The definitions of the entry may be weighted based on the English version of the example sentences. This could provide additional information to rank search results so that e.g. http://sino.study/terms/language would perhaps rank 语言 at the top rather than 话. Currently, the results are only ranked by word frequency, but there is no way of telling if the frequency of a specific word matches the desired definition or some other definition. By having a weight for each of the definitions an invididual frequency could be adjusted for each of the definitions.
A more immediately implementable solution and perhaps a necessary fallback is to simply divide the word frequency by the number of definitions (or perhaps the number of word classes when that metadata is available).
This feature depends on example sentences.
The definitions of the entry may be weighted based on the English version of the example sentences. This could provide additional information to rank search results so that e.g. http://sino.study/terms/language would perhaps rank 语言 at the top rather than 话. Currently, the results are only ranked by word frequency, but there is no way of telling if the frequency of a specific word matches the desired definition or some other definition. By having a weight for each of the definitions an invididual frequency could be adjusted for each of the definitions.
A more immediately implementable solution and perhaps a necessary fallback is to simply divide the word frequency by the number of definitions (or perhaps the number of word classes when that metadata is available).