今年新生开学在即，南京大学推出了一个新的宿舍分配方案。通过校园迎新网进行问卷调查，学校统计了新生的生活习惯，兴趣爱好等数据，然后通过大数据“推荐算法”，量化评估新生各项数据之间的相似度，为其分配志趣相投的舍友，从而帮助新生更好的适应大学新生活。
Chinese college says it will let algorithms assign roommates based on interests and habits
Nanjing University's new system cheered by netizens, but can you trust college students?
In the West, college dorm rooms are mostly shared by just two students. But in China, most dorm rooms are crammed[1] with 4 to 6 people, assigned only by their student numbers.
[1]cram: If people cram into a place or vehicle or cram a place or vehicle, so many of them enter it at one time that it is completely full. 挤满; 塞进
We crammed into my car and set off.
我们挤进我的汽车，出发了。
此外，cram还有“临时抱佛脚，突击准备 (考试)”的意思，英文解释为“If you are cramming for an examination, you are learning as much as possible in a short time just before you take the examination.”
She was cramming for her Economics exam.
她正为了应付经济学考试而临时抱佛脚。”
It means your odds of[2] sharing living space with roommates you don't get along with for four years is extremely high. Some of them may not want to turn on air conditioning in a 30˚C summer because they think it's bad for their health. (This may or may not be from personal experience.)
[2]the odds of ...的可能性，发生的几率
That's why netizens in China loved today's announcement from Nanjing University, who says they have a solution to the roommate nightmare.
The university says it will use “recommendation algorithms[推荐算法]” for new enrollments this year, assigning them with roommates with similar interests and living habits, a local newspaper reports.
问卷中，不仅有“作息时间”、“空调使用习惯”、“个人卫生习惯”、“共用物品和消费倾向”等调查选项，还有“兴趣爱好”一栏。
The newspaper's Weibo post drew more than 17,000 likes and thousands of comments, with most users praising the university and taking the chance to complain about their own horrible roommates.
The university explained in the report that it created a recommendation system similar to what's used on NetEase Music, a popular Spotify-like music streaming app in China. It says results will be based on online questionnaires filled out by the new students about their daily routine, air conditioner habits and personal hygiene[3].
Li Hao, a teacher in charge of enrollment at the university, said that the university will adopt the latent factor model[隐语义模型] algorithm to deal with the big data, and this algorithm will help match students with similar mindsets, instead of random choices.
“类似于网易云音乐的推荐算法，通过‘隐语义模型’，我们可以通过潜在特征联系新生和兴趣。”南大学工处招办主任李浩说，“即使这名新生并没有接触过某些兴趣爱好，我们也能根据他和其他同学填写的问卷，通过算法挖掘出这名同学与这些兴趣的潜在关联，从而可以量化评估新生之间的兴趣爱好相似度，就有更大的可能为他找到志趣相投的室友。”
[3]hygiene /'haɪdʒiːn,ˋhaɪdʒin/ the practice of keeping yourself and the things around you clean in order to prevent diseases 卫生；保健
the importance of personal hygiene 个人卫生的重要性
Recommender systems are a common use of machine learning[机器学习] technologies -- it's how Amazon recommends more products to you, and also how Netflix figures out what you're likely to watch next.
But users on Weibo point out that the data in this system might not be trustworthy, because it relies entirely on students being honest.
As one of the most liked user comments says, “who would say on the questionnaire that they don't have good personal hygiene?”
“谁会在问卷中把个人卫生习惯不好真实填上去...”好像很有道理的样子？
来源:AbacusNews
- ◆ -
网友：今后你的室友很可能是你的“情敌”?!
你觉得呢？