知识卡片
共评样本少会削弱相似证据
两个用户在极少量共同物品上的评分相近,并不能强力支持他们整体兴趣相同。选择相似度时应同时检查共评数量、缺失处理与评分分布,而不是只取一个接近1的数。不同方法利用的证据范围也不同。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- · 查看原书依据
suffer from few co-ratings and cold-start items [12, 18]. Most users show less interest in providing reviews leads to few co-ratings and low accurate similarity. For this reason, new similarity metrics were proposed to alleviate these drawbacks. Patra [12] presented a BCF metric that considers all ratings of an active user not to depend on co-ratings, unlike traditional methods. In the study [9], authors measured
参考来源
- 位置:《Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications》第08章《协同与内容电影推荐》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications/_epub-src/OEBPS/Text/9789815136746-C8.html)
- 结论依据:摘录中关于共评样本少会削弱相似证据的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:suffer from few co-ratings and cold-start items [12, 18]. Most users show less interest in providing reviews leads to few co-ratings and low accurate similarity. For this reason, new similarity metrics were proposed to alleviate these drawbacks. Patra [12] presented a BCF metric that considers all ratings of an active user not to depend on co-ratings, unlike traditional methods. In the study [9], authors measured