim. It combines: accept meta-category → retrieve result set from a provider selected from the group consisting of a local music collection, a central content server, and a peer-to-peer computer system → apply a collaborative filtering query to get a filtered result set → apply a reordering algorithm selected from the group consisting of a ranking algorithm, a random element removal algorithm, a retention of top-N-most-popular-elements algorithm, and a pairing sort algorithm → insert into the dynamic playlist → seed the next meta-category with the result set and repeat until done. Dependent claims 23–24 add the pairing-sort procedure and the hash-clustering-of-profiles procedure.

Shared dependent-claim machinery (claims 20–21 and 23–24), plain language:

• Pairing sort: walk the playlist two items at a time, insert any missing item into an elements table, increment a weight between the current pair (incrementing an existing pair-link, or creating one of strength 1 if absent, and bumping the second item's link counter), then slide forward one position (item 2 becomes item 1) and repeat until the list is exhausted. This learns the order in which items tend to be played together.
• Profile hash clustering: pick a seed user profile, compare it against all profiles, rank by similarity, cluster the most similar ones with the seed, count element frequencies across the cluster, build a hash profile of the most frequent items as the cluster representative, store it in a hash table, remove the seed and its cluster from the pool, and repeat with the next unclustered profile.

Drafting/prosecution notes (useful caveats)

• Claims 2 and 3 are materially duplicative and internally inconsistent. Both recite a component "configured to reorder the dynamic playlist after insertion," but claim 2 calls it a "fifth component" while claim 3 — which also depends from claim 1, not claim 2 — calls it a "sixth component." Treat the numbering as a drafting artifact rather than a meaningful hierarchy.
• Claim 5 contains the typo "paring sort" (for "pairing sort"), reproduced literally in the granted text.
• The specification expressly benchmarks itself against, and states that "any of the known relational algorithms can be used," specifically naming Firefly, U.S. Pat. No. 5,749,081; Hey, U.S. Pat. Nos. 4,870,579 and 4,996,642; and Rose, U.S. Pat. No. 5,724,567.
• EPO search report on EP 1 818 930 A1 (Vodafone) cites US 6,526,411 B1 (Ward, 2003-02-25) as an "X" category document against claims 1 and 9 — a good indicator of its perceived scope as prior art in the playlist-generation space.

Litigation / appellate status — with explicit uncertainty

• CAFC 2026 dockets: I found no Federal Circuit 2026 docket or appeal naming US 6,526,411, and I found no 2026 district court filing asserting this patent. The 2025–2026 Cedar Lane cases surfaced in my searches concern different patents — e.g., 8,577,782 and 8,438,485. I cannot rule out a 2026 docket that is not indexed in the sources I can reach, so treat this as "not found," not as "nonexistent."
• Known assertion history: per RPX (Nov. 2019), the '411 was asserted by Cedar Lane together with 6,502,194, 6,721,489 and 7,173,177 against Charter Communications (E.D. Tex. 2:19-cv-00368), Frontier Communications, and Starz Entertainment / Lions Gate (D. Colo. 1:19-cv-03189) over cable boxes. The Google Patents litigation tag list is consistent with a broad 2019–2021 campaign.
• The patent is expired. Adjusted expiration 2021-08-02, status "Expired – Lifetime," so any 2026 dispute would concern past damages only.
• Note that the '177 patent from the same Cedar Lane bundle was invalidated in IPR (Unified Patents v. Cedar Lane; PTAB final written decision 2021-02-12, institution 2020-03-18) — relevant context if you are assessing the asserted family's strength, though that holding is not about the '411.

Confidence assessment

High confidence on all bibliographic fields, claims text, abstract, chain of title, and the expiry date — these come from the granted patent text, the Google Patents record, and the republished claim set at uspto.report (https://uspto.report/patent/grant/6526411). Low confidence / unverified: whether a 2026 CAFC appeal involving the '411 exists; my searches returned nothing, and I am flagging this as an information gap rather than asserting its absence. Likewise, "Current Assignee = Cedar Lane Technologies Inc." reflects the last recorded assignment (2019-10-17) and Google Patents' listing; I did not independently confirm a later transfer.

I'll research litigation involving US 6,526,411. Let me start with searches.

I have good initial results. Let me search for more details on the specific cases.

Litigation Involving U.S. Patent No. 6,526,411

Important framing note: I am interpreting the patent number strictly as 6,526,411 ("System and method for creating dynamic playlists," Sean Ward, filed 2000-11-15, granted 2003-02-25, now owned by Cedar Lane Technologies Inc.). All cases below tie to that number. I was unable to verify the full defendant/outcome details for every docket within my search budget, so I have flagged what is confirmed versus what comes only from the patent record. I have not filled gaps with guesses.

1. Source of the litigation data

The authoritative Google Patents record for US6526411 (fetched from https://patents.google.com/patent/US6526411/en) carries a "Family has litigation" flag and a "Critical" litigation entry, drawing on Unified Patents' litigation dataset. It lists the following dockets (case numbers and jurisdictions as reported; the record did not supply defendant names or outcomes):

The record also links a "First worldwide family litigation" entry (Darts-ip, family ID 27496659): https://patents.darts-ip.com/?family=27496659

2. Cases I could confirm beyond the bare docket list

Plaintiff in all confirmed matters: Cedar Lane Technologies Inc. (Canadian NPE, current assignee of the '411 patent since 2019-10-17, per the assignment chain: Ward → Relatable, LLC → Rosenstrauss Applications V, LLC → Callahan Cellular L.L.C. → Intellectual Ventures Assets 120 LLC → Avinnov LLC → Cedar Lane Technologies Inc.).

Sources: IPWatchdog Patent Filings Roundups (https://ipwatchdog.com/2020/04/29/... and https://ipwatchdog.com/2020/10/21/...); UniCourt case page for Cedar Lane Technologies Inc. v. Aspiro Inc. (https://unicourt.com/case/pc-db5-cedar-lane-technologies-inc-v-aspiro-inc-[440647](/patent/440647)).

3. What I could not confirm

• Defendant names for the remaining ~14 dockets (the 2:19-cv-00368/00369, 1:19-cv-00373, 1:19-cv-03189, 1:20-cv-00472/00473/00501/01370/03037/03635/06021, 6:20-cv-00988, 3:20-cv-01103, 1:21-cv-00676). The Google Patents/Unified Patents record lists only court + case number for these.
• Outcomes for any specific case. Cedar Lane's litigation is overwhelmingly resolved by settlement: third-party NPE trackers report Cedar Lane's overall portfolio resolved ~96.6% by settlement (with a small remainder by procedural disposition), and a KIPO/Korean report pegged Cedar Lane at roughly 367 suits through 2022 (310 settled, 13 otherwise terminated, 44 pending). However, I cannot attribute a specific outcome to any particular '411 docket without further docket-level verification.

4. Ownership / status notes relevant to litigation

• The '411 patent is Expired – Lifetime (adjusted expiration 2021-08-02), so any residual litigation would concern past damages only.
• Current assignee of record: Cedar Lane Technologies Inc. (Google Patents "Current Assignee").
• Cedar Lane is a well-documented NPE; independent reporting describes 457 total suits since inception, concentrated in E.D. Tex. and D. Del., with 2026 Q1 activity noted. (These figures are for Cedar Lane's whole portfolio, not the '411 patent specifically.)

5. Bottom line

Yes — U.S. 6,526,411 has been litigated, entirely by Cedar Lane Technologies Inc. as plaintiff, against numerous defendants in 2019–2021 across E.D. Tex., W.D. Tex., N.D. Tex., D. Del., S.D.N.Y., D. Colo., and N.D. Ill. I confirmed the defendant identity for only one docket robustly (Cedar Lane v. Aspiro, 1:20-cv-03159, S.D.N.Y.) and partially for two others (Disney, 6:20-cv-00949; DISH, 6:20-cv-00947). For the remainder, I can supply the court and case number from the patent's own litigation record but not verified defendant names or outcomes, and I am declining to invent them.

Recommended next steps for docket-level certainty: query Unified Patents' case list (https://portal.unifiedpatents.com/litigation/caselist) and RPX Insight for Cedar Lane Technologies (https://insight.rpxcorp.com/entity/[1742191](/patent/1742191)-cedar-lane-technologies-inc), and pull the individual PACER dockets for the case numbers in the table in Section 1.

I'll verify the structured "no PTAB proceedings" default against live sources before writing this up.

Let me run a few more targeted checks — specifically for any IPR naming the '411, and for the Unified Patents/Cedar Lane IPR pattern.

One more confirmation pass — checking litigation/IPR databases for any challenge to the '411 specifically.

PTAB / AIA Trial Proceedings — US 6,526,411 (Ward, "System and method for creating dynamic playlists")

Source of record: The structured "PTAB proceedings on file" block (USPTO Open Data Portal) returns no AIA trial proceedings for this patent. I searched independently for IPR/PGR/CBM activity naming the '411 and found nothing that contradicts that default. Details and the one near-miss below.

Proceedings overview

Total AIA trial proceedings on US 6,526,411: 0. There are no IPRs, no PGRs, and no CBMs on file — not active, not invalidated, not sustained, not settled, not denied. The patent's 24 claims, including independents 1, 6, 12, and 22, are entirely untested at the PTAB. For a defendant, the bottom line is the inverse of the usual NPE posture: there is no FWD to hide behind, and no claim you can point to as already canceled. But the patent is also expired as of 2021-08-02, and it is (as detailed below) from the same acquisition pipeline that produced a sibling patent Unified Patents successfully killed in IPR — so the defensive playbook exists, it just hasn't been run against this patent.

Flag on a likely trap: Anyone researching this patent will quickly hit Unified Patents, LLC v. Cedar Lane Technologies Inc., IPR2020-00006 — the case where the Board held the asserted independent claim of U.S. Patent 7,173,177 unpatentable (FWD 2021-02-12; instituted 2020-03-18). That is a different patent. The '177 patent (§101-ish "owned/unowned playlist item" UI claim) is not the '411 patent (dynamic playlist generation with collaborative filtering). The two were asserted together by Cedar Lane against the same defendants, which is why they surface in the same searches. Do not carry the '177 result over to the '411 — it has no preclusive effect on it.

Proceedings on file

None. There is no proceeding to profile. Per the constraints of this section, I will not manufacture proceeding numbers, panels, or dispositions.

For completeness, the following are the AIA-adjacent events I did find in the record, and why none of them is a proceeding against the '411:

No Federal Circuit appeal involving the '411 was found. No appeal is possible from a proceeding that does not exist.

Strategic summary

Claim status: everything is UNTESTED. No claim of US 6,526,411 has been canceled, narrowed, disclaimed, or confirmed by the PTAB or the Federal Circuit. Independent claims 1, 6, 12, and 22 and dependents 2–5 and 7–11, 13–21, 23–24 are all live in the sense that no adjudicative body has voided them. Because the patent expired 2021-08-02, "live" is a misnomer in practical terms — the claims cannot be infringed going forward, and any exposure is limited to past damages within the § 286 six-year lookback from the filing of the complaint that asserted them. That temporal reality, not any PTAB loss, is why Cedar Lane's '411 campaign has gone quiet.

Estoppel landscape: a blank slate — and that cuts both ways. Because no IPR was ever instituted against the '411, § 315(e)(2) estoppel is empty. No petitioner is barred from raising any ground. A defendant today (or a defendant defending a legacy past-damages claim) faces no procedural obstacles to an IPR-style § 102/§ 103 attack — subject, however, to the timing realities that would now drive a discretionary-denial fight: the patent has been expired since 2021, and the Office has recently been receptive to § 314(a) settled-expectations arguments arguing against spending PTAB resources on old, expired claims (see, e.g., the Director's discretionary-denial decision in IPR2025-00408 denying institution in part on the patent's years in force and resulting settled expectations). Practically, invalidity is better litigated in the district court for a past-damages-only claim, using the same art a petition would have used, without the institution risk.

Pattern signals. Cedar Lane was a serial, high-volume filer (RPX and KIPI data place it among the most active NPE plaintiffs of 2020–2022, several hundred suits), and it did not generally have to defend PTAB challenges — the defensive-aggregator community (Unified Patents) targeted only the '177 patent out of the playlist bundle. That is the tell: well-asserted patents eventually attract IPRs, and the '411 did not. The most likely explanations are (a) the defendants chose district-court invalidity, ex parte reexam, or settlement over IPR; (b) the claims were asserted late enough that discretionary-denial risk made IPR unattractive; and/or (c) the damages at stake per defendant were too small to justify the ~$500K cost of an IPR petition. None of those reasons implies the claims are strong. If anything, the expiration date suggests the campaign ran out of runway.

Recommended next steps

Since there is no PTAB activity, say so plainly — and read the absence as a signal. The absence of IPRs is not a sign the '411 has been "hardened." It has never been stress-tested. The '177 patent in the very same portfolio was tested and its asserted independent claim was held unpatentable — so the pipeline that produced the '411 is demonstrably vulnerable to well-funded prior-art attack.

For a defendant facing a legacy '411 demand:

1. Check the accrual window first. The patent expired 2021-08-02. Under 35 U.S.C. § 286, only damages from the six years preceding the complaint are recoverable, and the complaint must have been filed within the limitations period for the accused conduct. If the demand concerns post-2021 conduct, there is no infringement to enjoin or to charge. This alone may end the matter.
2. Assert the § 101 / Alice defense aggressively in court. Precedent from the § 101 fight over the other Cedar Lane-family playlist patent — Samsung's invalidity contentions in the '411-family / 9,462,411 litigation, citing BSG Tech, Electric Power Group, and Beteiro — shows the standard playbook: the claims are directed to the abstract idea of building a playlist, and the "elements table"/"pairs table" and "meta-category" limitations are generic, conventional technology. Note the '411 patent's own specification admits that its collaborative-filtering algorithms are "well-known and commonly used" and expressly incorporates Firefly (U.S. 5,749,081), Hey (U.S. 4,870,579 and 4,996,642), and Rose (U.S. 5,724,567) — an unusually generous built-in obviousness/double-counting record.
3. Mine the prosecution and EP file history. The EPO search report on EP 1 818 930 A1 (Vodafone) cited US 6,526,411 B1 as an "X" category document against claims 1 and 9 — i.e., a third party's examiner treated Ward's own disclosure as anticipating-type art. The same references and the EPO's reasoning are available as § 102/§ 103 ammunition.
4. If an IPR is nonetheless contemplated (e.g., a co-defendant wants a stay), be aware that the exhaustion date removes most of the usual urgency and invites a § 314(a) discretionary denial framed around settled expectations for a patent expired since 2021. Expect the Fintiv/General Plastic and settled-expectations briefing to dominate any institution decision, and budget for a real fight at the institution stage rather than the merits.
5. Link the record, don't cite memory. If you take the "sibling patent was invalidated" point to a court or to the Board, cite the actual documents: Unified's FWD announcement at https://www.unifiedpatents.com/insights/2021/2/12/cedar-lane-patent-held-unpatentable and the institution decision at https://www.unifiedpatents.com/insights/2020/3/24/cedar-lane-patent-determined-to-be-likely-unpatentable — and expressly state that they concern U.S. 7,173,177, not 6,526,411. Getting that distinction wrong is the fastest way to lose credibility with the panel.

Confidence assessment

• High confidence that no AIA trial proceeding against US 6,526,411 appears in the USPTO ODP structured data, and that my independent searches corroborate this. The only IPR in the Cedar Lane playlist portfolio is against the '177 patent.
• Medium-high confidence on the negative result being complete: ODP coverage of older, terminated, or never-instituted petitions is generally good, and I ran multiple targeted queries, but I cannot prove a negative for a petition that was filed and dismissed without leaving an indexed record. Treat this as "none found," not "provably none."
• I did not verify the FWD date/panel composition of IPR2020-00006 at the source document level — I relied on Unified's own public statements for the 2021-02-12 FWD date and the 2020-03-18 institution. If you need those for a filing, pull the papers from PTAB E2E for IPR2020-00006.
• No contradictions with the prior sections of this analysis were identified; the earlier section's account of the '177 IPR correctly labeled it as a different patent.

I'll research the USPTO assignment record for US 6,526,411, focusing on reel/frame numbers and correspondents of record.

Assignment Record & NPE Analysis — US 6,526,411

Sourcing caveat up front (read before using the table below): I was able to reconstruct the chain of title, dates, conveyance types, assignors and assignees from the Google Patents legal-events transcription of the USPTO assignment record and an RPX corroborating report. I was not able to open the USPTO Patent Assignment Search (assignmentcenter.uspto.gov / assignment.uspto.gov) UI directly in this session to read the reel/frame numbers or the correspondent-of-record fields, and my searches did not surface them in any secondary source. Those two fields are therefore marked not retrieved rather than guessed. The chain itself is well-documented and consistent across sources, so the NPE analysis below stands on the dates and assignee identities, not on the missing fields.

Verification links: USPTO Assignment Center · USPTO Assignment Search · Google Patents legal events · RPX, "In a Matter of Months, Former IV Portfolio Bounces Off One NPE to Land in New Cedar Lane Complaints"

Inventors

Pattern notes:

• There is a name-field discrepancy worth flagging to the file: the grant front page and Google Patents read "Sean Ward," while third-party databases attribute the provisionals to "Sean Michnel Ward." This is a middle-name variance, not two inventors, but it is the kind of inconsistency that complicates chain-of-title diligence if a later assignee's name is checked against the patent face.
• No within-12-months departure pattern can be assessed — there is only one inventor, and he is simultaneously the founder of the operating company and the original assignee of record. The "inventor exit" heuristic simply does not apply.

Original assignee

At issuance (2003-02-25): Sean Ward, as an individual. Both Google Patents ("Original Assignee: Individual") and the printed front page (which lists Ward's residential address rather than a corporate assignee) indicate the patent issued to the inventor personally, unassigned. The operating company that embodied the work, Relatable, LLC, received the patent by assignment just 8 days after issue (recorded 2003-03-05), i.e. as a corrective post-issue assignment to the true operating owner rather than as a sale.

• Relatable, LLC — primary line of business: music/media recommendation and dynamic-playlist technology; a Virginia entity founded in 1999 by Sean Ward (per RPX, which describes Relatable as a Virginia entity founded by Ward in 1999). The specification's own framing — building playlists from abstract "meta-categories," seeding a live query against content providers, and collaborative-filter-based recommendation — is consistent with a small software/recommendation startup rather than a hardware manufacturer.
• Did Relatable ship a product embodying the claims? The specification describes a "jukebox style" dynamic-playlist player and references real third-party components (Oracle™ RDBMS, RealServer™, Icecast™), i.e. the claims were drafted against a contemplated commercial product. Whether a shipping Relatable product reached market I cannot confirm from the sources retrieved. Treat "operating company with a contemplated product" as the finding; "shipped at scale" as unverified.
• Current status: Relatable, LLC no longer appears in the chain after 2009 — it exited by assigning the patent out to Rosenstrauss Applications V, LLC (recorded 2009-06-04). Its present corporate status (dissolved/acquired) is not determinable from the records I retrieved; the patent-relevant fact is that Relatable ceased to be the owner in 2009.

Assignment timeline

Chronological, from the Google Patents legal-events transcription (all are recorded dates; executed dates are given where RPX supplies them). Reel/frame and correspondent columns are not retrieved — see sourcing caveat.

• Recorded 2003-03-05 — Reel/frame not retrieved

  • Conveyance: Assignment ("Assignment of Interest")
  • Assignor: Sean Ward (individually)
  • Assignee: Relatable, LLC (Delaware?/Virginia — entity state not confirmed)
  • Correspondent: not retrieved
  • Context: Internal correction / founder-to-startup — the inventor-owner formalizes ownership into his own operating company 8 days after issue.

• Recorded 2009-06-04 — Reel/frame not retrieved

  • Conveyance: Assignment
  • Assignor: Relatable, LLC
  • Assignee: Rosenstrauss Applications V, LLC — an Intellectual Ventures-affiliated holding entity (the "Rosenstrauss Applications" family are IV acquisition vehicles)
  • Correspondent: not retrieved
  • Context: Transfer to a monetization aggregator — the operating startup sells the asset out to an IV acquisition vehicle. This is the first point the chain leaves operating-company hands.

• Recorded 2016-01-14 — Reel/frame not retrieved

  • Conveyance: Merger
  • Assignor: Rosenstrauss Applications V, LLC
  • Assignee: Callahan Cellular L.L.C. — another Intellectual Ventures entity
  • Correspondent: not retrieved
  • Context: Internal corporate reorganization within the IV family (merger of one IV holding vehicle into another). No third party introduced.

• Recorded 2018-12-18 — Reel/frame not retrieved

  • Conveyance: Nunc pro tunc assignment
  • Assignor: Callahan Cellular L.L.C.
  • Assignee: Intellectual Ventures Assets 120 LLC (note: AVInnov's later assignment names "Intellectual Ventures Assets 120/125 LLC")
  • Correspondent: not retrieved
  • Context: Intra-IV parsing of the portfolio in preparation for divestiture — "nunc pro tunc" backdating a transfer to make the record effective as of the intended date.

• Recorded 2019-02-06 — Reel/frame not retrieved

  • Conveyance: Assignment
  • Assignor: Intellectual Ventures Assets 120/125 LLC
  • Assignee: AVInnov LLC — the Delaware NPE that acquired a portfolio of "more than three dozen" patents from IV in December 2018 (RPX)
  • Correspondent: not retrieved
  • Context: Transfer-to-asserter — IV sells the portfolio into a purpose-built asserting vehicle.

• Executed 2019-10-07 / recorded 2019-10-17 — Reel/frame not retrieved

  • Conveyance: Assignment
  • Assignor: AVInnov LLC — signed by Steven R. Pederson as manager (RPX)
  • Assignee: Cedar Lane Technologies Inc. — a British Columbia company formed 2019-04-29, with attorney Greg Benoit as director (RPX); described by RPX as a Canadian NPE running multiple litigation campaigns on a growing body of former IV patents "now numbering close to 100"
  • Correspondent: not retrieved
  • Context: Transfer-to-asserter / pre-litigation positioning — AVInnov's entire portfolio is flipped to Cedar Lane, which reasserts the same patents roughly a month later.

No subsequent assignment is recorded; Google Patents shows "Current Assignee: Cedar Lane Technologies Inc." as of the last recorded transfer (2019-10-17).

Timeline diagram

timeline
    title Ownership of US 6526411
    1999 : Provisional applications filed
         : Sean Ward founds Relatable
    2000 : Non provisional application filed
    2003 : Patent issues to Ward as individual
         : Assigned to Relatable LLC
    2009 : Assigned to Rosenstrauss Applications V
    2016 : Callahan Cellular acquires by merger
    2018 : Assigned to Intellectual Ventures Assets 120
    2019 : Assigned to AVInnov LLC
         : Assigned to Cedar Lane Technologies
         : Cedar Lane sues Charter
    2021 : Patent expires

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT. The asset passes from an operating recommendation startup (Relatable, LLC) into a sequence of licensing-only vehicles: Rosenstrauss Applications V, LLC (rec. 2009-06-04), Callahan Cellular L.L.C. (merger, rec. 2016-01-14), Intellectual Ventures Assets 120 LLC (rec. 2018-12-18), AVInnov LLC (rec. 2019-02-06). The "Applications / Cellular / Assets / LLC" naming and the nature of the transfers (merger, nunc pro tunc, portfolio sale) are the concrete tells — no products in commerce, no operating activity. Caveat: the registered-agent-address and single-member-LLC particulars are "not retrieved"; the signal rests on the conveyance types and the IV/AVInnov/Cedar Lane identities, not on naming alone.

2. Known asserter in the chain — PRESENT (strong). The chain contains Intellectual Ventures (Rosenstrauss Applications V / Callahan Cellular / IV Assets 120), AVInnov LLC, and Cedar Lane Technologies Inc. — all entities that appear on RPX / Unified Patents asserter directories as high-frequency plaintiffs. Cedar Lane specifically is described by RPX as an NPE running multiple campaigns from ~100 former IV patents. Per the earlier section of this analysis, Cedar Lane asserted the '411 (with 6,502,194; 6,721,489; 7,173,177) against Charter Communications, Frontier Communications, and Starz/Lions Gate in 2019.

3. Repeat correspondent across the chain — INSUFFICIENT DATA / not assessable. I could not retrieve the correspondent-of-record field for any of the six recordings. This is the single most probative field for tying the anonymous LLCs together, and it is the one I cannot fill from the sources reachable here. Action item for verification: query the Assignment Center by patent number and read the "correspondent" column on each of the six records; recurring counsel across Rosenstrauss → Callahan → IV Assets 120 → AVInnov → Cedar Lane would convert this from "unclear" to "present."

4. Cascading transfers — PRESENT. Three consecutive transfers in roughly 10 months — IV Assets 120 (rec. 2018-12-18) → AVInnov (rec. 2019-02-06) → Cedar Lane (rec. 2019-10-17) — plus a 2016 IV-internal merger, and a 2018 "nunc pro tunc" clean-up. Rapid chain-through-LLCs within 24 months is exactly the pattern flagged.

5. Pre-litigation transfer — PRESENT. Cedar Lane's assignment was executed 2019-10-07 (recorded 2019-10-17) and, per RPX, Cedar Lane reasserted the same four patents against Charter about a month later. The 2019 Texas E.D. filings (e.g. 2:19-cv-00368 / 2:19-cv-00369) fall within the six-month window that this heuristic targets — the transfer was arranged immediately before assertion.

6. Bankruptcy fire-sale — NOT PRESENT. No bankruptcy of any assignor or assignee appears in the record or in the retrieved sources. Relatable's exit (2009) reads as a portfolio sale to IV, not a Chapter 7/11 disposition.

7. Privateering — UNCLEAR. The classic privateering fact pattern is operating company → NPE that asserts on the operating company's behalf against that company's competitors. Here the operating company (Relatable) sold out entirely to IV in 2009 and had no continuing economic stake surfaced; IV then monetized on its own account. There is no evidence Relatable directed or benefited from the later Cedar Lane assertions, so the "privateering-for-an-operating-patron" signal is not established (though the IV→AVInnov→Cedar Lane leg is a textbook monetization pipeline).

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. It terminates at Cedar Lane Technologies Inc., an affirmative asserting NPE. Inverse finding: the patent was not neutralized; it was weaponized.

Verdict

NPE — high confidence.

Four independent strong signals converge on the same record. (i) Shell-entity transfers: the asset passes Relatable, LLC → Rosenstrauss Applications V (rec. 2009-06-04) → Callahan Cellular (merger, rec. 2016-01-14) → IV Assets 120 (rec. 2018-12-18) → AVInnov (rec. 2019-02-06), i.e. out of operating hands into licensing vehicles. (ii) Known asserters in the chain: Intellectual Ventures, AVInnov, and Cedar Lane all appear on RPX/Unified asserter directories. (iii) Cascading transfers: three assignments in ~10 months across 2018–2019. (iv) Pre-litigation transfer: Cedar Lane's executed 2019-10-07 assignment was followed by reassertion against Charter roughly a month later. The only data gap is the correspondent-of-record / reel-frame field (signal 3), which I could not retrieve and which I flag as insufficient data rather than a negative finding.

Flagged inconsistency with the earlier-generated section: the prior section states RPX "notes an IV affiliate acquired the '411 in May 2009," whereas the Google Patents legal events show the Rosenstrauss Applications V assignment recorded 2009-06-04. An execution date in May 2009 with a June 2009 recording would reconcile the two, but the record I can see shows only the June recording date. Minor, but worth confirming against the reel.

Verification: re-run the search at USPTO Assignment Center or assignment.uspto.gov by patent number 6,526,411 and read the reel/frame and correspondent columns for each of the six records — that will close the one open field and, if the correspondent recurs across Rosenstrauss → Callahan → IV Assets 120 → AVInnov → Cedar Lane, will upgrade signal 3 to "present."

I'll research the USPTO record for US 6,526,411 and its cited prior art. Let me run several targeted searches.

I now have the examiner-cited prior art list from the patent's front page. Let me verify the identities and disclosures of the key references before analyzing them.

Prior Art Analysis — U.S. Patent No. 6,526,411 (Ward)

Note on scope and method. I retrieved the granted patent's own front-page "References Cited" list (i.e., the prior-art citations of record) and corroborated it against the Google Patents record, uspto.report, FreePatentsOnline, and the copy of the printed front page filed in Cedar Lane v. LiveXLive (D. Del. 1:20-cv-01370, D.I. 1-3). I did not obtain a live PatFT/PatentCenter text dump inside this session, and I was unable to pull the full text of every cited reference. Where I describe a reference's substance, I mark my confidence; where I could not verify it, I say so rather than guess. The citation data (numbers, inventors, grant dates) below is verbatim from the printed list.

One clarification that matters for this task: the patent's Google Patents page shows "Cited By (399)" — those are forward citations (later documents citing the '411), not prior art. The prior art is the "References Cited" block reproduced below.

1. The critical date (what "anticipates" means here)

• Priority: 1999-11-15 (provisional 60/165,726); filing: 2000-11-15.
• This is a pre-AIA patent. Prior art therefore comes from:
  • §102(b): U.S. patents printed, or publications dated, before 1999-11-15 (one year before the U.S. filing date); and
  • §102(a)/(e): patents/applications by another with an effective date before the applicant's invention date — for §102(e) the operative date is the reference's filing date, which is why several references granted after the '411 (e.g., 6,192,340; 6,356,971; 6,421,651) still qualify.
• Anticipation under §102 requires that a single reference disclose every limitation of a claim. Combinations of references go to §103 (obviousness), which is outside the strict §102 question you asked, though I note it where relevant.

2. Complete list of references cited of record (31 U.S. patents)

Asterisk () = flagged by the examiner with the footnote " cited by examiner" on the printed front page.

OCR/transcription variants to be aware of (do not treat as different patents): "Whites"/"Whiteis"; "Hoonan"/"Hooban"; "Rorberts"/"Roberts"; "Jkeunab" (unrecoverable); examiner name printed as "Meijahic" in one OCR copy vs. "Metjahic" (Safet Metjahic) in the record — this is a scan artifact, not a different examiner. (This also reconciles with the earlier-generated bibliographic section, which already lists Metjahic.)

3. §102 analysis — claim-by-claim, reference-by-reference

3a. Bottom line up front

On the face of the record, none of the examiner-cited references appears to anticipate independent claims 1, 6, 12, or 22 in their entireties. The reason is structural: every independent claim requires (i) a "meta-category" = a set of criteria, (ii) retrieval of a metadata result set from a content provider, and (iii) seeding the next meta-category with the result set and looping until all meta-categories are processed. The cited playlist/jukebox art teaches ordered queues but not tag/seeding recursion; the cited collaborative-filtering art teaches recommendation but not meta-category seeding or provider retrieval. That is consistent with the patent having issued over them.

Where a single cited reference does map onto limitations is at the dependent-claim level — principally the "elements table / pairs table + increment-a-pair-weight" feature (claims 1 storage element; 20; 23) and the collaborative-filtering-with-feedback features (claims 12, 13, 15).

3b. The strongest §102 candidate: Whiteis, US 5,749,081 (Firefly)

• Citation: US 5,749,081 A, Whiteis, "System and method for recommending items to a user," Firefly Network, Inc.; filed 1995-04-06; granted 1998-05-05. (§102(b) art — issued more than a year before the '411 filing.)
• Disclosure (verified): input list of favored items → generate all pairs → store each pair in a LINKS table with a weight = number of times the pair co-occurred (in Increment step 107 an existing link's weight is incremented by 1; a new link gets weight 1) → maintain a LINKSTO table storing, per item, the sum of link weights ("LinksTo Weight") → produce a ranked RESULTS table.
• Mapping to the '411:
  • Claim 1's "elements table" and "pairs table" limitation, and claim 20/23's "determining if both elements are in an elements table … incrementing a weight between the first element and the second element," are both strongly suggestive of Whiteis's LINKSTO/LINKS tables and weight-increment routine. This is the single most plausible §102 challenge point in the record — aimed at claims 20 and 23, and (for the storage component only) claim 1.
  • Why it likely still fails on the independents: Whiteis operates on a co-occurrence list of favorites, not a playback playlist, and discloses no meta-category, no content-provider retrieval of a metadata result set, and no seeding of a next meta-category. Those are claim-1/6/12/22 limitations absent from Whiteis.
  • Note the specification itself concedes Firefly as a known "relational algorithm" (see §4 below), which weakens any argument that the '411's filtering concept is novel relative to Whiteis.

3c. Secondary §102 candidates (collaborative filtering / feedback)

• US 5,867,799 (Lang et al.) and US 5,983,214 (Lang et al.) — integrated content-based + collaborative filtering with adaptive profiles and user feedback updating the profiles. Relevant to claim 12 (collaborative filtering query), claim 13 (play-pattern / user intervention), and claim 15 (rating data). However, they are document/informon filters, not playlists: no elements/pairs table, no meta-category, no content-provider retrieval of playable items, no seed-and-repeat over categories. No anticipation of an independent claim; useful §103 art against claims 13/15.
• US 6,041,311 / 6,049,777 / 6,092,049 (Chislenko & Sheena, Microsoft; all filed ~1995-06-30) — the Microsoft automated-collaborative-filtering family for recommending items using ratings. Relevant to claim 12 and, via rating inputs, claims 15. Same structural gap: no meta-category, no playlist seeding, no elements/pairs table. §103 material rather than §102 anticipation.
• US 5,918,223 (Blum et al.) — content-based audio analysis/retrieval. Potentially relevant to "meta-data having association with a content item" and to content-based (as opposed to collaborative) filtering, but not to playlist construction or claim 6/12/22's provider/seed loop. Low anticipation value.

3d. Playback/queue art

• US 6,118,450 (Proehl et al.), US 5,341,350 (Frank), US 5,633,839 (Alexander), US 5,872,747 (Johnson), etc. — playlist/queue and playback-control art. These bear on the "playlist" environment but, as cited, do not appear to disclose dynamic meta-category expansion via a content-provider query or the seed-the-next-category loop that distinguishes the independents.
• The Roberts-family patents (5,987,525; 6,154,773; 6,230,192; 6,230,207; 6,240,459) and 6,192,340 (Abecassis) — networked/streamed media delivery and playback control. Background art for the FIG. 5 "jukebox" embodiment and for the "content provider system 120"; I did not verify text sufficient to assert §102 anticipation of any specific claim.

3e. Claims with no apparent §102 reference in the record

• Claims 21 and 24 (hash-clustering of user profiles: rank, cluster, count element frequency, build a hash profile, place in a hash table, remove clustered profiles, iterate). I found no cited reference that appears to disclose this specific sequence. The Lang references disclose community/interest grouping and profile adaptation, but not building a hash profile per cluster. My working view is that claims 21/24 were not anticipated by anything in the cited list.
• Claims 2–5, 7–11, 14, 16–19 (reordering/culling, provider groups "local collection / central server / peer-to-peer," random reorder): the group-recitation "peer-to-peer computer system" (claims 11, 14, 22) is notable because the file-sharing/peer-to-peer prior art of 1999–2000 is essentially absent from the cited list — a real gap if you were building an invalidity case.

4. Admitted prior art that is not in the examiner's cited list (flagged as a gap)

The '411 specification expressly benchmarks itself against, and states "any of the known relational algorithms can be used … such as, for example":

• Firefly — U.S. Pat. No. 5,749,081 (this one is in the cited list — see §3b);
• Hey — U.S. Pat. Nos. 4,870,579 and 4,996,642 (not in the front-page cited list); and
• Rose — U.S. Pat. No. 5,724,567 (not in the front-page cited list).

These are admissions of prior art in the specification, so they are prior art of record even though they were not printed in the "References Cited" block. The Hey patents (ratings-based recommendation) and Rose (5,724,567) are §102(b) art (all issued well before 1999-11-15) and should be part of any complete §102/§103 analysis; their absence from the printed list is a notable completeness gap in the examination record. (Flagging this explicitly: it does not contradict the earlier-generated summary, which correctly noted the specification names Firefly/Hey/Rose.)

Also worth recording (not prior art, for completeness): US 6,748,395 (Microsoft, "System and method for dynamic playlist of media," filed 2000-07-14) appears in the forward-citation ("Cited By") list. Because its filing is after the '411's 1999-11-15 priority, it is not prior art to the '411.

5. Confidence and caveats

• High confidence: the identity, inventor, and grant date of every reference in the table (taken verbatim from the printed front page), the five examiner-asterisked entries, and the substance of Whiteis (5,749,081) and Lang (5,867,799; 5,983,214).
• Medium confidence: the subject-matter descriptions for the Microsoft collaborative-filtering family (6,041,311 / 6,049,777 / 6,092,049, sharing a ~1995-06-30 filing date per a citing document's reference list), Blum 5,918,223, the Roberts/RealNetworks family, Abecassis 6,192,340, and Proehl 6,118,450.
• Low / unverified: references #1–#6, #8, #10, #11, #13, #14, #16, #18, #22, #30, #31. In particular US 5,959,945 — both sources I retrieved give the inventor as a garbled token ("Jkeunab"), and I could not confirm its title or subject matter. US 6,421,651 (Tedesco) and US 6,356,971 (Katz) carry class codes in the scanned copy that look corrupted ("705/5", "710/301"), so their apparent relevance may be a scan artifact.
• No fabrication: I have deliberately not assigned titles or disclosures to the low-confidence entries. A definitive §102 opinion would require pulling the full text (and file history) of each reference — I could not do that for all 31 within this session, and I am flagging it rather than filling the gaps.
• Contradiction check with earlier sections: none material. The earlier bibliographic section listed the primary examiner as Safet Metjahic; one scanned copy renders it Meijahic. This is an OCR artifact, not a conflict.

Recommended next step for docket-grade certainty: retrieve each reference's full text from USPTO PatentCenter/PatFT (or Espacenet) and, for the Whiteis and Lang references specifically, compare their disclosure line-by-line against claims 1, 6, 12, 20, 22, and 23.

I'll ground the obviousness analysis in the actual prior-art references before drafting. Let me verify the key references cited in the patent and the pre-priority-date candidates.

Good — the key references are confirmed. Let me run a final targeted batch on the remaining structural features and the EPO "X" citation.

Obviousness Analysis — U.S. Patent 6,526,411 under 35 U.S.C. § 103

Two housekeeping flags before the analysis:

• Date conflict. The task header states "Current Date: April 26, 2026," while the patent page I was given was fetched 2026-09-28 and the system date is 2026-09-28. I proceed on the 2026 dates; nothing in the analysis turns on this.
• Source discipline. The "Cited By (399)" table on the Google Patents record is not prior art. Nearly every entry in it (Microsoft's US 6,748,395, Gracenote's US 2003/0135513, Platt's US 2003/0221541, Apple's US 2004/0224638, etc.) post-dates the '411 and in several cases was blocked by it. The only "Prior Art" the page itself supplies is the family the specification names: Firefly (US 5,749,081), Hey (US 4,870,579 and US 4,996,642), and Rose (US 5,724,567). I build on that list and supplement with pre-1999-11-15 art located in search.

1. Framework and threshold determinations

Statute. The application was filed 2000-11-15 with a claimed provisional priority of 1999-11-15 (Prov. 60/165,726; Prov. 60/165,727 also filed 1999-11-15; Prov. 60/166,039 filed 1999-11-17). It is therefore governed by pre-AIA § 103(a). The analysis follows Graham v. John Deere Co., 383 U.S. 1 (1966), as refined by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007).

Critical date. Assuming the provisional supports the claims (the specification's own cross-reference paragraph is garbled — it recites "the present inventor's application Ser. No. 09/712,261 and the No. 60/165,727 Provisional Application) filed concurrently herewith," which is a drafting artifact — but the incorporated provisionals are of record), the effective date is 1999-11-15.

PHOSITA. A person with a bachelor's degree in computer science or electrical engineering and about two years of experience in networked media systems / information filtering, or equivalent practical experience. This is consistent with the stipulation the parties reached for the analogous patents in I/P Engine, Inc. v. AOL Inc. (J.A. 39, quoted in the Fed. Cir. opinion) — "an individual with a bachelor's degree in computer science with at least two years of experience."

Binding specification admissions. The following statements in the '411 are prior-art admissions that bind the patentee for § 103 purposes (Constant v. Advanced Micro-Devices, 848 F.2d 1560, 1570 (Fed. Cir. 1988); PharmaStem Therapeutics v. ViaCell, 491 F.3d 1342, 1362 (Fed. Cir. 2007)):

This is precisely the posture the Federal Circuit condemned in I/P Engine v. AOL, 576 F. App'x 982 (Fed. Cir. 2014) — where the patentee's own specification called content-based and collaborative filtering "conventional," and the court held the combination obvious as a matter of law. The '411 is in the same trap: it admits the filtering engines and the shuffle, and claims only the chaining of them into a playlist.

2. The prior-art record

Prior-art status of Boulter (2003/0046399). Its publication date (2003-03-06) is after the '411's filing, so it is not a § 102(a)/(b) printed publication. Its operative date is therefore § 102(e), and under In re Giacomini, 612 F.3d 1380 (Fed. Cir. 2010), a U.S. patent/publication gets § 102(e) effect as of its provisional filing date — here 1999-11-10, five days before the '411's 1999-11-15 priority. The margin is thin; a § 103 combination should not rest on Boulter alone. Its utility filing (09/709,234, 2000-11-09) independently predates the '411's 2000-11-15 filing and supplies a fallback.

3. Independent Claim 1 (system)

Conclusion — Claim 1 is obvious. Every element is disclosed, and the only arguably novel twist (feeding a prior result set forward as the seed for the next criterion) is the everyday relevance-feedback loop that a PHOSITA would apply for a predictable benefit (better-targeted subsequent retrieval). KSR, 550 U.S. at 416–17.

4. Independent Claim 6 (basic method)

Claim 6 is claim 1 minus any filtering: accept → retrieve → insert → seed-and-repeat. Motorola '127 alone discloses metadata-attribute-driven selection and playback from a server, and Boulter '399 discloses the same loop over a content library keyed on an artist/genre criterion. Adding "retrieve results → use as seed for the next criterion" is a simple substitution of a known iterative technique (MPEP 2143(III)) with a predictable result. Obvious.

5. Independent Claim 12 (recommendation method) — the cleanest § 103 case

Claim 12 = claim 6 + "calculating a filtered result set by application of a collaborative filtering query algorithm."

Primary combination: Hey '642 (or '579, or Firefly '081) + Rose '567 + Motorola '127 / Boulter '399.

• The collaborative filter. Hey '642 discloses recommending "items sampled by one or more of the users in the group but not sampled by the selected user," which is verbatim the '411's stated purpose ("providing to the user recommendations of items which have not yet been sampled by the user"). Firefly '081 discloses the same. Both are admitted in the '411 specification.
• Hybrid content + collaborative. Rose '567 explicitly teaches combining content-based and correlation-based prediction, and the '411's own Summary calls its system "a hybrid content based and collaborative filtering system." The Federal Circuit has held this exact combination (content-based filtering + collaborative filtering applied to a query/result list) obvious as a matter of law, relying on the patentee's admissions. I/P Engine v. AOL, 576 F. App'x 982, 988–92 (Fed. Cir. 2014) ("These references, individually and collectively, teach the clear advantages of combining content-based and collaborative filtering.").
• The playlist/retrieval scaffold. Motorola '127 (server + player + profile + attribute filtering) or Boulter '399 (artist/genre criterion → community preference → content selection → re-bias).
• Motivation. (i) Improving recommendation quality is the express purpose of both Hey and Firefly; (ii) the '411 itself admits the filters are interchangeable building blocks and states "a system is needed that is easy to use, adapts to personal tastes, and can easily add or subtract music or videos"; (iii) a PHOSITA would expect predictable improvement by inserting a known ranking layer between retrieval and playback. No reference teaches away — Firefly's disparagement of Hey is an efficiency critique (storage/processing), not a teaching that their combination is inoperative.

Claim 12 is, in my assessment, clearly obvious.

6. Independent Claim 22 (narrowest)

Adds two Markush groups:

1. Provider group — "a local music collection, a central content server, and a peer-to-peer computer system." Motorola '127 discloses server → player content delivery; the '411's own FIG. 5 describes "local content 22 and available streamable content 22 from content provider systems 120" (its own admission that local + central were conventional). Peer-to-peer distribution of media was well known and commercially deployed by 1999 (Napster launched in 1999). Selecting among these is obvious to try and a mere recitation of known alternatives.
2. Reordering group — "a ranking algorithm, a random element removal algorithm, a retention of top N most popular elements algorithm, and a pairing sort algorithm." Firefly '081 discloses ranking by link weight and the pair-link incrementing scheme; Hey '642 discloses keeping "the most highly recommended item[s]" (top-N retention) and culling the rest; random shuffle is an admitted prior art ("consisting mainly of randomizing the order"). The '411 calls these "common playlist manipulation algorithms." Reciting a closed list of admitted techniques does not impart patentability.

Obvious.

7. Dependent claims

8. Motivation to combine — the KSR rationales

1. Same field / familiar elements. Every reference is in networked personalized recommendation or media playback; all are analogous art.
2. Predictable result. The '411's independent claims recite no result beyond "the retrieved content has a high correlation with the user's preference" — the stated goal of Hey, Firefly, Rose, and Boulter.
3. Market and design pressure. The '411 admits the need ("A system is needed that is easy to use, adapts to personal tastes…"). The Amazon interrogatory exhibit of record describes the pre-1999 commercial landscape: "Net Perceptions, Firefly, or LikeMinds… commercializing such systems for web e-commerce, and pervasive statements in the literature."
4. Specification admissions of interchangeability. When a patentee says "any of the known relational algorithms can be used… the Firefly system… the Hey systems… or… the Rose system," he has supplied the combination rationale himself.
5. No teaching away; no unexpected results. Firefly's criticism of Hey is a performance critique, not a teaching that combining CF with a playlist is inoperative. Nothing in the record establishes criticality of the "elements table + pairs table" pairing or of the recursion.
6. Obvious to try / finite alternatives. The reordering and provider Markush groups are precisely the "finite number of identified, predictable solutions" KSR treats as obvious.

9. Where the § 103 case is weakest (candid assessment)

• The "seed the next meta-category with the result set" recursion (all independents). No single reference I located teaches chaining criterion-to-criterion with the prior result set as the seed. The closest analogue is IR relevance feedback / query expansion (Rocchio 1971; Salton & McGill 1983, incorporated into Rose '567 by reference). The combination is still strong, but it requires an explicit motivation statement; it is not a one-reference read.
• "Hash profile… placed in a hash table" (claims 21, 24). The clustering steps are well supported (Hey, Bergh '186, Xerox '797), but the specific hash-profile-as-cluster-representative artifact may not appear in the located references. The step is defensible as a routine data-structure choice, but a challenger would want a primary reference teaching hashing of a per-cluster preference profile.
• Boulter's 5-day priority margin. If the '411's provisional is held not to support the claims (in which case the effective date moves to 2000-11-15), Boulter's § 102(e) date remains 2000-11-09 — still prior art, but the margin shrinks and the Giacomini analysis matters more.
• I was unable to retrieve the EPO search report text for EP 1 818 930 A1 (Vodafone), which the earlier section reports as citing US 6,526,411 B1 as an "X" document against claims 1 and 9. My search budget ran out. That citation, if confirmed, is significant corroboration of perceived breadth but I am relying on the earlier section and flagging it as unverified here.
• Mootness note. The patent is Expired – Lifetime (adjusted expiration 2021-08-02), so any § 103 challenge is relevant only to past damages or to a reexamination-style record; the claim set cannot be prospectively asserted.

10. Bottom line

Overall confidence: high that independent claims 6, 12, and 22 are invalid under pre-AIA § 103(a), and that dependent claims 7–11, 13–20, and 23 fall with them. Moderate on claims 1 and 2–5 (they require an explicit motivation for the element/pairs-table + seeding combination, which the specification's own admissions largely supply), and moderate on 21/24 pending a reference that squarely discloses the hash-profile cluster representative.

Key counter-consideration the patentee would press: the "hybrid content-based + collaborative" framing is not itself novel — the '411 admits as much, and I/P Engine v. AOL holds that admission dispositive. The patentee's best remaining argument is narrow: that no reference teaches the specific data structures (elements table + pairs table) coupled to the meta-category seeding loop. Because Firefly '081 supplies essentially those data structures (LINKS table with incrementing pair weights) and the seeding loop is textbook relevance feedback, I regard that argument as unlikely to succeed.

Caveat on scope: This is a technical invalidity assessment, not a legal opinion, and it assumes the provisional applications support the claims as issued. If they do not, the critical date moves to 2000-11-15 and the art set expands (e.g., Music Choice's 2000-03-08 filings, Logan/Gotuit's 2000-03-28 filings, and the Microsoft dynamic-playlist work would come into play) — which would only strengthen the obviousness case.