Obviousness

US patent 6029195 describes a system for customized electronic identification of desirable objects. The core aspects of the invention, as defined and detailed in the patent, include:

1. Automatically constructing a "target profile" for each desirable object, often based on attributes such as word frequency (e.g., TF/IDF for textual objects).
2. Automatically constructing a "target profile interest summary" for each user, describing their interest level in various types of target objects.
3. Evaluating target profiles against user interest summaries to generate a user-customized rank-ordered listing of desirable objects.
4. Automatically updating user target profile interest summaries on a continuing basis to reflect changing interests, often through relevance feedback.
5. Grouping target objects into clusters based on similarity.
6. Providing privacy for user profiles using a cryptographically based proxy server.

A Person Having Ordinary Skill in the Art (POSITA) in the field of information retrieval or electronic media systems around the priority date of US6029195 (November 29, 1994) would possess knowledge of fundamental concepts such as content analysis, user feedback mechanisms, data clustering, and network security.

Here are combinations of prior art references that would likely render the claims of US6029195 obvious under 35 U.S.C. § 103:

Combination 1: Apple's Advanced Technology Group Interface ("A pile metaphor for supporting casual organization of information" by Mander et al., 1992, and "Content awareness in a file system interface: implementing the pile metaphor for organizing information" by Rose et al., 1993) combined with the generally known concept of relevance feedback in information retrieval systems.

• Teachings of Apple's "pile of articles" system: This prior art explicitly describes a system that uses word frequencies and TF/IDF to automatically file (cluster) articles into subpiles, determine keywords for indexing, and label these piles. This directly teaches the automated creation of content-based "target profiles" for desirable objects (articles) and their organization into clusters based on content similarity [cite: "Apple's Advanced Technology Group has developed an interface based on the concept of a "pile of articles". ... The Apple interface uses word frequencies to automatically file articles by picking the pile most similar to the article being filed. This system functions to cluster articles into subpiles, determine key words for indexing by picking the words with the largest TF/IDF (where TF is term (word) frequency and IDF is the inverse document frequency) and label piles by using the determined key words."].
• Motivation for Combination: The US6029195 patent itself identifies a key limitation of the Apple system, stating, "However, no feedback from users is collected or stored, so no performance improvement occurs over time." A POSITA in information retrieval would be well aware that user relevance feedback is a standard and effective technique for improving the performance and personalization of information retrieval and filtering systems. The patent explicitly acknowledges this general knowledge: "a filtering system uses relevance feed back to refine its knowledge of the user's interests: whenever the filtering system identifies a target object as potentially interesting to a user, the user (if an on-line user) provides feedback as to whether or not that target object really is of interest. Such feedback is stored long-term in summarized form, as part of a database of user feedback information, and may be provided either actively or passively." Given a system (Apple's) that effectively organizes information by content, a POSITA would be motivated to enhance its utility by making it adaptive to individual user preferences, thereby addressing the stated deficiency.
• Obviousness: It would have been obvious for a POSITA to combine the content profiling and clustering capabilities of Apple's "pile of articles" system with known relevance feedback mechanisms. This combination would involve:
  1. Collecting active or passive feedback from users regarding the articles presented by the Apple system.
  2. Using this feedback to construct and continuously update a "target profile interest summary" for each user.
  3. Comparing the content-based "target profiles" of articles (as derived by Apple's methods) against these dynamically updated "user interest summaries" to generate a user-customized, rank-ordered listing of articles most likely to be of interest.
     This combination directly teaches the core method of US6029195 for customized electronic identification of desirable objects through profiling, matching, and dynamic user preference updating.

Combination 2: Apple's Advanced Technology Group Interface (Mander et al., 1992, Rose et al., 1993) combined with U.S. Pat. No. 5,331,554 (Graham et al.).

• Teachings of Graham et al. (U.S. Pat. No. 5,331,554): This patent discloses "a method for retrieving segments of a manual by comparing a query with nodes in a decision tree." This teaches the use of decision trees as an efficient method for navigating and retrieving information from structured data.
• Motivation for Combination: Apple's system organizes articles into clusters (piles). The US6029195 patent describes hierarchical clustering as "often useful" and notes that a "hierarchical cluster tree therefore serves as a decision tree for identifying target objects." A POSITA, having a system that organizes information into clusters, would seek efficient methods for users to navigate and retrieve specific information from this organized structure. Applying known data retrieval techniques to improve user experience is a common motivation in the art.
• Obviousness: It would have been obvious for a POSITA to combine the article clustering (which can be hierarchical) provided by Apple's system with the decision tree retrieval methods taught by Graham et al. This would enable more efficient and structured navigation through the clustered articles, allowing users to rapidly identify desirable objects by traversing a decision tree built upon the cluster hierarchy.

Combination 3: Apple's Advanced Technology Group Interface (Mander et al., 1992, Rose et al., 1993) combined with U.S. Pat. No. 5,331,556.

• Teachings of U.S. Pat. No. 5,331,556: This patent "addresses techniques for deriving morphological part-of-speech information and thus to make use of the similarities of different forms of the same word (e.g. 'article' and 'articles')." This teaches methods for improving text processing by normalizing word forms.
• Motivation for Combination: Apple's system relies on word frequencies and TF/IDF for profiling articles. The accuracy of such content profiles can be negatively affected by variations in word forms (e.g., "run," "running," "runs" being treated as distinct terms). The US6029195 patent itself mentions "a related technique is to replace each word by its morphological stem, so that 'staple', 'stapler', and 'staples' are all replaced by 'staple.'" A POSITA would be motivated to improve the robustness and accuracy of content-based profiles by ensuring that morphologically related words are treated consistently, thereby enhancing the quality of similarity measurements.
• Obviousness: It would have been obvious for a POSITA to incorporate the morphological text processing techniques taught by U.S. Pat. No. 5,331,556 into the text analysis component of Apple's system. This would lead to more accurate "target profiles" derived from word frequencies and TF/IDF, as the system would more effectively identify the underlying topics by treating different forms of the same word as equivalent.

Combination 4: The overall system for customized electronic identification of desirable objects (as made obvious by Combination 1) combined with generally known network security and privacy techniques (e.g., proxy servers and cryptography).

• Teachings of the overall system (from Combination 1): Provides a system for personalizing information retrieval by generating and updating user interest profiles based on feedback.
• Motivation for Combination: As systems collect detailed personal preferences and behaviors (like "target profile interest summaries"), privacy concerns become significant. In the mid-1990s, the importance of user privacy and the use of technical means to protect it (e.g., anonymity, data encryption, access control) were increasingly understood in the context of online services and data processing. The patent itself highlights the need for privacy protection: "a particular user may not wish to make public all of the interests recorded in the user's target profile interest summary" and "It is therefore necessary that data in a user's target profile interest summary be protected from unwanted disclosure except with the user's agreement."
• Obviousness: A POSITA designing such a personalized system, aware of the privacy implications of collecting user data, would find it obvious to integrate known security and privacy mechanisms. The use of a proxy server to dissociate a user's true identity from a pseudonym via cryptographic techniques, as described in US6029195, would be an obvious application of established network security principles to protect sensitive user information. [cite: "Additionally, a cryptographically based proxy server is provided to ensure the privacy of a user's target profile interest summary, by giving the user control over the ability of third parties to access this summary and to identify or contact the user.", "the disclosed system provides a solution to the privacy problem by using a proxy server which acts as an intermediary between the information provider and the user.", "the proxy server dissociates the user's true identity from the pseudonym by the use of cryptographic techniques."].