Parallel Resource Map for Multilingual Search Results
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multilingual searchers often fail to find relevant information in their native language when using search engines, as they may not be aware of parallel resources, such as high-quality translations or original documents in their preferred language, leading to poor search results due to limitations in machine translation and lack of awareness of relevant content in other languages.
Innovation Solution
A system and method that identifies and provides visual alerts for parallel resources in a preferred language within search results, using a resource index module to generate search results and a parallel resource map module to relate resources in one language to their high-quality translations or original documents in another language, based on user preferences or implicit cues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If machine translation is used to translate search results to the user's preferred language, then the user can access information in their native language, but the translation quality is poor and key points are missed
Solution Approach 1:
The patent uses machine translation to create translated copies of search results, allowing users to access information in their native language. While the translation quality may not be perfect, this approach provides a practical solution by generating accessible copies of the original content without requiring manual translation of each result.
2Productivity
If search results are provided only in the query language (English), then the search engine can return comprehensive results, but non-English speakers cannot understand or utilize these results effectively
Solution Approach 1:
The patent introduces machine translation as an intermediary layer between the English search results and non-English speaking users. The translation system acts as a mediator that converts the comprehensive English results into the user's preferred language, enabling both comprehensiveness and understandability.
Solution Approach 2:
The patent changes the language parameter of the search results based on the user's detected or selected language preference. By dynamically adjusting the language parameter through machine translation, the system maintains comprehensive results while adapting them to be understandable for users in different languages.
3Ease of operation
If users search in their native language using cross-lingual information retrieval, then they can find relevant resources, but ambiguities arise and search quality deteriorates
Solution Approach 1:
Instead of translating the user's native language query to English and searching (which causes ambiguity), the patent inverts the approach by keeping the original English queries and translating the results to the user's native language. This avoids the ambiguity of cross-lingual query translation while still enabling native language search experience.
4Loss of time
If automatic translation services are used, then users can quickly access translated content, but the translation quality is inferior to human translation
Solution Approach 1:
The patent employs machine translation as a disposable, automated solution that provides adequate translation quality for search results without requiring expensive or time-consuming human translation. The machine translation serves as a practical, cost-effective intermediary that balances speed and acceptable quality for informational search purposes.
Data Source
AI summary
Various aspects can be implemented for providing parallel resources in search results. In general, one aspect can be a method that includes receiving a search query in a first language. The method also includes generating a search result based on the search query, the search result including a location reference to an identified resource in the first language. The method further includes identifying a parallel resource in a second language, the parallel resource including one or both of a high-quality translation of the identified resource, and an original resource originating in the second language on a similar topic as the identified resource. Other implementations of this aspect include corresponding systems, apparatus, and computer program products.