Intent-Oriented Browsing via ML Prediction Server
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users, especially those with visual impairments, face challenges in navigating websites due to inconsistent architectures and visual structures, leading to unnecessary work and difficulty in completing tasks, as existing technologies focus more on accessibility rather than supporting high-level user goals.
Innovation Solution
A system and method for intent-oriented browsing that uses machine learning and crowdsourcing to direct users to desired webpages across different websites by allowing users to enter their browsing intent, with an intent repository, prediction server, and user interface, which predicts target hyperlinks and provides real-time human help.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional browsing navigation is used on websites with inconsistent architectures, then users can access web pages, but users have to browse through multiple pages and take additional steps to achieve their objectives
Solution Approach 1:
The patent introduces an intermediary system consisting of a prediction server, intent repository, and browser extension that acts as a mediator between the user and website navigation. This intermediary analyzes the current page, predicts target pages based on user intent, and directly navigates users to destination pages, eliminating the need for manual browsing through multiple intermediate pages.
Solution Approach 2:
The system performs preliminary actions by pre-analyzing website structures and pre-predicting navigation paths before users actually need to navigate. The prediction server maintains a repository of learned navigation patterns and uses them to proactively suggest and execute navigation decisions, so users don't have to manually explore multiple pages to find their destination.
2Adaptability or versatility
If websites maintain different architectures and visual structures, then websites can have unique designs, but users face unexpected distractions and information overload
Solution Approach 1:
The patent segments the browsing process into distinct components: intent detection, prediction, and navigation execution. The browser extension separates concern for website-specific architectures from user goal-oriented navigation by isolating the prediction logic in a separate server component, allowing each website to maintain its unique structure while providing consistent user experience across different sites.
Solution Approach 2:
The prediction server implements a universal navigation system that works across multiple websites with different architectures. The system learns and adapts to various website structures while providing a consistent intent-based navigation interface, making it universally applicable to e-commerce, news, social media, and other website types.
3Reliability
If users with visual impairments exhaustively scan pages for navigation, then they can find target pages, but they experience information overload and navigation difficulty
Solution Approach 1:
The system enables self-service navigation by automatically analyzing the current page context and predicting the target page based on user intent. For users with visual impairments, the browser extension autonomously performs what would otherwise require exhaustive manual scanning, reducing navigation complexity while maintaining reliable task completion through automated page analysis and prediction.
4Loss of information
If personal assistants provide search results, then users can see available information, but users have to click through results rather than being taken directly to their goal
Solution Approach 1:
The system performs preliminary navigation actions by predicting and executing the path to the target page before the user even sees search results. Instead of presenting multiple search results and waiting for user selection, the system proactively navigates directly to the most relevant destination based on user intent, eliminating the clicking-through process while maintaining information availability through context-aware prediction.
Data Source
AI summary
A system that provides intent-oriented browsing powered by machine learning and crowdsourcing. The system allows users to enter their intents, which are then assigned to target pages via supervised learning models based on hyperlinks and contributions made by other users. The system has a prediction server that is programmed to receive hyperlinks from a website and return target hyperlinks based on known intent, a user interface for inputting user intent, and a browser programmed to connect to the intent repository and to the prediction server via a user script. The list of supported intents can grow over time based on correct page marks for intent-page mappings as well as via continuous training of machine learning models.