System and method of automatically generating building dialog services by exploiting the content and structure of websites
First Claim
Patent Images
1. A system for providing a spoken dialog interface to a website in at least one of an HTML markup language and an XML markup language, the system comprising:
- a first module controlling a processor to compute a summary for each respective node in the website based on a list of anchor texts, each anchor text in the list of anchor texts being in a hyperlink of another webpage that points to each respective node, each anchor text selected according to a term density associated with a percentage of salient words within a phrase relative to a total number of words in the phrase;
a second module controlling the processor to identify hyperlinks associated with each respective node and a weight associated with each respective hyperlink, wherein the weight relates to a percent of salient words in the anchor text of the respective hyperlink relative to total words in anchor texts of the hyperlinks associated with each respective node; and
a third module controlling the processor to use the hyperlinks, weights, and the summary for each respective node to enable the spoken dialog interface to interact with a user in a live spoken dialog, wherein nodes in the website are matched with open-ended user spoken input and prompts from the spoken dialog interface seeking the open-ended user spoken input are independent of the nodes in the website.
4 Assignments
0 Petitions
Accused Products
Abstract
A method and system are disclosed for providing a dialog interface for a website. The method comprises at each node in a website, computing a summary, a document description and an alias. A dialog manager within a spoken dialog service utilizes the summary, document description and alias for each website node to generate prompts to a user, wherein nodes in the website are matched with user requests. In this manner, a spoken dialog interface to the website content and navigation may be generated automatically.
22 Citations
18 Claims
-
1. A system for providing a spoken dialog interface to a website in at least one of an HTML markup language and an XML markup language, the system comprising:
-
a first module controlling a processor to compute a summary for each respective node in the website based on a list of anchor texts, each anchor text in the list of anchor texts being in a hyperlink of another webpage that points to each respective node, each anchor text selected according to a term density associated with a percentage of salient words within a phrase relative to a total number of words in the phrase; a second module controlling the processor to identify hyperlinks associated with each respective node and a weight associated with each respective hyperlink, wherein the weight relates to a percent of salient words in the anchor text of the respective hyperlink relative to total words in anchor texts of the hyperlinks associated with each respective node; and a third module controlling the processor to use the hyperlinks, weights, and the summary for each respective node to enable the spoken dialog interface to interact with a user in a live spoken dialog, wherein nodes in the website are matched with open-ended user spoken input and prompts from the spoken dialog interface seeking the open-ended user spoken input are independent of the nodes in the website. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11)
-
-
12. A spoken dialog system for interfacing via natural language with a website in at least one of an HTML markup language and an XML markup language, the spoken dialog system comprising:
-
a first module controlling a processor, at each node of the website, to compute a summary for each respective node based on a list of anchor texts, each anchor text in the list of anchor texts being in a hyperlink of another webpage that points to each respective node, each anchor text selected according to a term density associated with a percentage of salient words within a phrase relative to a total number of words in the phrase; a second module controlling the processor to identify hyperlinks associated with each respective node, wherein the weight relates to a percent of salient words in an anchor text of the respective hyperlink relative to the total words in anchor texts of all links pointing to each respective node; and a third module controlling the processor to use the identified hyperlinks, weights, and the summary for each respective node to enable interaction with a user in a live spoken dialog, wherein nodes in the website are matched with open-ended user spoken input and prompts from the spoken dialog interface are seeking the open-ended user spoken input are independent of the nodes in the website. - View Dependent Claims (13, 14, 15)
-
-
16. A system for providing a spoken dialog interface to a website in at least one of an HTML markup language and an XML markup language, the system comprising:
-
a first module controlling a processor, for each node within the website, to select a list of anchor texts that describe content beneath each respective node, wherein each anchor text is selected based on a term density associated with a percentage of salient words within a phrase relative to a total number of words in the phrase and on whether a respective anchor text contains at least one similar salient word to anchor texts already in a node summary comprise at least two anchor texts selected from the list of anchored texts; a second module controlling the processor to organize the list of anchor texts into the node summary; and a third module controlling the processor to utilize the node summary to generate prompts to a user via the spoken dialog interface in a live spoken dialog, wherein nodes in the website are matched with open-ended user spoken input and prompts from the spoken dialog interface are seeking the open-ended user spoken input are independent of the nodes in the website. - View Dependent Claims (17, 18)
-
Specification