Obviousness

Obviousness Analysis of US Patent 8239358 Under 35 U.S.C. § 103

This analysis identifies combinations of prior art references that would render the independent claims of US Patent 8239358 obvious to a person having ordinary skill in the art (PHOSITA) at the time of the invention (priority date of February 6, 2007).

Person Having Ordinary Skill in the Art (PHOSITA)

A PHOSITA in 2007 would be a software engineer or researcher with experience in information retrieval, natural language processing (NLP), and web application development. They would possess knowledge of how search engines function, various document summarization techniques (including single and multi-document summarization), and standard user interface (UI) design principles for displaying information and facilitating user interaction on the internet.

Problem Addressed by US8239358

The background of US8239358 identifies a key problem with existing internet search methods: users face a "large amount of Web pages" and "spends a considerable time before finding satisfactory results," often dealing with "contradictory" information or content "irrelevant" to their specific query. [cite: US8239358 Description, Background of the Invention] The patent aims to improve user satisfaction by presenting an automatically generated multi-document summary as a primary search result, providing a concise digest of relevant information with direct links to original sources, and offering extensive user interaction capabilities. [cite: US8239358 Description, Detailed Description of the Preferred Embodiments]

Obviousness Combination

The independent claims (1, 5, and 7) of US8239358 would have been obvious to a PHOSITA by combining the "iResearch Reporter" system (as described in the Wikipedia entry on Multi-document summarization), US Patent 5,623,652 (Apple Computer, Inc.) [cite: US5623652A Abstract], US Patent 6,775,677 (International Business Machines Corporation) [cite: US6775667B1 Abstract], and general knowledge of user interface design common in the art.

Motivation for Combination: A PHOSITA, motivated to address the identified problems of information overload and poor user satisfaction in web search, would logically seek to integrate known multi-document summarization capabilities into search engine interfaces. The existence of systems like "iResearch Reporter" explicitly combining search, multi-document summarization, and source linking demonstrates that this integration was known and desired. Enhancing such a system with common user interaction features found in other information retrieval systems would be a natural and obvious step to further improve usability and user control.

Analysis Against Independent Claim 1 (Method Claim)

Claim 1 describes a method for searching and summarizing multiple documents with extensive user interaction.

• Sending a query to a search component and receiving results: The "iResearch Reporter" system "accepts user-entered query, passes it on to Google search engine, retrieves multiple relevant documents." This directly teaches the initial steps of query submission and receiving search results. US5623652A further teaches a "method and apparatus for searching for information" and "displaying to a user... representations of portions of search results." [cite: US5623652A Abstract]
• Passing results to a summarization component to produce a multi-document summary (digest) on a query-specified topic, by analyzing syntax/semantics, and extracting sentences/text fragments: The "iResearch Reporter" system "produces categorized, easily readable natural language summary reports covering multiple documents in retrieved set." This clearly describes the creation of a multi-document summary responsive to the query. Multi-document summarization techniques, including extraction based on semantic analysis, were well-known prior to the priority date, as evidenced by US5384703A (Xerox) which describes "summarizing one or more documents according to a theme specified by a user query" by "identifying and scoring concepts" and "extracting important sentences" [cite: US5384703A Abstract], and by the existence of Document Understanding Conferences (DUC) which began in 2001 addressing multi-document summarization challenges.
• Each sentence/text fragment in the summary comprises an attached reference link to its source document: The "iResearch Reporter" explicitly states that "all extracts linked to original documents on the Web." This directly teaches the inclusion of source links within the multi-document summary fragments.
• Displaying the multi-document summary as a product of the query: The "iResearch Reporter" "returns the summary to a client-side system for display, for example, within a web browser." Presenting this summary as the direct output of a search query is a straightforward integration into a search interface.
• Receiving and displaying search results in addition to the multi-document summary: A PHOSITA would find it obvious to display both the generated summary (as from "iResearch Reporter") and the conventional list of search results (as taught by US5623652A [cite: US5623652A Abstract]). This provides users with flexibility, allowing them to quickly grasp the main points or delve into individual documents, addressing the problem of varying user information needs.
• Extracting and displaying semantic concepts to a user: The "iResearch Reporter" produces "categorized" summary reports, which implies the extraction and presentation of concepts. Furthermore, US6775677B1 teaches a "system, method, and program product for identifying and describing topics in a collection of electronic documents." [cite: US6775677B1 Abstract] A PHOSITA would understand the value of presenting these concepts to a user for better understanding of the summarized content.
• Enabling extensive user interaction (refine query, navigate summary, view source documents via links/results, request another summary with exclusions/inclusions, choose display preferences, modify/regroup summary, save, and rate): These user interaction features are common and well-known in web-based information retrieval systems by the priority date.
  • Refining the query and viewing source documents from search results are fundamental functionalities of any search engine (e.g., US5623652A [cite: US5623652A Abstract]).
  • Viewing source documents via attached reference links in the digest is explicitly taught by the "iResearch Reporter".
  • Navigating parts of the summary (e.g., scrolling, clicking headings) is standard functionality for any displayed text on a web page.
  • Requesting another summary with modifications (e.g., excluding/including documents or fragments, modifying/regrouping content): Given the existence of multi-document summarization systems ("iResearch Reporter", US5384703A [cite: US5384703A Abstract]) and the general trend toward user customization in web applications and "interactive searching techniques" (US7707220B2 [cite: US7707220B2 Abstract]), a PHOSITA would find it obvious to implement controls allowing users to refine the summary output. This directly addresses the problem of users not finding satisfactory results and wanting more control over the information they receive.
  • Choosing a display preference for attached reference links (e.g., inline, at the end, full URL vs. short title) is a routine UI customization option.
  • Saving and rating the currently displayed multi-document summary for feedback: These are common features in web applications for content management and collecting user feedback, motivated by a desire to enhance user experience and improve system performance (as stated in US8239358's description for "self-learning features").

Analysis Against Independent Claim 5 (Method Claim)

Claim 5 presents a method similar to Claim 1, focusing on receiving and displaying the multi-document summary as the primary product of the query, along with similar user interaction steps. For the same reasons outlined for Claim 1, the combination of the "iResearch Reporter" system, US5623652A [cite: US5623652A Abstract], US6775677B1 [cite: US6775677B1 Abstract], and general UI design principles would render Claim 5 obvious. The core inventive concept of providing a multi-document summary with source links alongside traditional search results and offering interactive controls is anticipated or obvious in light of this prior art.

Analysis Against Independent Claim 7 (System Claim)

Claim 7 describes a system (comprising "means for" various functions) to achieve the methods of Claims 1 and 5.

• Means for interfacing with a user: Inherently present in any web-based system, as taught by the "iResearch Reporter" demo nature and US5623652A's user display [cite: US5623652A Abstract].
• Means for searching the documents using a query from the user: Taught by the "iResearch Reporter" which "passes [query] on to Google search engine", and US5623652A's searching apparatus. [cite: US5623652A Abstract]
• Means for generating a multi-document summary comprising a digest with attached reference links: Taught by the "iResearch Reporter" system, which "produces...summary reports covering multiple documents...all extracts linked to original documents on the Web." This directly covers the generation of the digest with linked fragments.
• Means for presenting said multi-document summary to the user as a product of said query: Taught by the "iResearch Reporter" which "returns the summary to a client-side system for display."
• Means for receiving and presenting said results from said search component in addition to presenting said multi-document summary: This "means" would be an obvious combination of the "iResearch Reporter's" summary output and the standard search result display component of a search engine (e.g., as taught by US5623652A [cite: US5623652A Abstract]), motivated by user choice and comprehensive information delivery.
• Means for extracting and displaying semantic concepts to the user: Taught by the "categorized" reports of "iResearch Reporter" and the topic identification capabilities of US6775677B1. [cite: US6775677B1 Abstract]
• Means for enabling user interaction (as detailed in Claim 1 and 5): These "means" correspond to standard software modules and user interface components that a PHOSITA would readily implement in any interactive web application or information retrieval system. For example, the means for viewing source documents via reference links is directly taught by the functionality described for "iResearch Reporter". The other interactive "means" are well-understood design choices for enhancing user experience in a search environment, particularly when a novel display (like a multi-document summary) is introduced.

In conclusion, the combination of the "iResearch Reporter" system, US Patent 5,623,652, US Patent 6,775,677, and the general knowledge of a PHOSITA regarding user interface design would render independent claims 1, 5, and 7 of US8239358 obvious.