Prior art — US Patent 8069073

As a senior US patent analyst, I have reviewed the prior art cited by the USPTO examiner during the prosecution of US patent 8,069,073 ('073 patent). The '073 patent, with a priority date of December 23, 1999, is directed to a system for bilateral or multilateral decision-making where preference profiles for both a "party" and a "counterparty" are generated using forced-choice questions and conjoint analysis, and then compared to find a "closeness of fit."

Below is an analysis of the most relevant prior art references cited against the '073 patent and their potential to anticipate the patent's claims under 35 U.S.C. § 102.

Analysis of Cited Prior Art

The core of the invention claimed in the '073 patent is the bilateral preference analysis. Unlike unilateral systems where a producer models consumer preferences, the '073 patent claims a method where preference profiles are generated for parties on both sides of a potential transaction (e.g., a job seeker and an employer) and then matched. The key elements of the primary independent claim include:

Supplying forced-choice questions to both a party and a counterparty.
Eliciting responses from both.
Determining preference profiles for both using conjoint analysis.
Delivering a match-list based on an analysis of both profiles.

1. U.S. Patent 5,758,328 A

Full Citation: US 5,758,328 A, "Method and apparatus for matching a person to a product or service"
Publication Date: May 26, 1998 (Filed: Aug 2, 1996)
Brief Description: This patent describes a system for matching individuals with products or services, specifically in the context of financial products like insurance policies or investment vehicles. The system gathers data from a person (e.g., age, income, risk tolerance) and data about various financial products (e.g., policy terms, investment risk). It then uses a computer to generate a ranked list of suitable products for that specific person.
Potential Anticipation Analysis:
- This reference discloses a system for matching a person (a "party") with a product/service (offered by a "counterparty"). It involves collecting data from the user and comparing it against a database of product attributes.
- However, US 5,758,328 appears to describe a unilateral preference system. It focuses on eliciting the consumer's preferences and characteristics to match against static, predefined product attributes. It does not explicitly describe a process where the "counterparty" (e.g., the insurance provider) also answers a series of forced-choice questions to generate a dynamic "success profile" or preference profile for the type of customer it seeks. The analysis is one-sided.
- Therefore, this patent likely does not anticipate the core claims of the '073 patent, because it fails to teach the bilateral step of supplying forced-choice questions to the counterparty to generate a corresponding preference profile for matching.

2. U.S. Patent 5,832,497 A

Full Citation: US 5,832,497 A, "Automated collaborative filtering system"
Publication Date: November 3, 1998 (Filed: May 2, 1997)
Brief Description: This patent, assigned to Net Perceptions, Inc., describes a method for providing personalized recommendations to users. It operates by collecting preference data (ratings) from a large group of users on various items. To make a recommendation for a specific user, the system identifies other users with similar tastes (a "neighborhood") and recommends items that those similar users liked but the target user has not yet rated.
Potential Anticipation Analysis:
- This reference is a foundational patent in the field of collaborative filtering, a method for making recommendations. It clearly discloses a system for eliciting user preferences and using those preferences to match users with items.
- However, the matching is based on user-to-user similarity, not a direct party-to-counterparty preference profile match as claimed in the '073 patent. The "counterparty" (the item being recommended) is passive; it does not have a preference profile. The system recommends items based on the aggregated preferences of other similar users. It does not, for example, have a movie (the counterparty) generate a profile of its ideal viewer to match against a specific user's (the party's) profile.
- Because it lacks the generation of a preference profile for the counterparty side of the transaction, this reference does not anticipate the claims of the '073 patent.

3. U.S. Patent 5,970,475 A

Full Citation: US 5,970,475 A, "Computer-aided, multi-attribute selection process"
Publication Date: October 19, 1999 (Filed: Sep 26, 1997)
Brief Description: This patent discloses a computer system that helps a user make a selection from a set of alternatives where each alternative has multiple attributes (e.g., selecting a car based on price, color, and MPG). The system elicits the user's preferences for each attribute, including the relative importance of each attribute. It then calculates a utility score for each alternative based on the user's stated preferences and presents a ranked list. The method it describes is consistent with the principles of conjoint analysis, where trade-offs between attributes are evaluated.
Potential Anticipation Analysis:
- This reference strongly teaches the "party" side of the '073 patent's claims. It describes eliciting multi-attribute preferences from a user and using a utility-based calculation (the foundation of conjoint analysis) to rank options.
- However, like the other references, it appears to be a unilateral system. The process is centered entirely on the decision-maker (the "party"). The alternatives (the "counterparties") are described by their fixed attributes; the system does not involve gathering preferences from the alternatives. For example, in a job-seeking context, this patent would describe the job seeker's process of ranking companies, but not the companies' process of creating a preference profile for their ideal candidate.
- Therefore, this reference does not anticipate the claims because it is missing the bilateral matching element.

4. U.S. Patent 6,029,195 A

Full Citation: US 6,029,195 A, "System for matching buyer and seller of goods and services in a network"
Publication Date: February 22, 2000 (Filed: April 8, 1998)
Brief Description: This patent describes an electronic marketplace where both buyers and sellers can input their requirements. A buyer can specify the attributes of a product they wish to purchase, and a seller can specify the attributes of a product they wish to sell. The system's "matching engine" then compares the specifications from both sides to identify potential matches and facilitate a transaction. This system is explicitly two-sided.
Potential Anticipation Analysis:
- This is arguably the most relevant prior art reference cited by the examiner. It explicitly discloses a bilateral matching system where both a "party" (buyer) and a "counterparty" (seller) provide input about their needs and offerings. A central system then performs the match.
- The central question for anticipation is whether this reference teaches the use of "forced choice questions" to determine "preference profiles" via "conjoint analysis." The '195 patent speaks of buyers and sellers inputting "specifications," "requirements," and "attributes." This language suggests a more direct, explicit statement of needs rather than the indirect elicitation of underlying utility values through the trade-off questions characteristic of conjoint analysis. For example, a buyer might specify "Color: Blue, Price: <$20," which is different from being forced to choose between a blue, $25 product and a green, $18 product to reveal their underlying preference structure.
- While the '195 patent teaches a bilateral matching system, it likely does not fully anticipate the claims of the '073 patent. A strong argument can be made that it does not disclose the specific method of preference elicitation (forced choice/conjoint analysis) that is a required limitation of the '073 claims. It describes matching based on explicit specifications, not implicitly derived preference profiles.