Constructing Uyghur Commonsense Knowledge Base by Knowledge Projection

Although considerable effort has been devoted to building commonsense knowledge bases (CKB), it is still not available for many low-resource languages such as Uyghur because of expensive construction cost. Focusing on this issue, we proposed a cross-lingual knowledge-projection method to construct an Uyghur CKB by projecting ConceptNet’s Chinese facts into Uyghur. We used a Chinese–Uyghur bilingual dictionary to get high-quality entity translation in facts and employed a back-translation method to eliminate the entity-translation ambiguity. Moreover, to tackle the inner relation ambiguity in translated facts, we made a hand-crafted rule to convert the structured facts into natural-language phrases and built the Chinese–Uyghur lingual phrases based on the similarity of phrases that corresponded to the bilingual semantic similarity scoring model. Experimental results show that the accuracy of our semantic similarity scoring model reached 94.75% for our task, and they successfully project 55,872 Chinese facts into Uyghur as well as obtain 67,375 Uyghur facts within a very short period.

Translation ambiguity in Mandarin-English bilinguals

Research focused on the cognitive processes surrounding bilingual language representation has revealed the important role that translation ambiguity plays in how languages are stored in memory (Tokowicz & Kroll, 2007). In addition, translation of emotionally related information has been shown to be challenging because a direct translation does not always exist (Basnight-Brown & Altarriba, 2014). The focus of the current study was to explore the processing of ambiguous words for translations that differ in orthography. In Experiment 1, Chinese-English bilinguals translated concrete and abstract words that differed in the number of translations across languages. In Experiment 2, emotion words were introduced into the context, in order to examine differences in emotion translation across languages. The results revealed that words with a single translation were produced faster and more accurately than words that had multiple translations. Finally, translation of emotional stimuli was faster when translating Chinese words as compared to English words.

Determinants of translation ambiguity

Ambiguity in translation is highly prevalent, and has consequences for second-language learning and for bilingual lexical processing. To better understand this phenomenon, the current study compared the determinants of translation ambiguity across four sets of translation norms from English to Spanish, Dutch, German and Hebrew. The number of translations an English word received was correlated across these different languages, and was also correlated with the number of senses the word has in English, demonstrating that translation ambiguity is partially determined by within-language semantic ambiguity. For semantically-ambiguous English words, the probability of the different translations in Spanish and Hebrew was predicted by the meaning-dominance structure in English, beyond the influence of other lexical and semantic factors, for bilinguals translating from their L1, and translating from their L2. These findings are consistent with models postulating direct access to meaning from L2 words for moderately-proficient bilinguals.

Document Representation with Statistical Word Senses in Cross-Lingual Document Clustering

Cross-lingual document clustering is the task of automatically organizing a large collection of multi-lingual documents into a few clusters, depending on their content or topic. It is well known that language barrier and translation ambiguity are two challenging issues for cross-lingual document representation. To this end, we propose to represent cross-lingual documents through statistical word senses, which are automatically discovered from a parallel corpus through a novel cross-lingual word sense induction model and a sense clustering method. In particular, the former consists in a sense-based vector space model and the latter leverages on a sense-based latent Dirichlet allocation. Evaluation on the benchmarking datasets shows that the proposed models outperform two state-of-the-art methods for cross-lingual document clustering.