Browse Prior Art Database

Improving Search Results By Analyzing A User's Sentiments When Performing A Search Disclosure Number: IPCOM000240437D
Publication Date: 2015-Jan-30
Document File: 2 page(s) / 23K

Publishing Venue

The Prior Art Database


A method and system is disclosed for improving search results by analyzing a user's sentiments when performing a search.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 2

Improving Search Results By Analyzing A User '


Typically, online search engines process search strings using a proprietary algorithm. If an issue is detected with a user entered search string, an alternate string is proposed to the user, and results for such string are retrieved. However, user reaction to the proposed alternate string and search results is not taken into account.

Disclosed is a method and system for improving search results by analyzing a user's sentiments when performing a search in order to take into account the user's reaction. The method and system receives information about the user's facial expression. Thereafter, the method and system uses Natural Language Processing (NLP) and semantic analysis on a partial search string entered by the user to infer what the user is searching for and to provide alternate or more complete strings.

In accordance with the method and system, as a user logs in, facial recognition is employed to identify the user. Responsive to the user accessing a search application and entering a query, the method and system monitors the user's facial expression. The method and system determines if the user requires any assistance regarding the query based on facial expressions of the user. The method and system continuously monitors facial expressions of the user to determine alternate strings to use for searching. The method and system questions the user to narrow down the scope of search strings. Questioning methods may include check boxes, radio buttons, filters and properly formulating questions to the user. Alternatively, the method and system may perform searches directly on alternate strings and offer results to the user. The method and system detects negative sentiments reflected in the user's facial expression if the results are not of user's expectations. Thereafter, the method and system may propose alternate search strings. The method and system associates search strings with search results and facial expression to determine success of a search. The method and system learns over time common facial expressions of a user to further refine sentiment analysis.

Consider a scenario, where User A launches a search a...