Comment Graph Construction via Semantic Subject-Opinion Pairs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine comment systems rely on conventional information retrieval techniques, resulting in repetitive comments that lack understanding of news content and often fail to recall unpopular news, leading to low recall rates and irrelevant responses due to heavy dependencies on timeliness, quality, and size of comment databases.
Innovation Solution
An artificial intelligence-based method and apparatus for constructing a comment graph that determines a comment text, identifies comment subjects and opinions, generates subject opinion pairs with emotional tendencies, and creates comment labels based on news information, establishing a comment graph with semantically associated nodes to improve comment relevance and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional information retrieval techniques are used to retrieve comments based on news keywords, then the system can quickly retrieve relevant comments, but the comments are repetitive and lack understanding of news content, resulting in low recall rates for unpopular news
Solution Approach 1:
The patent introduces an intermediary mechanism (comment graph with subject-opinion pairs and emotional tendencies) between the news retrieval system and the comment database. This intermediary structures comments semantically rather than relying on simple keyword matching, enabling the system to understand news content and retrieve relevant comments even for unpopular news topics.
Solution Approach 2:
The patent transforms the comment retrieval approach by changing the parameter from keyword-based matching to semantic structure-based matching. By organizing comments into subject-opinion pairs with emotional tendencies and establishing comment graphs, the system shifts from surface-level text matching to deeper semantic understanding, improving recall rates while maintaining retrieval efficiency.
2Adaptability or versatility
If the system lifts restrictions on relevance to provide more diverse comments, then coverage of comment opinions expands, but irrelevant replies are generated, creating heavy dependencies on timeliness, quality and size of comment databases
Solution Approach 1:
The patent applies preliminary action by pre-structuring comments into subject-opinion pairs with emotional tendencies and organizing them into comment graphs before retrieval. This preprocessing creates a semantic framework that guides relevant comment selection, allowing the system to expand coverage without generating irrelevant replies and reducing dependency on database quality.
Solution Approach 2:
The patent replaces the mechanical keyword-matching system with a semantic understanding system based on subject-opinion pairs and emotional tendencies. This substitution enables the system to distinguish relevant from irrelevant comments more effectively, expanding coverage while maintaining relevance without heavy dependency on database quality.
Data Source
AI summary
The present disclosure discloses an artificial intelligence based method and apparatus for constructing a comment graph. A specific embodiment of the method comprises: determining a comment text based on comment data on a network page; identifying a comment subject and a comment opinion in the comment text, based on a characteristic in the comment text; generating a subject opinion pair including the comment subject, the comment opinion and an emotional tendency based on the comment subject and the comment opinion; generating a comment label associated with the subject opinion pair based on news information; and generating a comment graph, based on the comment label and the subject opinion pair. This embodiment improves the pertinence and the accuracy of the comment and the control to the comment emotion, when providing comments externally.