CMM
由美国卡内基梅隆大学软件工程研究所于1980年代提出的软件开发能力分级评估框架，将组织能力分为五个等级，是全球软件工程领域最权威的能力评估标准之一
Timeline (last 90 days)
IDM performs global alignment in the item space, while CMM performs local alignment in the code space; the two objectives are complementary and together constitute a complete distillation framework.
Experiments show that IDM improves Recall@10 by an average of 15.39%, and CMM improves NDCG@10 by an average of 8.28%.
IDM is better at improving recall, indicating that reconstructing the item distribution helps the model cover more relevant items; CMM performs better on ranking quality (NDCG), indicating that code marginal matching can optimize ranking more finely.
All Facts (3)
IDM performs global alignment in the item space, while CMM performs local alignment in the code space; the two objectives are complementary and together constitute a complete distillation framework.
50%UnverifiedExperiments show that IDM improves Recall@10 by an average of 15.39%, and CMM improves NDCG@10 by an average of 8.28%.
50%UnverifiedIDM is better at improving recall, indicating that reconstructing the item distribution helps the model cover more relevant items; CMM performs better on ranking quality (NDCG), indicating that code marginal matching can optimize ranking more finely.
50%