ROUGE (Recall-Oriented Understudy for Gisting Evaluation) is a family of metrics for comparing automatically generated texts – especially summaries – with reference texts written by people. It measures how many words or word sequences overlap, for example single words in ROUGE-1 or the longest common subsequence in ROUGE-L. The score is quick to compute, but it does not reliably capture paraphrases with the same meaning or factual accuracy.
The BLEU score (Bilingual Evaluation Understudy) is a metric developed at IBM in 2002 to evaluate machine translations automatically. It measures how many word sequences in a translation also appear in one or more human reference translations and gives a value between 0 and 1 (often shown as 0–100). BLEU is fast and comparable but captures neither meaning nor style well – good paraphrases can score poorly.
In automatic text summarisation, AI condenses long texts to their key points. Extractive methods copy important sentences verbatim, while abstractive methods rephrase the content; large language models work mainly abstractively. This is useful for minutes, reports or long email threads; for important decisions the summary should be checked against the original.
In AI, benchmarking means systematically comparing models or systems using defined tasks and criteria. Public benchmarks measure things like general knowledge, reasoning or coding ability; for practical use, however, test cases from your own day-to-day work matter most. Typical criteria are accuracy, completeness, response time and cost. This helps decide whether a smaller, cheaper model is good enough for a task.
Natural language generation (NLG) is the area of language processing that produces understandable text from data, bullet points or other inputs. Earlier NLG systems relied on templates and rules; today most use large language models. Typical uses include weather and stock market reports, product descriptions, summaries of key figures and chatbot answers. It is important that figures and facts from the inputs are carried over correctly and unchanged.
