Machine Translation

Machine translation (MT) is the use of software to translate text or speech from one language into another without human intervention. It is one of the oldest and most visible applications of artificial intelligence to language, and its history reads like a map of the field's own paradigm shifts: from hand-written rules to statistical learning and, finally, to the neural networks that now power large language models.

The problem is harder than it looks. Translating is not swapping words but preserving meaning, register and context across linguistic systems that rarely map one to one. That difficulty explains why MT has needed three generations of technology to approach human quality.

From rules to statistics

The earliest systems, rule-based machine translation (RBMT), dominated the field's opening decades. Linguists hand-coded bilingual dictionaries and grammatical rules; the output was predictable but rigid, unable to absorb the ambiguity and variation of real language. In the 2000s the paradigm shifted to statistical machine translation (SMT): instead of rules, systems learned from large bilingual corpora which translations were most probable. The word-alignment models developed at IBM and, above all, the phrase-based approach formalized by Philipp Koehn, Franz Och and Daniel Marcu in 2003 set the industry standard for more than a decade. Documentation: original phrase-based translation paper.

The neural era: from the Transformer to LLMs

A representative neural approach is GNMT: an attention-based encoder-decoder learns to generate the target sequence conditioned on the source. It models probabilities; it does not guarantee semantic fidelity.

How it is measured: BLEU and its limits

Scoring a translation automatically is a problem in its own right. The historical benchmark is BLEU, proposed by Kishore Papineni and colleagues in 2002: it compares a system's output against one or more human translations and computes the overlap of n-grams (sequences of one to four words), with a brevity penalty. It is fast, cheap and language-independent, but it measures surface word overlap, not adequacy or meaning: a correct translation that uses different synonyms can score poorly. That is why neural metrics such as COMET (Rei et al., 2020) have gained ground, comparing source, hypothesis and reference in a semantic space and correlating better with human judgment. Documentation: original BLEU paper; original COMET paper.

Real limits

Machine translation does not remove the need for review: register, specialized terminology and fidelity to the source must be checked when errors carry a high cost.