Microsoft Technology Licensing, LLC

Dong Han Wang

Dilek Zeynep Hakkani-Tur

Gokhan Tur

Larry Paul Heck

Jun Liu

Arun Sacheti

Nanshan Zeng

Parthasarathy Govindarajen

Xiaodong Fan

Jinghua Chen

Diane Shambaugh

Nektarios Ioannides

Avinash Vemuluru

Nikita Astrakhantsev

Liang Tang

Alexander Jack Sunell

Alexandra Victoria Foote

Elena Usinskiene

Qianyue You

Souvick Sarkar

Adam Jeffrey Curtis

Jeffrey Roger Devries

Xiaofei Huang

The present disclosure relates to systems, devices, and methods for a visual exploration of search engine results. The systems, devices, and methods may present, in response to receiving the search engine results for a query to perform a task, a search result webpage with a cascading waterfall layout that includes a plurality of images from a plurality of webpages included in the search results. The search result webpage may also present, adjacent to the plurality of images, structured data corresponding to a selected image from the plurality of images where the structured data includes details from the webpage to complete the task.

Image reranking and presentation for visual exploration

Techniques for dynamically generating deeplink search results in response to navigational search queries. In an aspect, to address user search queries, a general-purpose search engine is provided in parallel with a dedicated engine for specifically ranking deeplinks. Upon identifying a received query as a navigational query, a parallel query is generated from a common domain and user query, and provided to the dedicated engine. The engine accesses relevant deeplink URL's from a search index, which may be frequently refreshed and updated with the most recent Web contents. Ranking of deeplink URL's may be performed according to an algorithm that processes query-level features and document-level features of URL's to be ranked. In an aspect, the algorithm may be trained from search engine logs and/or Web browser logs, by calculating a Log-based Normalized Discounted Cumulative Gain (LNDCG) designed to quantify relevance of search results to queries based on user click behavior.

Dynamic deeplinks for navigational queries

A dialog system for use in a multi-user, multi-domain environment. The dialog system understands user requests when multiple users are interacting with each other as well as the dialog system. The dialog system uses multi-human conversational context to improve domain detection. Using interactions between multiple users allows the dialog system to better interpret machine directed conversational inputs in multi-user conversational systems. The dialog system employs topic segmentation to chunk conversations for determining context boundaries. Using general topic segmentation methods, as well as the specific domain detector trained with conversational inputs collected by a single user system, allows the dialog system to better determine the relevant context. The use of conversational context helps reduce the domain detection error rate, especially in certain domains, and allows for better interactions with users when the machine addressed turns are not recognized or are ambiguous.

Multi-user, multi-domain dialog system

Growing community of inventors

Bellevue, WA, United States of America

Dong Han Wang