Memory Network Question Answering with External Search Evidence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current question-answering systems lack the ability to effectively utilize external knowledge to generate accurate answers, as they rely solely on internal databases and fail to leverage external search results efficiently.
Innovation Solution
A memory network-based system that receives search results from a search engine, encodes search evidence and queries into vectors, and decodes these representations to generate answers, utilizing an encoder-decoder framework to dynamically incorporate external knowledge.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If question-answering systems rely solely on internal databases, then system complexity is reduced, but answer accuracy and relevance to external knowledge deteriorates
Solution Approach 1:
The patent combines internal databases with external search engines into a unified question-answering system. The search engine results are integrated with internal database queries, allowing the system to leverage both internal knowledge bases and external web information to generate more accurate and comprehensive answers while maintaining a cohesive system architecture.
Solution Approach 2:
The patent introduces an intermediary processing layer that receives queries from the user interface, coordinates between the internal database and external search engine, and synthesizes results. This intermediary component manages the complexity of interacting multiple information sources without exposing it to the user, thereby improving answer accuracy while containing system complexity within a manageable architecture.
2Loss of information
If question-answering systems incorporate external search engines, then access to external knowledge is improved, but information processing time increases
Solution Approach 1:
The patent implements preliminary action by pre-processing and caching frequently accessed external knowledge from search engines. The system anticipates common queries and pre-loads relevant external information into temporary storage or cache structures, so that when these queries are submitted, the system can retrieve pre-fetched external knowledge without performing time-consuming real-time search operations.
Solution Approach 2:
The patent employs periodic action by implementing batch processing of search queries and updating external knowledge bases at scheduled intervals rather than continuously. The system periodically synchronizes with external search engines to update knowledge caches, and uses these updated caches to answer user queries, thereby reducing the frequency of direct external searches and minimizing processing time for individual user interactions.
3Measurement precision
If search results are extensively analyzed to generate search evidence, then answer quality is improved, but computational complexity increases
Solution Approach 1:
The patent applies partial action by selectively analyzing only the most relevant portions of search results rather than processing entire documents. The system extracts and analyzes specific text snippets, titles, and key segments from search results that are directly relevant to the query, ignoring extraneous information. This partial analysis approach maintains high answer quality by focusing computational resources on the most informative elements while reducing overall computational complexity.
Solution Approach 2:
The patent segments the analysis process into distinct stages: first extracting titles and text snippets from search results, then analyzing these segmented elements to generate search evidence, and finally synthesizing the evidence into answers. This segmentation allows the system to process information in manageable chunks, applying different analysis techniques to different types of content (titles vs. snippets), thereby improving answer quality through thorough analysis while controlling computational complexity through structured processing.
Data Source
AI summary
A computer-implemented method for utilizing external knowledge and memory networks in a question-answering system includes receiving, from a search engine of a question-answering system, one or more search results based on a search query associated with a question submitted via a user interface associated with a computing device, analyzing the one or more search results to generate search evidence as a source of external knowledge for generating an answer to the question, the search evidence including one or more titles and one or more corresponding text snippets, encoding the search evidence and the search query to generate vectors stored in a memory network, obtaining a final vector representation based on the encoding, and decoding the final vector representation to obtain the answer to the question.