Conversation-Based Search System Using Separated Touch Interface
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines face challenges in accurately understanding users' intentions and providing relevant search results due to variations in query inputs, which can be vague, incomplete, or overly broad, leading to irrelevant search results.
Innovation Solution
A conversation-based search system and method that proposes utterance phrases to users to refine their queries, allowing for a guided conversation to clarify the user's intention and reform the query, ultimately providing a more accurate search result.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a search engine directly processes user queries without additional interaction, then the search process is fast and simple, but the search results may not accurately reflect user intention due to vague or incomplete queries
Solution Approach 1:
The patent introduces a conversation window as an intermediary interface between the user and the search engine. This conversation window facilitates multi-turn interactions where the system can ask clarifying questions and guide users to refine their queries, thereby improving understanding of user intention without requiring fundamental changes to the core search engine architecture
Solution Approach 2:
The system performs preliminary actions by proactively analyzing the initial query and identifying missing information before executing the search. It generates utterance phrases to elicit additional details from users in advance, transforming vague queries into more precise search terms before the actual search is performed, thus improving result accuracy without increasing the complexity of the search algorithm itself
2Measurement precision
If the search system engages in conversation with the user to clarify queries, then the search results become more relevant to user intention, but the search time increases
Solution Approach 1:
The system applies partial action by selectively engaging in conversation only when the initial query is identified as vague, incomplete, or ambiguous. For clear and complete queries, the system proceeds directly to search without additional interaction, thus avoiding unnecessary time delays. The conversation is triggered only when needed to clarify specific aspects of the query
Solution Approach 2:
The conversation window allows users to skip certain clarification steps if they prefer to proceed with their original query. The system can also skip redundant questions based on information already obtained in previous turns, efficiently moving through the clarification process and reducing overall interaction time while still improving query precision
3Ease of operation
If the search system provides multiple utterance phrases to guide users, then the query refinement becomes more effective, but the interface complexity increases
Solution Approach 1:
The search system segments the query refinement process into multiple discrete utterance phrases, each addressing a specific aspect of the query. Instead of presenting all possible clarification options at once, the system divides the interaction into manageable steps, with each utterance phrase focusing on one particular piece of information needed to improve the search query
Solution Approach 2:
The interface dynamically adapts the conversation flow based on user responses. The system adjusts which utterance phrases are presented next based on the information already obtained, making the interface flexible and responsive rather than rigid and predetermined. This dynamic adaptation simplifies the user experience by only presenting relevant follow-up questions
Data Source
AI summary
A conversation based search method includes the steps of: proposing an utterance phrase corresponding to a query input by a user; and reforming the query by means of an answer from the user and offering a search result corresponding to the reformed query, wherein the user answer is presented by the user using a touch-detection surface separated from a user terminal search window in which the user receives the search results.