| The case study aims to compare and evaluate the two outputs of HT（human translation）and MT+LPE（machine translation and light post-editing）,taking a chapter of the web novel,Joy of Life（Qing Yu Nian）as an example.The research is divided into 2 parts.First,it collects English readers’ ratings on the two outputs via a questionnaire to see their quality disparity and uses Coh-Metrix 3.0 text analysis to find in what respects the disparity is reflected.Second,it tries to select the better solution（between HT and MT+LPE）to CWN translation given the constraints of 3 factors pre-selectedtranslation quality,speed and budget-via the analytic hierarchy process（AHP）,to explore the value of the MT+LPE mode for translating CWNs.Contrary to conventional wisdom,the results show that the MT+LPE mode scores higher and nearly 40%of the questionnaire respondents think that there is a significant quality disparity between the two translations.Text analysis by Coh-Metrix 3.0 shows that the two versions both have merits.MT+LPE is weaker in narrativity but outperforms in word concreteness,syntax complexity,referential cohesion and LSA cohesion（a measure of semantic coherence）,which are probably the reasons why MT+LPE earns a higher rating.In terms of the mode selection for CWN translation,translation quality has the highest weight among the 3 factors that affect the decisionmaking,significantly higher than the other two.The AHP result shows that MT+LPE is the better solution to CWN translation under the constraints of the factors,winning by an extremely narrow margin.The sensitivity analysis demonstrates that the result is not stable because a tiny rise in the weight of translation quality will change the better solution to HT.Nonetheless,it is shown that the MT+LPE mode is of value,able to compete with human translation and has a huge potential for market application. |
PDF Full Text Request