Fine-tuning machine translation quality-rating scales for new digital genres: The case of user-generated content

With the active participation of users in product review platforms, online consumer-generated content, and, more specifically, user-generated reviews, have become a clear reference purchasing decision-making processes, which sometimes exceed impact advertising campaigns. A common feature most tourism platforms is use machine translation (MT) systems to immediately make reviews available various languages. However, quality MT output these varies greatly, primarily due subjective and unstructured nature this digital genre. Different studies confirm that there are no universal rating scales. The assessment usually depends on factors such as purpose text or value given immediacy translation. New neural been revolution increase translated output; however, new lines research opening up verify whether paradigm can be assessed with existing scales, mainly from previous rule-based statistical translation, it necessary develop metrics specifically for intelligent systems. On other hand, one questions remain resolved context large amounts textual data training effective less but higher better-adjusted specialty type used. Based hypothesis each genre requires specific work identifies error patterns characteristics user corpus-based approach analysis will contribute adapting scales