Table 3 - uploaded by Tariq Khan
Content may be subject to copyright.
Results for English-Telugu PSS
Source publication
The paper describes an initial attempt to use Levin's verb classes for the task of preposition sense selection in English to Indian language machine translation. Two language pairs have been selected to describe the approach, English-Hindi and English- Telugu. We exploit the correspondence of verb class's semantics vis-à-vis some prepositions it ta...
Contexts in source publication
Context 1
... get this information from the English language analyzer of the Shakti MT system 5 . To make sure that the mistakes made by the analyzer do not affect our system, we manually correct the errors of attachment, etc. Table 3 shows the result, for all the verb classes we get very high accuracy with an average overall performance of 93.6%. The results point towards the robustness of their use. ...
Context 2
... get this information from the English language analyzer of the Shakti MT system 5 . To make sure that the mistakes made by the analyzer do not affect our system, we manually correct the errors of attachment, etc. Table 3 shows the result, for all the verb classes we get very high accuracy with an average overall performance of 93.6%. The results point towards the robustness of their use. ...
Similar publications
Obtaining meaningful quality scores for machine translation systems through human evaluation remains a challenge given the high variability between human evaluators, partly due to subjective expectations for translation quality for different language pairs. We propose a new metric called XSTS that is more focused on semantic equivalence and a cross...
In recent years, non-autoregressive machine translation has attracted many researchers’ attentions. Non-autoregressive translation (NAT) achieves faster decoding speed at the cost of translation accuracy compared with autoregressive translation (AT). Since NAT and AT models have similar architecture, a natural idea is to use AT task assisting NAT t...
A probabilistic method is used for word sense disambiguation where the features taken are the surrounding six words. As their surface forms are used, no syntac- tic or semantic analysis is required. De- spite its simplicity, this method is able to disambiguate the noun interest accurately. Using the common data set of (Bruce & Wiebe 94), we have ob...
English machine translation is a natural language processing research direction that has important scientific research value and practical value in the current artificial intelligence boom. The variability of language, the limited ability to express semantic information, and the lack of parallel corpus resources all limit the usefulness and popular...
Large Language Models (LLMs) have shown to be effective evaluators across various domains such as machine translations or the scientific domain. Current LLM-as-a-Judge approaches rely mostly on individual assessments or a single round of pairwise assessments, preventing the judge LLM from developing a global ranking perspective. To address this, we...
Citations
The preposition sense disambiguation is a critical task in any reliable Machine Translation (MT) system. The pervasive use of preposition or its equivalent in most of the languages makes it a crucial element during translation. Unlike English, there is no concept of preposition in Kannada. English prepositions are translated to Kannada by attaching appropriate inflections to the head noun of the prepositional phrase. Further post-positional words may also appear in Kannada translation for some prepositions. The choice of the appropriate post-positional word depends on the WordNet synset information of the head noun. The paper proposes an algorithm to disambiguate sense of a simple preposition in English to Kannada MT. It uses properties of the head noun and complement of the preposition for disambiguation. To the best of our knowledge, this is the first attempt towards introducing an algorithm to disambiguate sense of the preposition during English to Kannada MT. Experiments were conducted and the result obtained has been described. The performance of an algorithm is proved to be reliable and scalable.
