Adaptive multi-channel content selection with behavior-aware query analysis

• US 8,751,481 B2
• Filed: 04/16/2008
• Issued: 06/10/2014
• Est. Priority Date: 04/16/2008
• Status: Expired due to Fees

• Current Assignee: IAC Search & Media Incorporated (IAC/InterActiveCorp)
• Original Assignee: IAC Search & Media Incorporated (IAC/InterActiveCorp)
• Inventors: Camoglu, Orhan, Chrabakh, Wahid, Ding, Lei, Miao, Jingheo, Agrawal, Divyakant, Bhatnagar, Neeraj, Yang, Tao
• Primary Examiner(s): Hoang, Son T
• Application Number:
  US12/148,190
• Publication Number:
  US 20090265325A1
• Time in Patent Office: 2,246 Days
• Field of Search: 707/722, 707/713, 707/736
• US Class Current: 707/713
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

Preferred embodiments of the invention include systems and methods for selecting content from a plurality of channels in response to a received query are described. The invention embodiments include systems and methods that optimize content selection based on received queries and an identification of the user together with analysis of historical data. Embodiments of the invention also include methods for storing historical data including a plurality of queries and user data from a plurality of user databases, analyzing the stored historical data to classify the relevancy of a channel to each of the plurality of queries and user data, storing the classification, and displaying content to user based on such factors as user habits, queries, timing of searches, and content preferences.

Claims

1. A method implemented by at least a computer system for adaptive multi-channel content selection with behavior-aware query analysis comprising:

receiving a search query from a user;

determining, for each of a plurality of channels wherein each channel has a plurality of web results, a probability that the user will select a web result of the channel for the search query;

making at least one selection decision among the plurality of channels based on corresponding probabilities that the user will select a respective web result of at least a respective channel based on historical data of the user, wherein

the historical data comprises past click results and access time range associated with the query, and

the selection decision comprises at least selecting channel placement of the at least respective channel for presentation to the user;

determining an optimal number of web results to display for each of the plurality of channels based at least in part on the corresponding probabilities that the user will select the respective web result of the at least respective channel based on the historical data for each channel and the search query; and

presenting a number of web results to the user, a wherein the number of web results displayed for each of the plurality of channels being based on the optimal number determined for each of the plurality of channels.

2. The method of claim 1, wherein the time range comprises an event.

3. The method of claim 1, further comprising:

storing historical data of queries and selections; and

analyzing the historical data based on the search query and the user to determine the probability for each channel.

4. The method of claim 1, further comprising:

identifying query concepts from the search query;

analyzing the historical data based on each of the query concepts to determine a recommended channel selection; and

aggregating the recommended channel selection for each of the query concepts.

5. The method of claim 1, wherein the selection decision comprises dynamically determining a recommended channel selection.

6. The method of claim 1, further comprising:

if the historical data cannot be analyzed then the determining the probability for each channel is implemented by;

partitioning the received search query to identify concept blocks;

classifying each concept block to determine a classification decision;

aggregating the classification decision of each concept block; and

computing a channel relevancy for the received search query.

7. The method of claim 1, further comprising:

tracking and mining historical user data associated with a plurality of data channels;

tracking and mining relevancy information appropriate of each data channel based on historical data of each user;

tracking and mining relevancy information appropriate of each data channel for a given query and user or group of queries and users;

tracking and mining click paths among the plurality of channels;

classifying query-channel relevancy, user past interests, and channel usage patterns;

collecting data for instant recommendation of channel sources based on classification results; and

optimizing channel placement for best retention and pick rates for targeted.

8. A computer implemented adaptive multi-channel content selection system comprising:

a computer memory for storing a set of instructions;

a computer processor connected to the memory;

a web search engine module stored in the computer memory and executable by the computer processor to receive a search query from a user and present web results corresponding to the search query to the user; and

a channel selection system module stored in the computer memory and executable by the computer processor to;

(i) determine, for each of a plurality of channels wherein each channel has a plurality of web results, a probability that the user will select a web result of the channel for the search query,

(ii) make at least one selection decision among the plurality of channels based on corresponding probabilities that the user will select a respective web result of at least a respective channel based on historical data of the user, wherein

the historical data comprises past click results and access time range associated with the query, and

the selection decision comprises at least selecting channel placement of the at least respective channel for presentation to the user and

(iii) determine an optimal number of the web results to display for each of the plurality of channels based at least in part on the corresponding probabilities that the user will select the respective web result of the at least respective channel based on the historical data for each channel and the search query, and

(iv) display a number of the web results to the user, wherein the number being based on the optimal number determined for each of the plurality of channels.

9. The system of claim 8, further comprising a plurality of databases located in the computer memory, the plurality of databases coupled with the web search engine module.

10. The system of claim 8, further comprising:

a plurality of database records consisting of multi-channel content;

a user-centric database, consisting of user centric click records, the click records used as training data to develop independent classifiers;

a query-channel relevancy classifier configured to analyze categorization of queries into groups and identify a relevancy degree of each channel'"'"'s content to a the search query or class of queries;

a channel-path classifier configured to determine an access path from one channel to another;

a user-bias classifier configured to analyze user search habits pertaining to the multi-channel content; and

a recommendations aggregator configured to collect database records from multi-channel content sources based on the classification results from the query-channel classifier, channel-path classifier and user-bias classifier.