Dynamically filtering search results using contextual user interaction data

Patent metadata

Field	Value
Publication number	US-12411898-B2
Application number	US-202318173778-A
Country	US
Kind code	B2
Filing date	Feb 23, 2023
Priority date	Feb 23, 2023
Publication date	Sep 9, 2025
Grant date	Sep 9, 2025

Abstract

A computerized method filters search result content using user interaction data. Search result content, including search result entries, is presented. User interaction data is received that is indicative of a user's interactions with the presented search result content and a result filter rule is determined using the user interaction data. Filtered search result content is generated using the determined result filter rule and the generated filtered search result content, including a portion of the search result entries, is presented. In some examples, the user is prompted to accept the filtering of the search result content prior to the generation of the filtered search result content. Thus, the search result content is dynamically filtered based on the user's interactions with that content while the user reviews the content.

First claim

What is claimed is:

1. A system comprising:
a processor of a user device;
a user interface (UI) of the user device; and
a memory comprising computer program code, the memory and the computer program code configured to cause the processor to:
present a portion of search result content on the UI, the search result content including search result entries;
receive user interaction data indicative of an interaction of a user with the presented search result content, the user interaction data including scrolling data;
generate a result filter rule using the scrolling data, wherein the result filter rule uses the scrolling data to remove at least one of the search result entries from the search result content;
provide a rules prompt using the result filter rule on the UI;
receive a rules prompt response to the provided rules prompt;
generate filtered search result content using the generated result filter rule and based on the received rules prompt response indicating acceptance of the generated result filter rule; and
present the generated filtered search result content, wherein the filtered search result content includes another portion of the search result entries.

2. The system of claim 1 , wherein generating the result filter rule includes:
determining a likely user intent using the user interaction data, wherein the likely user intent indicates that the user associated with the user interaction data is likely not interested in a class of search result entries of the search result content; and
generating the result filter rule to filter out the class of search result entries at which the determined likely user intent is directed.

3. The system of claim 2 , wherein determining the likely user intent using the user interaction data includes determining the likely user intent using a relevance model that has been trained using machine learning techniques.

4. The system of claim 1 , wherein the memory and the computer program code configured to, with the processor, further cause the processor to generate the result filter rule using user context data, wherein the user context data includes context data from at least one of a user profile of the user from whom the user interaction data is received and a past search of the user from whom the user interaction data is received.

5. The system of claim 1 , wherein the result filter rule filters the search result content based on at least one of the following: search result entry type, search result entry author, and date.

6. The system of claim 1 , wherein the memory and the computer program code configured to, with the processor, further cause the processor to:
determine that the user lacks interest in a currently presented portion of search result entries using the user interaction data;
identify a group of similar search result entries in the search result content that have not been presented yet;
provide an entry group prompt using the identified group of similar search result entries on the UI;
receive an entry group prompt response to the provided entry group prompt; and
present the group of similar search result entries on the UI based on the entry group prompt response indicating that the user wants to see the group of similar search result entries.

7. The system of claim 1 , wherein the scrolling data includes data indicative of at least one of:
when the user scrolls through at least one of the search result entries;
when the user scrolls past at least one of the search result entries;
when the user stops scrolling and the search result entries that are currently displayed on the UI;
how the search result content is being scrolled through;
the search result entries that are displayed as scrolling occurs; or
the search result entries that are displayed when scrolling stops.

8. The system of claim 1 , wherein generating the result filter rule using the scrolling data comprises generating the result filter rule based on the user having scrolled past at least one of the search result entries.

9. The system of claim 1 , wherein generating the result filter rule using the scrolling data comprises generating the result filter rule based on the user having scrolled past at least two search result entries that share at least one of a feature or a property.

10. The system of claim 1 , wherein generating the result filter rule using the scrolling data comprises inferring that the user is not interested in at least two search result entries that share at least one of a feature or property.

11. A computerized method comprising:
presenting search result content;
receiving user interaction data indicative of a user interaction with the presented search result content, the user interaction data including scrolling data;
generating a result filter rule using the scrolling data, wherein the result filter rule uses the scrolling data to remove at least one search result entry from the search result content;
filtering the presented search result content using the generated result filter rule; and
presenting the filtered search result content.

12. The computerized method of claim 11 , wherein generating the result filter rule includes:
determining a likely user intent using the user interaction data, wherein the likely user intent indicates that a user associated with the user interaction data is likely not interested in a class of search result entries of the search result content; and
generating the result filter rule to filter out the class of search result entries at which the determined likely user intent is directed.

13. The computerized method of claim 12 , wherein determining the likely user intent using the user interaction data includes determining the likely user intent using a relevance model that has been trained using machine learning techniques.

14. The computerized method of claim 11 , further comprising generating the result filter rule using user context data, wherein the user context data includes context data from at least one of a user profile of a user from whom the user interaction data is received and a past search of the user from whom the user interaction data is received.

15. The computerized method of claim 11 , wherein presenting the filtered search result content includes maintaining search result entries of the search result content that have already been presented and presenting search result entries of the filtered search result content when a new batch of search result entries is to be presented.

16. The computerized method of claim 11 , wherein the user interaction data includes at least one of the following: entry selection data, entry hiding data, viewed time data, and page selection data.

17. A computer storage medium has computer-executable instructions that, upon execution by a processor, cause the processor to at least:
cause search result content including search result entries to be presented;
receive user interaction data indicative of a user interaction with the presented search result content, the user interaction data including scrolling data;
determine a result filter rule using the scrolling data, wherein the result filter rule uses the scrolling data to remove at least one of the search result entries from the search result content;
obtain filtered search result content using the determined result filter rule; and
cause the filtered search result content to be presented, wherein the filtered search result content includes a portion of the search result entries.

18. The computer storage medium of claim 17 , wherein determining the resu

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Olsen-Kristiansen Bernt Ivar
• Meling Jon
• Lilleby Bjørnstein
• Valvåg Steffen Viken

Classifications

• H04L67/535

  Tracking the activity of the user (network monitoring arrangements H04L43/00; recording of computer activity G06F11/34) · CPC title

• G06F16/954

  Navigation, e.g. using categorised browsing · CPC title

• G06F16/9535 Primary

  Search customisation based on user profiles and personalisation · CPC title

• G06F16/9538 Primary

  Presentation of query results · CPC title