Data-driven learning (DDL) has emerged as an inductive, learner-centred approach in language education—particularly at the tertiary level—sparking renewed interest in how it compares with traditional dictionary consultation. This entry synthesizes empirical studies that directly contrast DDL and dictionary use across four domains: vocabulary learning, writing proficiency, grammar learning, and reading comprehension. The evidence indicates that DDL often produces deeper vocabulary retention and stronger improvements in collocational and written accuracy, whereas dictionaries continue to offer clear benefits for explicit grammar instruction and for supporting lower-proficiency readers. This entry also identifies methodological shortcomings in existing comparisons. Finally, the entry discusses an integrated approach that blends corpus data with lexicographic information in DDL tasks—enhanced by Artificial Intelligence (AI) interfaces—to harness both inductive and deductive pathways for more effective L2 learning.