Local Translation Agent Using Remote Prediction Utility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional computer-aided translation systems face delays and inconveniences due to the time required to process and transmit translation memories over networks, often hindered by insufficient bandwidth or storage capacity, preventing timely access to remote translation memories for translators.
Innovation Solution
A method and system for local, computer-aided translation using remotely-generated translation predictions, where a remote translation memory determines the utility of a translation for a document portion before a request is made, and transmits this translation along with an identification to a local machine, enabling the local machine to generate subsequent document portions through reuse, thus reducing reliance on network transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If translation memory is stored remotely on a server, then storage capacity is sufficient, but network transmission delay increases translation time
Solution Approach 1:
The system performs preliminary actions by sending source text portions to the remote translation memory before the translator actually requests translation. The remote system processes and returns translation predictions in advance, so that when the translator needs the translation, it is already available locally, eliminating network delay during actual translation work.
Solution Approach 2:
The system creates local copies of translation predictions on the translator's machine. Instead of requiring continuous network access to the remote translation memory, the local copy allows the translator to work offline or with minimal network dependency, significantly reducing transmission delays.
2Speed
If translation memory is transmitted to local machine, then access speed increases, but network bandwidth requirements increase
Solution Approach 1:
The system extracts only the necessary translation predictions from the remote translation memory and sends them to the local machine, rather than transmitting the entire translation memory database. This selective extraction minimizes network bandwidth consumption while still providing fast local access to relevant translations.
Solution Approach 2:
The system sends slightly more translation data than immediately needed (excessive action) to preemptively cache translations that may be needed soon. This approach balances network bandwidth usage by sending data in manageable portions while ensuring fast access when translations are actually required.
3Reliability
If translator waits for remote translation memory response, then translation accuracy improves, but productivity decreases
Solution Approach 1:
The system performs translation predictions in advance and sends them to the local machine before the translator completes their work. This preliminary action ensures that accurate translations are ready when needed, eliminating waiting time and maintaining both accuracy and productivity.
Solution Approach 2:
The system implements feedback by monitoring translator behavior and usage patterns, then sending additional translation predictions proactively. This feedback mechanism ensures that accurate translations are available exactly when the translator needs them, improving productivity without sacrificing accuracy.
Data Source
AI summary
A method for local, computer-aided translation using remotely-generated translation predictions includes the step of receiving, by a remote translation memory, at least one portion of a document in a source language for translation into a target language. The remote translation memory determines, prior to receiving a request from a translator for a translation of a second portion of the document, that a translation of the at least one portion is useful in translating the second portion, or in translating a portion of a second document. A local machine receives the translation of the first portion and an identification of the utility of the translation. A translation agent on the local machine generates a translation through reuse of the translation of the first portion of the document, responsive to the received identification of the utility of the translation of the first portion of the document.