Data Feed Aggregation Using Semantic Embeddings Across Services
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Individuals face the cumbersome and time-consuming task of sifting through multiple data feed services to obtain semantically related information, particularly when using voice user interfaces, due to the habitual nature of seeking information across different sources and the frequent alterations in data presentation.
Innovation Solution
Implementing domain-specific machine learning models to translate between data feed service action spaces and a data feed-agnostic semantic embedding space, allowing for the aggregation of information from multiple data feed services in response to a single natural language query.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If individuals manually navigate multiple data feed services to obtain semantically related information, then they can access diverse perspectives, but the process becomes cumbersome and time-consuming
Solution Approach 1:
The patent combines multiple data feed services into a unified interface that accepts single natural language queries. The system merges data from disparate sources (news websites, social networking data feeds, etc.) and presents aggregated results, eliminating the need for users to manually navigate each service separately while maintaining access to diverse perspectives.
Solution Approach 2:
The patent introduces an intermediary system that acts as a mediator between users and multiple data feed services. This intermediary accepts natural language queries, translates them into service-specific queries, aggregates results from multiple sources, and presents unified answers, thereby reducing user effort and time investment.
2Adaptability or versatility
If data feed services frequently alter data presentation and accessibility, then they can optimize their services, but individual efforts to obtain information are impeded
Solution Approach 1:
The patent creates a universal interface that can interact with multiple different data feed services through a common natural language query mechanism. This universal layer abstracts away service-specific changes in data presentation and accessibility, allowing users to interact with all services through the same ease-of-use interface regardless of individual service alterations.
Solution Approach 2:
The system dynamically adapts to changes in individual data feed services by maintaining flexible query translation capabilities. When services alter their data presentation or accessibility, the intermediary system can adjust its query strategies and aggregation methods accordingly, ensuring continuous ease of operation for users without requiring them to adapt to each service's changes.
3Adaptability or versatility
If users seek information from multiple sources habitually, then they obtain different perspectives, but the process becomes cumbersome particularly with voice user interfaces
Solution Approach 1:
The patent merges multiple information sources and interaction steps into a single unified interface that processes one natural language query at a time. This is particularly beneficial for voice user interfaces, where users can speak a single query to retrieve diverse perspectives from multiple sources without needing to navigate through complex multi-step processes across different services.
Data Source
AI summary
Implementations are described herein for aggregating information responsive to a query from multiple different data feed services using machine learning. In various implementations, NLP may be performed on a natural language input comprising a query for information to generate a data feed-agnostic aggregator embedding (FAAE). A plurality of data feed services may be selected, each having its own data feed service action space that includes actions that are performable to access data via the data feed service. The FAAE may be processed based on domain-specific machine learning models corresponding to the selected data feed services. Each domain-specific machine learning model may translate between a respective data feed service action space and a data feed-agnostic semantic embedding space. Using these models, action(s) may be selected from the data feed service action spaces and performed to aggregate, from the plurality of data feed services, data that is responsive to the query.


