Obviousness

Obviousness Analysis of US Patent 8527274 under 35 U.S.C. § 103

The independent claims of US patent 8527274 (Claims 1 and 13) describe a method and system, respectively, for delivering targeted advertisements and tracking interactions in voice recognition contexts. A key feature is the use of user interaction with selected advertisements to interpret incomplete or ambiguous natural language requests. A person having ordinary skill in the art (PHOSITA) in the field of natural language processing (NLP), speech recognition, and online advertising at the time of the invention (priority date: February 6, 2007) would have been motivated to combine existing technologies to arrive at the claimed invention.

Claim Breakdown and Relevant Prior Art

The independent claims (Claim 1 for method, Claim 13 for system) generally comprise the following steps:

1. Receiving a natural language utterance from a voice-enabled device.
2. Processing the utterance to identify one or more requests.
3. Determining that a request is incomplete or ambiguous.
4. Selecting one or more advertisements associated with the utterance.
5. Outputting the selected advertisements.
6. Monitoring user interaction with the advertisements.
7. Interpreting the incomplete or ambiguous request based on that interaction.

Several prior art references, particularly those explicitly cited within US8527274, disclose elements of these claims.

• Natural Language Processing and Conversational Voice User Interfaces: Numerous patents by VoiceBox Technologies Corp., the original assignee, cover the fundamental aspects of natural language speech processing, identifying requests, and handling conversational interactions. These include:

  • U.S. Pat. No. 7,398,209 (priority June 3, 2003): "Systems and Methods for Responding to Natural Language Speech Utterance."
  • U.S. Pat. No. 7,693,720 (priority June 15, 2003): "Mobile Systems and Methods for Responding to Natural Language Speech Utterance."
  • U.S. Pat. No. 7,640,160 (priority August 5, 2005): "Systems and Methods for Responding to Natural Language Speech Utterance."
  • U.S. Pat. No. 7,949,529 (priority August 29, 2005): "Mobile Systems and Methods of Supporting Natural Language Human-Machine Interactions."
    These references collectively teach the first two elements of the claims: receiving natural language utterances and processing them to identify requests.

• Handling Incomplete or Ambiguous Requests: The prior art also addressed the challenges of ambiguous or incomplete voice inputs:

  • U.S. Pat. No. 7,620,549 (priority August 10, 2005): "System and Method of Supporting Adaptive Misrecognition in Conversational Speech." This patent explicitly teaches reinterpreting utterances based on subsequent information, especially when initial interpretations are incorrect.
  • U.S. Pat. No. 8,073,681 (priority October 16, 2006): "System and Method for a Cooperative Conversational Voice User Interface." This patent describes systems that engage users in "cooperative, conversational interaction" to resolve requests and advance a conversation, including determining if requests are "incomplete or ambiguous" and prompting users for clarification. The '274 patent itself states that its conversational language processor may use techniques described in the '681 patent to identify conversation types, interpret utterances, and identify requests.

• Targeted Advertising and Interaction Tracking: By 2007, online advertising systems were well-established, offering targeted advertisements and sophisticated tracking mechanisms. For example, Google's AdWords, launched in 2000, allowed advertisers to deliver targeted ads based on keywords and user context, and critically, tracked user interactions such as click-through rates (CTR) and conversions. Such systems encompassed the concepts of selecting advertisements, outputting them (typically graphically, but also audibly in contexts like telematics, as the '274 patent mentions), and monitoring user interaction.

Obviousness Combination and Motivation

A strong combination for rendering the claims obvious would be U.S. Pat. No. 8,073,681 (Cooperative Conversational Voice User Interface) in view of general knowledge of online advertising systems (pre-2007).

Reasoning for Combination:

1. Shared Problem and Goal: US8073681 addresses the problem of effectively handling natural language speech utterances, particularly those that are incomplete or ambiguous, to foster a "cooperative, conversational interaction." The explicit goal is to "resolve requests and advance a conversation." The '274 patent's background further highlights the "missed opportunities for providing valuable and relevant information to users" due to inadequate voice user interfaces and the need for "new and effective ways to reach consumers" through advertising. This demonstrates a clear motivation to improve conversational interfaces and to integrate commercial aspects.

2. Completing the Conversational Loop: US8073681 teaches determining when requests are "incomplete or ambiguous" and taking action, such as generating a "system response [that] may prompt a user to clarify an incomplete and/or ambiguous request." A PHOSITA, aiming to enhance the cooperative nature and commercial viability of such a system, would be motivated to integrate targeted advertisements into this disambiguation process.

3. Advertising as Implicit Feedback for Disambiguation: It was a well-known principle in 2007 that user interaction provides valuable feedback. Online advertising systems demonstrated that users interact with relevant advertisements. A PHOSITA would recognize that presenting a highly targeted advertisement (selected based on the ambiguous request and user profile, similar to how conventional online ads are targeted) and then monitoring the user's interaction with that ad (e.g., a voice confirmation, a touch, a click) offers a powerful form of implicit or explicit feedback to "interpret the request that is incomplete or ambiguous." For instance, if a user ambiguously requests "Find me a good place to eat," and an ad for "Chang's Chinese Restaurant" is presented, a user's interaction with that ad (e.g., "Yes, call Chang's") directly resolves the ambiguity of the original request.

4. Synergistic Benefits: The combination yields synergistic benefits:

   • Improved User Experience: By using ad interaction for disambiguation, the system becomes more intelligent and responsive, reducing the need for explicit clarification prompts and leading to a smoother user experience, aligning with the "cooperative" goals of US8073681.
   • Monetization and Enhanced Ad Relevance: The voice interface gains a new revenue stream through targeted advertising. Furthermore, by making advertisements integral to resolving user requests, their contextual relevance and engagement are increased, leading to higher effectiveness for advertisers. The '274 patent explicitly states that advertisements can "improv[e] the consumers' interaction with a device."

Therefore, a PHOSITA, seeking to build upon the robust conversational and ambiguity-handling capabilities of systems like those described in US8073681 and motivated by the economic opportunities of targeted advertising and the desire to improve user interaction, would find it obvious to combine these known elements. The integration of ad selection, presentation, and interaction monitoring to specifically refine the interpretation of ambiguous natural language utterances represents a logical extension of existing technologies and practices.