《Mahout in Action》中对content-based推荐技术的评价

本文探讨了基于内容的推荐系统的技术特点及限制,指出这类系统虽然有效但仍存在领域局限性,难以跨领域应用。
部署运行你感兴趣的模型镜像
There’s nothing wrong with these techniques( content-based,本人注 ); on the contrary, they can work quite well. They are  necessarily domain - specific approaches, and would be hard to meaningfully codify into a framework. To  build an effective content -based book recommender, one would have to decide which attributes of a  book  --   page count, author, publisher, color, font  --   are meaningful, and to what degree. None of this  knowledge translates into any other domain; recommending books  this way doesn’t help in recommend  pizza toppings. 

For this reason, Mahout will not have much to say about this sort of recommendation. 


写的不错,对吧~^^

转载于:https://my.oschina.net/liangtee/blog/84212

您可能感兴趣的与本文相关的镜像

Stable-Diffusion-3.5

Stable-Diffusion-3.5

图片生成
Stable-Diffusion

Stable Diffusion 3.5 (SD 3.5) 是由 Stability AI 推出的新一代文本到图像生成模型,相比 3.0 版本,它提升了图像质量、运行速度和硬件效率

评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值