Translation(more commonly called machine translation) convert text or speech from one language to another, with similar meaning. Some approaches to perform translation are rule based approach, statistical approach or a hybrid approach. Rule based approaches require dictionary of source to target language. Statistical approaches require bilingual corpora such that same text is present in both the languages. Hybrid approaches uses the power of rule based and statistical approach.
Transliteration is the process of converting a word written in one language into another language, phoneme by phoneme. In other words, the letters in the target language replace the ones that most closely correspond to the name in the source language. Enabling transliteration for a search engine allows the site visitors to type a query phonetically in one language and have that query appear in another language.
Text like names, prices, dates, location names can be same in both the languages and these are called named entities. Drawback of translation is that these named entities might be translated if they are not recognized as named entities. In such cases it makes sense to use transliteration. However, people have tried to incorporate transliteration in the process of translation for named entities but did not succeed.