Resolving ambiguous queries
The patent describes a system that resolves ambiguous search queries by using historical user data to determine the most likely meaning of the query and rank search results accordingly. This read more
Presentation of Local Results
The patent describes a system that processes search queries by generating both local and non-local results, prioritizing local results based on their relevance to the query. It optimizes the display read more
Enforcing category diversity
The patent outlines a system designed to ensure that search results for local points of interest, like restaurants or museums, show a diverse range of categories. It does this by read more
Query completions
This patent essentially proposes a method for search engines to provide more relevant and personalized query completions to users based on their previous search activities, ultimately aiming to enhance the read more
Systems and methods that match search queries to television subtitles
The technology described matches search queries with TV subtitles to enhance the experience of using second screens like smartphones or tablets while watching television. It detects spikes in search queries read more
Systems and methods for machine-learned prediction of semantic similarity between documents
The system breaks down two documents into smaller text blocks, encodes each block using a machine-learned model, and then compares these encodings to calculate a similarity score, indicating how similar read more
Privacy-sensitive training of user interaction prediction models
This patent describes methods, systems, and computer programs for training a machine learning model to predict user interactions with web content, like text or images, based on search queries. The read more
Query categorization based on image results
The patent describes a method for categorizing search queries based on the images returned in the search results. It involves analyzing user behavior data associated with these images, such as read more
Evaluating an Interpretation for a Search Query
The patent describes a method for evaluating if a person’s understanding of a search query is correct. It trains a computer model with past search queries and human interpretations to read more
Modifying search result ranking based on implicit user feedback
The patent describes a system that improves search result rankings based on how long users view each result. It calculates relevance by comparing the number of longer views to shorter read more