Graph-Based Interaction Interfaces for Reusable Chat Context
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional chatbots are limited by their reliance on past interaction context, require repetitive interactions for similar outputs, and lack mechanisms to establish connections between different interactions, leading to inefficiencies in generating complex outputs and reusing outputs across multiple instances.
Innovation Solution
A graph-based interaction interface utilizing a generative AI model to establish connections between interactions, allowing context to be passed between nodes and enabling the propagation of updates, facilitating reusable workflows and efficient generation of complex outputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If conventional chatbots use context from past interactions in a current interaction, then the chatbot can maintain conversation coherence, but the scope of the current interaction is limited to that context and cannot include context from previous contents
Solution Approach 1:
The patent transitions from a linear, sequential interaction model to a graph-based multi-dimensional structure where interactions are represented as nodes connected by edges. This allows the system to access context from multiple previous interactions simultaneously, not just the immediate past, thereby expanding interaction scope while maintaining coherence through graph traversal and context aggregation mechanisms.
Solution Approach 2:
The graph-based interaction interface serves multiple functions: it maintains conversation coherence by tracking contextual relationships, enables reuse of interaction patterns across different contexts, and allows flexible navigation between various interaction histories. This multi-functional design resolves the contradiction by making the system adaptable to different interaction scopes while preserving coherence.
2Reliability
If conventional chatbots require repeating a whole interaction to generate similar outputs, then the interaction logic is preserved, but significant time is consumed to generate complex outputs
Solution Approach 1:
The patent segments the monolithic interaction process into discrete, reusable nodes within a graph structure. Each node represents a specific interaction pattern or logic unit that can be independently executed. This segmentation allows the system to reuse proven interaction logic across multiple contexts without re-executing entire interaction sequences, significantly reducing time consumption while maintaining logic consistency through standardized node definitions.
Solution Approach 2:
The system performs preliminary action by pre-defining and storing interaction patterns as reusable nodes in the graph. These nodes encapsulate validated interaction logic that can be directly applied to new situations, eliminating the need to re-derive or re-execute complete interaction sequences. This preliminary structuring of interaction logic enables rapid generation of complex outputs while preserving reliability.
3Reliability
If conventional chatbots require repeating a whole interaction for each complex output, then complete context is maintained, but user productivity is reduced
Solution Approach 1:
The patent implements copying by allowing interaction patterns to be replicated as nodes within the graph structure. Once an interaction pattern is defined and validated, it can be copied and instantiated multiple times across different parts of the graph, each maintaining the complete contextual logic of the original. This copying mechanism enables high productivity through reuse while ensuring context completeness through faithful replication of interaction patterns.
Solution Approach 2:
By segmenting interactions into reusable nodes, the system allows users to build complex workflows by composing smaller validated units rather than recreating entire interaction sequences. This segmentation maintains context completeness within each node while dramatically improving productivity through modular assembly and reuse of interaction components.
4Device complexity
If conventional chatbots have no mechanism to establish connections between different interactions, then the system is simple to implement, but resultant actions cannot depend on previous outputs
Solution Approach 1:
The graph structure serves as an intermediary layer between individual interactions, providing a formal mechanism to establish connections and dependencies. The graph's edge structure explicitly defines relationships between nodes, enabling resultant actions to depend on previous outputs through defined connection paths. This intermediary graph layer adds adaptability while maintaining manageable complexity through standardized connection rules and traversal algorithms.
Solution Approach 2:
The patent adds a structural dimension to the interaction model by organizing interactions as nodes in a graph with explicit connection relationships. This dimensional change from flat sequential processing to structured graph-based organization enables dependency tracking and cross-interaction references while keeping implementation complexity manageable through graph theory fundamentals and standardized algorithms.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method is provided. The method is executed by a connection engine implemented as a computer program within a computing environment. The connection engine executes a chat assistant utilizing generative artificial intelligence (AI) mode to perform the method. The method includes generating a graph-based interaction interface and processing interactions between a user and the chat assistant as nodes within the graph-based interaction interface. The method includes establishing connections between the nodes of the graph-based interaction interface. The connections enable context to be passed between the nodes. The method includes propagates an update to an interaction of a corresponding node as the context to other connected nodes.