Machine translation (MT) is automatic translation through a software program without human assistance. It has recently been at the center of controversies regarding reliability and quality, but few people know that there are actually different types of machine translation: statistical, rule-based, and hybrid.
Statistical Machine Translation (SMT)
SMT uses statistical models derived from the analysis of bilingual text corpora, usually finding a correspondence between one word and another. This is, for example, what Google Translate does. SMT tools might help you to understand the basic meaning of a text, but they will not produce good quality translation.
Rule-based Machine Translation (RBMT)
RBMT, as indicated by its name, is based on grammar rules. The system generates sentences in a target language after analyzing grammatically both the source and the target languages involved. This is, for example, what Systran does. These software programs usually require tedious proofreading, and as they need a lot of lexicons, they only become efficient after long term use.
Hybrid Machine Translation
Hybrid MT is a mix of SMT and RBMT that also includes a translation memory and thus produces translations of a better quality. This tool can be used by translators to avoid translating the exact same sentence twice.
However, linguists still doubt that this tool can help achieve higher quality translations or gain time because of the long editing part. This is why translation needs are still addressed by translators.