US 7209876

Here's a concise summary of US Patent 7209876:

US Patent 7209876: System and method for automated answering of natural language questions and queries

• Title: System and method for automated answering of natural language questions and queries
• Assignee: Iot Innovations LLC (Current Assignee as of August 2, 2022). The original assignee was Groove Unlimited LLC.
• Inventors: Michael J. Miller, Leonard Jay Wantz
• Filing Date: November 13, 2002
• Issue Date: April 24, 2007
• Abstract: The patent describes a method for searching an information repository to find answers to natural language questions or queries. This method involves generating an expected answer form based on the structure of the natural language question, searching for answers that fit this expected form, and then presenting the answer to the user.

Legal Status: The patent is listed as "Expired - Lifetime," with an expiration date of January 24, 2025. Therefore, as of April 26, 2026, the patent has expired. There is no specific mention of US Patent 7209876 in the CAFC dockets for 2026 within the search results, which is consistent with its expired status.

Plain-Language Overview of Independent Claims:

• Claim 1 (Method for Producing Answers): This claim describes a computer-implemented method that takes a search question or query containing placeholders (variable expressions) for the desired answer. It then performs an initial search in an information repository, narrows down the results to "answer phrases" that structurally resemble the query and fill the placeholders, and finally modifies these phrases to provide direct answers.
• Claim 13 (Method for Producing Answers - Keyword/Phrase Search Focus): This claim outlines a computer-implemented method where a search question or query indicates the subject. It initiates a keyword/phrase search in an information repository, obtains a dataset, and then retrieves "answer phrases" from this dataset that syntactically match the anticipated form of an answer to the query. These answer phrases are then provided as output.
• Claim 25 (Method for Processing Natural Language): This claim details a computer-implemented method for processing a natural language question. It starts by converting the natural language question into a "generic question form," then into an "expected answer form." It then uses keywords from the question to search an information repository, filters the results to find "answer phrases" that syntactically match the "expected answer form," and formats these to be direct answers to the original natural language question.
• Claim 32 (System for Producing Answers): This claim describes a computer system, specifically a server, hosting a data search service. This system is equipped with program code to:
  • Transform a natural language question into an "expected answer form" that includes specific data terms and answer placeholders.
  • Extract these data terms to generate a keyword/phrase search query and initiate a search in an information repository.
  • Filter the search results to pinpoint "answer phrases" that syntactically match the "expected answer form."
  • Format these "answer phrases" into direct answers for the original natural language question and present them as output.

As of April 26, 2026, US Patent 7209876 is listed as "Expired - Lifetime" with an expiration date of January 24, 2025. This means the patent has already expired.

Despite its expired status, there is known litigation involving US Patent 7209876. Based on the provided information from Unified Patents, the following cases are associated with this patent:

• Jurisdiction: Ohio Northern District Court
  • Case Number: 1:24-cv-01403
  • Status: Litigation
• Jurisdiction: Texas Eastern District Court
  • Case Number: 2:25-cv-00239
  • Status: Litigation
  • Case Number: 2:26-cv-00334
  • Status: Litigation
  • Case Number: 2:25-cv-01135
  • Status: Litigation
  • Case Number: 2:25-cv-01009
  • Status: Litigation
• Jurisdiction: Massachusetts District Court
  • Case Number: 1:24-cv-12631
  • Status: Litigation
  • Case Number: 1:23-cv-12528
  • Status: Litigation
  • Case Number: 1:23-cv-10352
  • Status: Litigation
• Jurisdiction: Georgia Northern District Court
  • Case Number: 1:23-cv-02475
  • Status: Litigation

Unified Patents also indicates "First worldwide family litigation filed" for US7209876(B2), with a link to Darts-ip. However, specific plaintiff(s), defendant(s), and filing dates for this "worldwide family litigation" are not detailed in the provided snippets. For the listed US District Court cases, the status is generally "litigation," and specific plaintiffs and defendants are not explicitly named in these summary entries.

It is important to note that while the Google Patents entry shows these litigation cases, specific outcomes are not provided in the directly available snippets. For detailed outcomes or current status, a deeper dive into each case through PACER or the respective court dockets would be necessary. PACER allows searching for cases by specific court or a nationwide index, with real-time updates for specific court filings and daily updates for the nationwide index.

Proceedings overview

As of today, June 1, 2026, there are no AIA trial proceedings (Inter Partes Review, Post-Grant Review, or Covered Business Method reviews) found on file for US Patent 7209876, according to both the USPTO Open Data Portal API and a comprehensive web search. This indicates that the patent's claims have not been challenged or adjudicated at the Patent Trial and Appeal Board (PTAB). For a defendant, this means the patent's claims are entirely untested in the PTAB forum.

Strategic summary

Since no PTAB proceedings have been identified for US Patent 7209876, all its claims (Claims 1-32) are considered untested in the context of AIA trials. There are no claims that have been canceled or sustained by the PTAB. Consequently, there is no estoppel landscape established under 35 U.S.C. § 315(e)(2), which would typically bar petitioners and their privies from raising grounds that were or reasonably could have been raised in an IPR. All potential prior-art grounds remain available for a defendant facing assertion of this patent. The absence of PTAB activity suggests that either the patent has not been extensively asserted in litigation that would provoke such challenges, or any assertions have not yet led to public PTAB filings. The mention of several district court litigations for this patent (from the initial patent summary) suggests that while it has been asserted, these assertions have not, as of yet, resulted in public PTAB challenges.

Recommended next steps

Given the absence of PTAB activity for US Patent 7209876, a defendant facing assertion of this patent has several strategic options. There are no existing PTAB Final Written Decisions to leverage. Therefore, a defendant could consider initiating an Inter Partes Review (IPR) if they identify strong prior art that was not considered by the USPTO during prosecution. Such a move would be based on a fresh analysis of the prior art landscape against the patent's claims. The fact that the patent expired on January 24, 2025, means that new IPRs might still be viable if the alleged infringement occurred before expiration or if there are ongoing damages claims. The absence of prior PTAB challenges means a petitioner would not be estopped by previous proceedings.

Inventors

• Michael J. Miller (Employer at time of filing: Not determinable from patent text)
• Leonard Jay Wantz (Employer at time of filing: Not determinable from patent text)

No unusual patterns, such as all inventors departing the original assignee within 12 months of filing, are discernible from the provided patent text.

Original assignee

The original assignee was Groove Unlimited LLC. Based on the provided patent information, it is not determinable if Groove Unlimited LLC shipped a product embodying the claims or their primary line of business. Their current status (operating, acquired, dissolved, in bankruptcy) is not stated in the patent text.

Assignment timeline

• 2007-03-09 (executed) / recorded 2007-03-12 — Reel 018992/0201

  • Conveyance: Assignment
  • Assignor: MILLER, MICHAEL J., WANTZ, LEONARD JAY
  • Assignee: GROOVE UNLIMITED, LLC
  • Correspondent: Not specified in this entry.
  • Context: Transfer from inventors to original assignee.

• 2008-08-27 (executed) / recorded 2008-10-08 — Reel 021640/0691

  • Conveyance: Assignment
  • Assignor: GROOVE UNLIMITED, LLC
  • Assignee: DATA SPEED II, LIMITED LIABILITY COMPANY
  • Correspondent: Not specified in this entry.
  • Context: Transfer from original assignee.

• 2015-08-26 (executed) / recorded 2021-01-26 — Reel 037584/0951

  • Conveyance: Merger
  • Assignor: DATA SPEED II, LIMITED LIABILITY COMPANY
  • Assignee: OL SECURITY LIMITED LIABILITY COMPANY
  • Correspondent: Not specified in this entry.
  • Context: Transfer of ownership via merger.

• 2022-07-21 (executed) / recorded 2022-07-21 — Reel 060585/0235

  • Conveyance: Assignment
  • Assignor: OL SECURITY LIMITED LIABILITY COMPANY
  • Assignee: INTELLECTUAL VENTURES ASSETS 184 LLC
  • Correspondent: Not specified in this entry.
  • Context: Transfer to Intellectual Ventures entity.

• 2022-07-28 (executed) / recorded 2022-08-02 — Reel 060698/0604

  • Conveyance: Assignment
  • Assignor: INTELLECTUAL VENTURES ASSETS 184 LLC
  • Assignee: IOT INNOVATIONS LLC
  • Correspondent: Not specified in this entry.
  • Context: Transfer from Intellectual Ventures entity to current assignee.

Timeline diagram

timeline
    title Ownership of US 7209876
    2007 : Assigned to Groove Unlimited LLC
    2008 : Assigned to Data Speed II LLC
    2015 : Merged to OL Security LLC
    2022 : Assigned to Intellectual Ventures Assets 184 LLC
         : Assigned to Iot Innovations LLC

NPE / troll-pattern signals

1. Shell-entity transfer — present. The transfer from OL Security LIMITED LIABILITY COMPANY to INTELLECTUAL VENTURES ASSETS 184 LLC (Reel 060585/0235) and then to IOT INNOVATIONS LLC (Reel 060698/0604) indicates shell entity behavior. Intellectual Ventures is a known patent licensing and assertion entity. Iot Innovations LLC also appears to be a licensing-focused entity, given the litigation history of the patent.
2. Known asserter in the chain — present. INTELLECTUAL VENTURES ASSETS 184 LLC is a known entity associated with Intellectual Ventures, which is a well-known patent licensing and assertion firm (NPE). This transfer is recorded on Reel 060585/0235.
3. Repeat correspondent across the chain — unclear. The provided assignment records do not specify the correspondent attorney or firm for any of the transactions.
4. Cascading transfers — present. There are two consecutive assignments in a short timeframe in 2022: from OL Security Limited Liability Company to Intellectual Ventures Assets 184 LLC (executed 2022-07-21, recorded 2022-07-21, Reel 060585/0235), and then from Intellectual Ventures Assets 184 LLC to Iot Innovations LLC (executed 2022-07-28, recorded 2022-08-02, Reel 060698/0604). These two transfers occurred within 12 days of each other (execution dates) or 13 days (recording dates), indicating a rapid change of ownership.
5. Pre-litigation transfer — present. The patent was assigned to Iot Innovations LLC on 2022-07-28 (executed) / 2022-08-02 (recorded), via Reel 060698/0604. Litigation cases against this patent were filed in Massachusetts District Court (e.g., 1:23-cv-12528, 1:23-cv-10352) and Georgia Northern District Court (1:23-cv-02475) in 2023, which is within 6 months of the assignment.
6. Bankruptcy fire-sale — not present. There is no information in the provided patent text to suggest a bankruptcy fire-sale.
7. Privateering — unclear. While the chain involves a known NPE, there is no explicit information from the provided data to determine if this is a privateering arrangement where an operating company transfers the patent to an NPE to assert on its behalf.
8. Defensive aggregator (anti-NPE) — not present. The chain ends with Iot Innovations LLC, which is not a known defensive aggregator.

Verdict

NPE — high confidence
The presence of Intellectual Ventures Assets 184 LLC in the assignment chain (Reel 060585/0235), followed by a rapid transfer to Iot Innovations LLC (Reel 060698/0604), strongly indicates an NPE pattern. Furthermore, the timing of these transfers immediately preceding multiple district court litigation filings in 2023 reinforces the conclusion that the patent was acquired for assertion purposes.

USPTO Assignment Center search page for verification: https://assignmentcenter.uspto.gov/

I'll verify the patent and its cited prior art through targeted searches before building the analysis.

Let me pull details on the remaining key cited references.

Scope note and verified identifiers

Target document (searched literally, no auto-correction): US 7,209,876 B2 — "System and method for automated answering of natural language questions and queries." Source: https://patents.google.com/patent/US7209876/en

⚠️ Contradictions to flag explicitly (do not silently reconcile):

1. Date inconsistency. This task header states "Current Date: April 26, 2026," while the system date and the live patent fetch are 2026-09-24. The previously generated litigation and PTAB sections also use April 26, 2026 / June 1, 2026. I flag this rather than correcting it.
2. Status inconsistency inside Google Patents' own page. One search-result header for this record renders "patent/US7209876B2/en active Active," while the page's legal-status block states "Expired – Lifetime" with adjusted expiration 2025-01-24 and 12th-year maintenance fee paid 2018-09-13. The authoritative legal-status field (Expired) controls.
3. "25 citations" table is internally duplicated. The page presents the same 25 references twice ("Citations (25)" and "Patent Citations (25)"). Only four carry the asterisk convention meaning "cited by examiner": US4688195, US5386556, US5963940, US6950793. The other 21 are third-party (likely applicant IDS) citations. This matters for §102 weight: examiner-cited art was substantively before the Office; the rest was merely of record.

Legal framework used. Because US10/293,743 was filed 2002-11-13 (before 2013-03-16), the AIA first-inventor-to-file provisions do not apply. Pre-AIA 35 U.S.C. §102 governs, with the 2001-11-13 priority date as the critical date (§102(a)/(b) art at/after 2001-11-13 is not §102(b); art published more than one year before, i.e., before 2000-11-13, is §102(b)). US patents/applications with pre-2001-11-13 US filing dates can qualify as §102(e) art.

Ranking: most relevant prior art (builds on the prior "Obviousness" section; does not repeat it)

Tier 1 — closest to the claimed subject matter

Tier 2 — US6088692A (U. Central Florida; NL search + ranking), US5873076A (Infonautics/Barr; query-processing & retrieval architecture), US6192338B1 (AT&T; NL knowledge servers as network resources — claim 32), US6026388A (Textwise/Liddy; retrieval UI/highlighting — claim 11), US6314411B1 (Pegasus; artificial-intelligence NL interface), US5592668A (Asymetrix; NL-like query constructs — parallel to the "fill-in-the-blank" embodiment), US5418948A (West/Turtle; concept matching of NL queries).

Tier 3 — background/NL-parsing or translation art: US4688195A, US5386556A, US5309359A, US5404295A, US5768603A, US6126306A, US5734889A, US5499335A, US6052656A, US6081774A, US5991713A, US6327589B1, US6950793B2.

Full citation list with dates, description, and potential §102 mapping

Each entry: full citation → publication/filing date → brief description → claim(s) it potentially implicates under §102. I expressly state where a reference does not fully anticipate, because §102 requires every element "as arranged as recited" in a single reference.

1. US4688195A — Texas Instruments Inc. — Natural-language interface generating system — filed 1983-01-28, issued 1987-08-18. (Examiner-cited.) Foundational system that accepts NL input and generates/constructs an interface or query to a data system. §102: potentially implicates claim 25(b) and claims 3/16 (converting NL input to a query form). No anticipation of claims 1, 13, 25 (missing the keyword search of a natural-language repository and the syntactic answer-phrase filtering of steps (d)/(e)).

2. US5309359A — Boris Katz — Method and apparatus for generating and utilizing annotations to facilitate computer text retrieval — filed 1990-08-16, issued 1994-05-03. Annotates stored text so NL queries can retrieve responsive text. §102: potentially implicates claims 1(a)–(b), 13(a)–(b) (text retrieval responsive to NL query). No anticipation — it relies on a pre-built annotation layer, not an expected-answer-form pattern matched at search time.

3. US5386556A — IBM (Hedin et al.) — Natural language analyzing apparatus and method — filed 1989-03-06, issued 1995-01-31. (Examiner-cited.) Syntactic/semantic parsing of natural-language input. §102: potentially implicates claim 25(b) (converting the NL question to a generic/question form). No anticipation of any independent claim — it is a parser, not a repository search-and-answer pipeline.

4. US5404295A — Boris Katz — Method and apparatus for utilizing annotations to facilitate computer retrieval of database material — filed 1990-08-16, issued 1995-04-04. Same Start/annotation family as the '359 patent; annotations used to retrieve stored material. §102: potentially implicates claims 1(b), 13(b). No anticipation for the same reason.

5. US5768603A — IBM — Method and system for natural language translation — filed 1991-07-25, issued 1998-06-16. Machine-translation of natural language. §102: tangential; at most background for claim 25(b) NL representation. No anticipation.

6. US6126306A — Ando, Shimon — Natural language processing method for converting a first natural language into a second natural language using data structures — filed 1991-09-11, issued 2000-10-03. Data-structure-driven NL-to-NL conversion. §102: tangential; background for claim 25(b)–(c) form conversion. No anticipation.

7. US5418948A — West Publishing Co. (Turtle) — Concept matching of natural language queries with a database of document concepts — filed 1991-10-08, issued 1995-05-23. Maps an NL query to a concept space and matches document concepts. §102: potentially implicates claims 3–4 and 16–17 (converting NL to an intermediate representation used to search). No anticipation — output is ranked documents, not a direct answer field.

8. US5734889A — NEC Corp. — Method and apparatus for retrieving data and inputting retrieved data to spreadsheet including descriptive sentence input means and natural language interface means — filed 1993-07-29, issued 1998-03-31. Descriptive-sentence (NL) input → data retrieval → output to a spreadsheet. §102: potentially implicates claims 25(a)–(c) and claim 31 (formatting output). No anticipation — structured-DB target, no answer-phrase pattern search.

9. US5499335A — [Microsoft Corp.](/litigations/by-plaintiff/Microsoft%20Corp.) — Method and system for providing standard resources in different natural languages — filed 1993-08-24, issued 1996-03-12. Localization of standard resources. §102: tangential. No anticipation.

10. US5592668A — Asymetrix Corp. — Method and apparatus for specifying a query to an information system using natural language-like constructs — filed 1993-08-25, issued 1997-01-07. User specifies a query to an information system via NL-like constructs (structured, partially filled query templates). §102: potentially implicates claims 13(a) and 1(a) (receiving a query containing a variable expression/placeholder) and claim 25(b). No anticipation — it does not disclose initiating a keyword search of a natural-language repository nor filtering for answer phrases syntactically matching an expected answer form.

11. US6052656A — Canon K.K. — Natural language processing system and method for processing input information by predicting kind thereof — filed 1994-06-21, issued 2000-04-18. Predicts the type/kind of input to route NL processing. §102: potentially implicates claims 4/17 (classifying/slotting the question). No anticipation.

12. US6088692A — University of Central Florida — Natural language method and system for searching for and ranking relevant documents from a computer database — filed 1994-12-06, issued 2000-07-11. NL querying plus relevance ranking. §102: potentially implicates claim 1(c)/13(c) and claim 12 (ranking output). No anticipation — ranks documents; does not extract a bound answer value to a variable expression.

13. US6282538B1 — Sun Microsystems (Woods) — Method and apparatus for generating query responses in a computer-based document retrieval system — priority 1995-07-07, issued 2001-08-28. "Hit passage" location with relaxation ranking (word order, adjacency, inflection, semantic distance; passage size scaled to query size) and hyperlinks back to the source document. §102: the strongest single-reference candidate for claim 13(c) and claims 8/21/30 ("answer data arranged in said natural language phrases with respect to said substantive data terms so as to syntactically match"), plus claim 12 ranking. Likely still not a complete anticipation of claims 1/13/25/32, because it ranks pre-existing passages rather than generating a placeholder-bearing "expected answer form" and binding the answer into it. Very strong §103 reference as well.

14. US5963940A — Syracuse University (Liddy, Paik, McKenna, Li) — Natural language information retrieval system and method — filed 1996-08-14, issued 1999-10-05. (Examiner-cited.) DR-LINK NL pipeline: generates an alternative representation of the query and of documents and matches with statistical evidence combination. §102: potentially implicates claims 2, 9, 15, 22 (keyword/phrase extraction and query formation) and claim 25(b). No anticipation — returns ranked documents/folders, not answer phrases.

15. US6026388A — Textwise, LLC (Liddy et al.) — User interface and other enhancements for natural language information retrieval system and method — priority 1995-08-16, issued 2000-02-15. Continuation-family UI enhancements (display, review, feedback) for the DR-LINK system. §102: potentially implicates claim 11 (highlighting) and claim 31 (formatting output). No anticipation of the search/extraction claims.

16. US5873076A — Infonautics Corp. (Barr et al.) — Architecture for processing search queries, retrieving documents identified thereby, and method for using same — filed 1995-09-15, issued 1999-02-16. Query-processing/retrieval architecture. §102: potentially implicates claims 2, 9, 15, 22 and claim 13(b). No anticipation.

17. US6418432B1 — AT&T Corp. — System and method for finding information in a distributed information system using query learning and meta search — filed 1996-04-10, issued 2002-07-09. §102(e) candidate. Query learning plus meta-search over a distributed information system. §102: potentially implicates claims 2, 9, 15, 22 and the multi-engine meta-search of step 200. No anticipation of the expected-answer-form/answer-phrase elements.

18. US6314411B1 — Pegasus Micro-Technologies, Inc. — Artificially intelligent natural language computational interface system for interfacing a human to a data processor having human-like responses — filed 1996-06-11, issued 2001-11-06. AI NL interface producing human-like responses to a data processor. §102: potentially implicates claims 3/16, 25(a)–(c), and claim 31 (formatted direct answers). No anticipation — the target is a local data processor, not a keyword-searched natural-language repository.

19. US5884302A — Ho, Chi Fai — System and method to answer a question — filed 1996-12-02, issued 1999-03-16. Grammatical-structure analyzer (pre-defined context-free structure), grammatical + topic + semantic tables, programming-steps generator/executor, and answer presenter; server + client computers. §102: potentially implicates claims 25(a)–(c), 4/17, 5/18, 7/20, 3/16, and claim 32 (server/client architecture). No anticipation of claims 1/13/25/32 in full — no key-word/phrase search of an external natural-language repository and no answer-phrase syntactic filtering of a result set. Highest-value §103 reference on the "question → form → answer" axis.

20. US6078914A — Open Text Corp. (Redfern, Darren M.) — Natural language meta-search system and method — filed 1996-12-09, issued 2000-06-20. Accepts NL query; parses to extract relevant terms (with synonyms/alternatives); forms per-search-engine queries; receives results; removes redundancies; retrieves complete sources; examines each for match ranking; identifies relevant portions; edits segments into substantially grammatically correct complete sentences/paragraphs; outputs in ranked order in order of the ranking. Its claim 1 (a)–(viii) is the closest published analogue to the patent's steps (a)–(d). §102: strongly implicates claims 1(a)–(b), 2/9/15/22, 12, 13(a)–(b), and 24/31; it is the single most dangerous reference. Assessment: probable anticipation candidate for claim 13 and for claim 1's first two steps, but full anticipation of claims 1/13/25/32 turns on whether Open Text's "relevant portions" are "answer phrases present in syntactically similar form to said search question or query" containing data that replaces a variable expression — on the abstract/claim text available, it edits and outputs passages rather than binding an answer value into an expected-answer-form template. Full-text claim-charting required (see limitations).

21. US6192338B1 — AT&T Corp. — Natural language knowledge servers as network resources — filed 1997-08-12, issued 2001-02-20. NL knowledge server exposed as a network resource. §102: potentially implicates claim 32(a) (server hosting a data search service accessible via client computers to a plurality of users) and claim 15/22. No anticipation of the extraction/matching code means.

22. US6081774A — Novell, Inc. — Natural language information retrieval system and method — filed 1997-08-22, issued 2000-06-27. NL information retrieval. §102: potentially implicates claims 2, 9, 15, 22 and claim 25(d). No anticipation.

23. US5991713A — IBM — Efficient method for compressing, storing, searching and transmitting natural language text — filed 1997-11-26, issued 1999-11-23. Compression/indexing/search of NL text. §102: tangential; background for claim 1(b) data-set handling. No anticipation.

24. US6327589B1 — Microsoft Corp. — Method for searching a file having a format unsupported by a search engine — filed 1998-06-24, issued 2001-12-04. Enables searching files whose format a search engine cannot index. §102: potentially implicates claim 1(b)/13(b) (initiating a search of an information repository and receiving a first data set). No anticipation of the answer-form elements.

25. US6950793B2 — IBM — System and method for deriving natural language representation of formal belief structures — priority 2001-01-12, issued 2005-09-27. (Examiner-cited.) Derives an NL representation from a formal structure — the inverse-direction analogue of the patent's NL-question → formal/expected-form transformation. §102(e) candidate (US priority 2001-01-12, before the 2001-11-13 critical date). §102: potentially implicates claims 4/17 and 25(b)–(c). No anticipation — it generates NL from formal logic; it does not retrieve answers from a repository.

Bottom-line §102 assessment

• No single cited reference appears to anticipate any of independent claims 1, 13, 25, or 32 in its entirety on the disclosures I could verify. Anticipation requires all elements "as arranged as recited"; each Tier-1 reference is missing at least the expected-answer-form generation with a variable expression and the post-search syntactic filtering of the result set into a bound answer value in the same document.
• Closest to a §102 chart: US6078914A (Open Text/Redfern) for claim 13 (and the first two steps of claim 1), with US6282538B1 (Sun/Woods) supplying the order/adjacency-sensitive syntactic passage matching.
• §102(e) qualifiers: US6418432B1 and US6950793B2 have US filing/priority dates preceding 2001-11-13 and can therefore be §102(e) art for their US disclosures.
• §102(b) vs §102(a): all 25 references published well before 2000-11-13 except US6418432B1 (2002-07-09) and US6950793B2 (2005-09-27), which rely on §102(e) rather than printed-publication dates.
• Examiner reliance: only US4688195, US5386556, US5963940 and US6950793 were examiner-cited; the other 21 were merely of record, so a defendant should not assume they were substantively applied during prosecution.

Limitations I will not paper over

1. Full texts (and full claim sets) of most of the 25 references were not retrievable within the available search budget; entries 1–12, 15, 16, 18, 21–24 are characterized from title/assignee/date metadata and abstract-level content. Only US6078914, US6282538, US5963940, US5884302, US6026388 and US6282538 were verified at claim/abstract level in this session.
2. A definitive §102 anticipation conclusion must be built by full-text element-by-element claim charts against each Tier-1 reference; the "potentially implicates" column is a screening result, not a legal conclusion.
3. Consistent with the operating rule "if you do not know something with high confidence, say so": I do not represent that any of the above is a confirmed, complete anticipation.

Cross-reference: This section deliberately does not repeat the previously generated "Obviousness" section; its three illustrative §103 combinations (Ho+Syracuse; Open Text+Sun; Asymetrix+UCF+Textwise) remain the correct vehicle for the obviousness case, and the §102 screening above supplies the reference-level grounding for them.

Under 35 U.S.C. § 103, an invention is considered obvious if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious to a person having ordinary skill in the art (POSA) at the time the invention was made. The analysis involves determining the scope and content of the prior art, identifying differences between the claimed invention and the prior art, and assessing the level of ordinary skill in the pertinent art. A motivation to combine prior art references is generally required to establish obviousness, which can come from the nature of the problem, the prior art itself, or the knowledge of a skilled artisan. Predictable results from combining known elements can also be enough to prove obviousness.

The priority date for US Patent 7209876 is November 13, 2001. Therefore, prior art references published or made publicly available before this date are relevant for an obviousness analysis.

Prior Art References Cited in US7209876:

The "Patent Citations" section of US7209876 lists numerous prior art references, including:

• US4688195A (1987-08-18) - Natural-language interface generating system
• US5386556A (1995-01-31) - Natural language analyzing apparatus and method
• US5309359A (1994-05-03) - Method and apparatus for generating and utilizing annotations to facilitate computer text retrieval
• US5404295A (1995-04-04) - Method and apparatus for utilizing annotations to facilitate computer retrieval of database material
• US5768603A (1998-06-16) - Method and system for natural language translation
• US6126306A (2000-10-03) - Natural language processing method for converting a first natural language into a second natural language using data structures
• US5418948A (1995-05-23) - Concept matching of natural language queries with a database of document concepts
• US5734889A (1998-03-31) - Method and apparatus for retrieving data and inputting retrieved data to spreadsheet including descriptive sentence input means and natural language interface means
• US5499335A (1996-03-12) - Method and system for providing standard resources in different natural languages
• US5592668A (1997-01-07) - Method and apparatus for specifying a query to an information system using natural language-like constructs
• US6052656A (2000-04-18) - Natural language processing system and method for processing input information by predicting kind thereof
• US6088692A (2000-07-11) - Natural language method and system for searching for and ranking relevant documents from a computer database
• US6282538B1 (2001-08-28) - Method and apparatus for generating query responses in a computer-based document retrieval system
• US5963940A (1999-10-05) - Natural language information retrieval system and method
• US6026388A (2000-02-15) - User interface and other enhancements for natural language information retrieval system and method
• US5873076A (1999-02-16) - Architecture for processing search queries, retrieving documents identified thereby, and method for using same
• US6418432B1 (2002-07-09) - System and method for finding information in a distributed information system using query learning and meta search (Note: Publication date is after priority date, but application date may be relevant for 102(e) if it has an earlier filing date)
• US6314411B1 (2001-11-06) - Artificially intelligent natural language computational interface system for interfacing a human to a data processor having human-like responses
• US5884302A (1999-03-16) - System and method to answer a question
• US6078914A (2000-06-20) - Natural language meta-search system and method
• US6192338B1 (2001-02-20) - Natural language knowledge servers as network resources
• US6081774A (2000-06-27) - Natural language information retrieval system and method
• US5991713A (1999-11-23) - Efficient method for compressing, storing, searching and transmitting natural language text
• US6327589B1 (2001-12-04) - Method for searching a file having a format unsupported by a search engine
• US6950793B2 (2005-09-27) - System and method for deriving natural language representation of formal belief structures (Note: Publication date is after priority date)

Obviousness Analysis (Illustrative Combinations):

Given the numerous prior art references in the field of natural language processing and information retrieval, a Person Having Ordinary Skill in the Art (POSA) in November 2001 would likely have been familiar with techniques for parsing natural language, conducting keyword searches, and presenting information. The core inventive concept of US7209876 revolves around transforming a natural language question into an expected answer form, using this form for a meta-search, and presenting direct answers.

Here are some illustrative combinations of prior art that could render certain claims of US7209876 obvious:

Combination 1: US5884302A (Ho) in view of US5963940A (Syracuse University) and general knowledge in the art.

• US5884302A (Ho): This patent describes a system and method to answer a question [cited in US7209876B2]. While the details of Ho's approach would need to be thoroughly examined, it provides a foundational understanding of systems designed to directly answer questions.

• US5963940A (Syracuse University): This patent discloses a natural language information retrieval system and method [cited in US7209876B2]. It likely teaches techniques for processing natural language queries and retrieving relevant information.

• Motivation to Combine: A POSA in 2001, seeking to improve direct question answering, would be motivated to combine a question-answering system like Ho's with natural language information retrieval techniques like those in Syracuse University. The goal would be to enhance the ability of Ho's system to understand and process natural language input more effectively, leading to more accurate and direct answers. If Ho's system did not explicitly detail converting a natural language question into an "expected answer form," a POSA would find it obvious to apply natural language parsing techniques (as suggested by Syracuse University) to derive structural expectations for answers. For example, if Ho processed a question to extract keywords, a POSA would readily understand that structuring these keywords into an "expected answer form" (e.g., "The capital of X is Y") would significantly improve the precision of finding direct answers, rather than just documents containing keywords.

Combination 2: US6078914A (Open Text Corporation) in view of US6282538B1 (Sun Microsystems, Inc.)

• US6078914A (Open Text Corporation): This patent describes a natural language meta-search system and method [cited in US7209876B2]. This directly addresses the concept of performing meta-searches based on natural language.

• US6282538B1 (Sun Microsystems, Inc.): This patent describes a method and apparatus for generating query responses in a computer-based document retrieval system [cited in US7209876B2]. This reference would likely teach mechanisms for formulating queries and processing results to provide responses.

• Motivation to Combine: A POSA would be motivated to combine Open Text Corporation's meta-search capabilities with Sun Microsystems' query response generation. The objective would be to leverage the broad searching power of a meta-search engine (as in Open Text) to gather a wider initial dataset, and then apply more refined query response generation techniques (as in Sun Microsystems) to extract and format direct answers. Specifically, if Open Text provided a raw list of search results, a POSA would naturally look to a system like Sun Microsystems, which focuses on generating responses from documents, to further process those results into direct answers. The "expected answer form" of US7209876 could be seen as an obvious way to bridge the gap between a broad meta-search and a targeted answer extraction.

Combination 3: US5592668A (Asymetrix Corporation) in view of US6088692A (University Of Central Florida) and US6026388A (Textwise, Llc)

• US5592668A (Asymetrix Corporation): This patent describes a method and apparatus for specifying a query to an information system using natural language-like constructs [cited in US7209876B2]. This suggests the use of user-friendly, structured queries.

• US6088692A (University Of Central Florida): This patent discloses a natural language method and system for searching for and ranking relevant documents from a computer database [cited in US7209876B2]. This reference would teach methods for searching and ranking.

• US6026388A (Textwise, Llc): This patent describes user interface and other enhancements for natural language information retrieval system and method [cited in US7209876B2]. This points to the idea of improving the user experience and presentation of results.

• Motivation to Combine: A POSA would be motivated to combine these references to create a more user-friendly and effective natural language question-answering system. Asymetrix Corporation's "natural language-like constructs" could be easily adapted or interpreted as an "expected answer form." The University of Central Florida's work on searching and ranking would provide the backend for finding and prioritizing information. Textwise, Llc's focus on user interface enhancements would naturally lead a POSA to format the retrieved answers directly and clearly for the user, perhaps highlighting the answer data, similar to what is claimed in US7209876. The motivation is to create a seamless user experience, from natural language input to a direct, formatted answer, which would involve integrating the "natural language-like constructs" with search, ranking, and presentation improvements.

General Motivation to Combine:

The field of natural language processing and information retrieval was well-established by 2001. A POSA would be motivated to combine existing techniques to overcome known limitations of traditional keyword-based search engines, such as the difficulty in formulating effective queries and the irrelevance of many search results. The stated problem in US7209876 itself—that prior art keyword searches "rarely mirrors a user's natural language" and often yield "irrelevant" results—highlights a widely recognized need in the art. The motivation to provide "direct answers" instead of mere links was a clear objective in the field. The concept of using grammatical or syntactic patterns to extract specific answers from text, rather than just documents, would be an obvious next step for a POSA working to improve the precision and directness of search results. The use of "variable expressions" in an "expected answer form" is a predictable application of pattern-matching techniques to achieve this goal, as pattern matching in text was a known technique at the time.

Therefore, many claims of US7209876, particularly those related to transforming natural language questions into expected answer forms, conducting searches based on these forms, and providing direct answers, could be considered obvious in light of the extensive prior art in natural language processing and information retrieval available before November 2001.