Widget searching utilizing task framework

• US 7,996,783 B2
• Filed: 03/02/2006
• Issued: 08/09/2011
• Est. Priority Date: 03/02/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Katariya, Sanjeev, Ramsey, William D.
• Primary Examiner(s): Bashore; William L
• Assistant Examiner(s): Distefano; Gregory A
• Application Number:
  US11/367,292
• Publication Number:
  US 20070209013A1
• Time in Patent Office: 1,986 Days
• Field of Search: 715/769, 715/224, 715/826, 715/764, 345/470
• US Class Current: 715/764
• CPC Class Codes: G06F 16/972 Access to data in other repository systems, e.g. legacy data or dynamic Web page generation

Abstract

A task framework and a semantic reasoning engine are combined to provide a scalable mechanism for dealing with extremely large numbers of widgets, allowing users to both find a widget and automatically fill-in whatever functionality is available on the widget. Calling applications are employed to obtain task information from each widget. The calling application also receives user queries that can be resolved by a widget. A task reasoning process based on an adaptive semantic reasoning engine utilizes the task information to select a widget best suited to respond to a user'"'"'s query. The task reasoning process can also be employed to determine “best-guess” slot filling of the selected widget. The calling application can then invoke the selected widget and, if available, fill appropriate slots with information to facilitate user interaction with the selected widget. Instances can be client- and/or server-side based.

Claims

1. A computer readable storage medium having computer-executable components stored thereon that facilitates widget searching, comprising:

a calling application component that receives a natural language query and is associated with a plurality of widgets configured to expose parametric task data to the calling application component and including one or more widgets associated with the natural language query, wherein the calling application interfaces with the plurality of widgets to obtain parametric task data from each of the plurality of widgets, the parametric task data includes at least one of task capabilities, types of acceptable slots, or information to facilitate filling of widget slots based on the query; and

an adaptive task reasoning component that receives the natural language query from the calling application component and utilizes the obtained parametric task data to automatically select one or more appropriate widgets from among the plurality of widgets in response to the natural language query, wherein the adaptive task reasoning component further determines appropriate slot filling for the selected widgets from the obtained parametric task data, and provides the slot filling for each of the selected widgets to the calling application component;

wherein the calling application component invokes a widget using the slot filling for that widget to allow a user to utilize the functionality of the one or more selected widgets, the widget selected from among the one or more widgets selected by the adaptive task reasoning component in response to the natural language query, and wherein the adaptive task reasoning component employs user feedback related to at least one selected widget to facilitate in subsequent widget selections and slot filling in response to subsequent natural language queries.

2. The computer readable storage medium of claim 1, the calling application component receives the natural language query directly from a user or via a Web browser.

3. The computer readable storage medium of claim 1, the calling application component or the adaptive task reasoning component resides on a server.

4. The computer readable storage medium of claim 1, the natural language query comprising a modality independent query input.

5. A server-side widget search system that utilizes the computer readable storage medium of claim 1 to facilitate in providing at least one appropriate widget to a user via a Web browser.

6. The computer readable storage medium of claim 1, wherein, upon determination that the one or more widgets includes a single widget, the calling application component provides the selected widget to the user without providing another widget concurrently.

7. A computer-implemented method for facilitating widget searching, comprising:

employing a calling application to obtain task associated data exposed by each of a plurality of widgets, each widget residing on a client or server and configured to expose the task associated data to the calling application, the task associated data includes widget task capabilities, types of acceptable slots, or information to facilitate widget slot filling;

receiving a user query associated with one or more of the plurality of widgets, wherein the user query is a natural language query;

passing the task associated data from the calling application to a task reasoning process that is based, at least in part, on an adaptive semantic reasoning engine;

employing the adaptive task reasoning engine to apply adaptive task reasoning and the task associated data obtained from the plurality of widgets to facilitate in selecting one or more appropriate widgets responsive to the user query;

utilizing the task associated data to determine slot filling for the selected widgets;

employing selected widget feedback from a user to facilitate in determining the one or more appropriate widgets and slot filling in response to the user query; and

calling at least one of the selected widgets with appropriate slots filled.

8. The method of claim 7 further comprising:

receiving the user query associated with one or more of the plurality of widgets directly from a user and/or via a Web browser.

9. The method of claim 7 further comprising:

relaying the selected widgets and slot filling for the selected widgets to the calling application; and

invoking at least one of the selected widgets with filled slots for utilization by a user.

10. A widget search method that employs the method of claim 7 to facilitate client-side widget searching.

11. A computer readable storage medium that facilitates widget searching, comprising:

means for receiving a natural language query processible by a widget;

means for determining an appropriate widget and slot filling for the natural language query, employed by an adaptive task reasoning process;

means for invoking the determined widget for interaction with a user, the determined widget configured to expose metadata describing tasks performed by that widget;

means for obtaining user feedback, slot filling data, or metadata related to the system, or a combination thereof, from interaction with the determined widget;

means for determining appropriate slot filling for the selected widget based on the natural language query;

means for filling the appropriate slots of the selected widget based on the natural language query;

means for employing an adaptive semantic reasoning engine to facilitate improvement on the adaptive task reasoning based upon the obtained user feedback, upon the obtained slot filling data, upon the obtained metadata, or a combination thereof, wherein the adaptive semantic reasoning engine employs statistical learning models; and

means for employing the user feedback relating to the determined widget to facilitate in future widget selections and slot filling in response to natural language queries.