TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Giving Away Our Recommendation Engine

297 点作者 kky大约 11 年前

7 条评论

mck-大约 11 年前
Fwiw, here are two light-weight feature-based recommendation engines I built for Node.js (for situations where you have the cold-start problem and therefore can&#x27;t rely on user&#x2F;item based collaborative filtering): Alike [1] and Look-Alike [2]<p>[1] <a href="https://github.com/axiomzen/Alike" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;axiomzen&#x2F;Alike</a><p>[2] <a href="https://github.com/axiomzen/Look-Alike" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;axiomzen&#x2F;Look-Alike</a>
评论 #7561829 未加载
contingencies大约 11 年前
So hang on, what exactly is a recommendation engine?<p>They give examples of LinkedIn (<i>people you may know</i>) and Amazon (presumably <i>other people who bought this</i>, <i>so-and-so&#x27;s list of such-a-subject books</i>).<p>That makes sense, though the segment of businesses that may actually benefit seems limited. Social stuff, sure. Most of us? What&#x27;s the minimum recommendable-entity&#x2F;category-or-user threshold that this makes sense for? Is success with these sorts of engines merely a reflector of poor UI design in your normal UX? (Of the above examples, the first seems very unidimensional - in that it&#x27;s basically a simple graph distance - and the latter also rather rudimentary and often irrelevant).<p>So what exactly is this thing providing? Graph analysis? I think not. It reads more like some kind of raw timestamped user behavioural event data processing to infer relationships between users or products they interact with. Reading through the docs it seems this is a layer on top of Apache Pig (<a href="https://pig.apache.org/" rel="nofollow">https:&#x2F;&#x2F;pig.apache.org&#x2F;</a>) - <i>a high-level language for expressing data analysis programs, coupled with infrastructure for evaluating these programs</i>. I think clarity in explaining this thing could be improved, particularly selling clearly what a recommendation is and when its useful. Using phrases like &quot;award winning&quot; doesn&#x27;t help.<p>PS. Why all the downvotes? Sheesh.
评论 #7560754 未加载
评论 #7560546 未加载
评论 #7562258 未加载
olidb2大约 11 年前
FWIW we&#x27;ve been using the mortar platform to run large pig jobs without a fuss at <a href="http://datadog.com" rel="nofollow">http:&#x2F;&#x2F;datadog.com</a> and we&#x27;ve been very happy with it. Glad to see them contribute their recommender code too.
评论 #7560606 未加载
pixelmade大约 11 年前
I&#x27;m curious what the business case was for open sourcing the code. Maybe to create an ecosystem?
评论 #7560092 未加载
评论 #7560239 未加载
dsheth大约 11 年前
Anyone know of any comparisons between this and Apache Mahout? I&#x27;ve used Mahout&#x27;s Item-Item recommender in the past, and it&#x27;s worked well, just wondering if there were advantages to this recommender.
ASquare大约 11 年前
I&#x27;m sure plenty of good karma (even the non-HN kind) is headed your way - kudos.
X4大约 11 年前
WOW, Awesome Documentation and Product!! Kudos and Greetings from Germany 😊<p>Those who know what Hadoop, Pig and the whole &quot;Data Science Stack&quot; is, will find this surely useful.