昨天看到 
Hive account@oflyhigh在介紹SteemData，看著看著，想想拿Steemit的數據做一些分析應該也挺有趣的。於是便提取了過去一星期的數據，拿大家的文章和留言做了文本分析，看看哪些文字在CN版出現得比較頻密。最後製造出了以下的文字雲：
I saw a post from Hive account@oflyhigh introducing SteemData yesterday, and while reading it I thought it may be interesting to do some analytics on Steemit data. So I extracted data on posts and comments with the CN tag in the last week, based on which text analytics was performed to find out what the frequently used words were. The word cloud was thus produced as follows:
（文字雲中出現愈大的字代表出現率愈高）
(the larger the word in the cloud, the more frequent the word was used)
數據是以用了CN作標籤之一的帖子跟以CN為第一標籤的帖子中的留言組成。經過簡單的數據清理，就用Python把文字雲作出來。結果雖然不是很完美（因為中文的斷詞有時不是做得很好，且停止詞的列表還有待完善），但基本上還是OK啦。從文字雲中你觀察到什麼呢？
The data extracted comprises posts with CN being one of the tags and comments with CN being the major tag. After some data cleaning, I used Python to proceed to the creation of word cloud. The result is not perfect, as the Chinese text segmenting is much more complicated than the English one and the stopword list is yet to be completed, but I’d say it’s OK. So what do you observe from the cloud?
我的一些觀察：
Some observations of mine:
- “Chinese”這個字出現最多，其實我是沒有預料到的。回去看一下數據，發現
Hive account@incrediblesnow非常努力到很多帖子提醒作者注意一下CN標籤的用法（很欣賞他的付出！），每個留言就有2個“Chinese”。當然同時還有很多人在用這個字，讓這個字的出現率佔第一位。
The word “Chinese” is the most frequently used, which wasn’t quite expected. I checked on the raw data and found that Hive account@incrediblesnow has been very hard working in advising authors the correct use of the CN tag (which I really appreciate!). The fact that his comments each has 2 “Chinese” and many others are actually using the word makes it top the list.
- “Cnfund”、“novel”跟“chapter”的佔比挺大的，這不意外，
Hive account@cnfund在這個星期出了13章小說。
The proportion of “Cnfund”, “novel” and “chapter” is considerable, which is not surprising considering that Hive account@cnfund has released 13 chapters of his novel in the last week.
- 除了 Hive account@cnfund，
Hive account@ace108、 Hive account@oflyhigh跟 
Hive account@sweetsssj也出現在文字雲裏(大家能放大圖片找找看)，不難理解他們是CN版的名人啊！
Hive account@cnfund aside, Hive account@ace108, Hive account@oflyhigh and Hive account@sweetsssj also have their names in the cloud (you may enlarge the picture and try to find where they’re located). No doubt they are famous in the CN category!
- 「香港」榜上有名！作為香港人看到這還是有種莫名的滿足。
Hong Kong is in the cloud! As a Hong Konger this does offer me some satisfaction.
下一個星期可以再看看大家討論的跟這個星期有什麼分別，希望會有一些有趣的發現。謝謝收看！
I may work on another text analytics next week when the data are ready and see if there will be some significant difference in the most used words between last week and this week. Hopefully there will be some interesting findings. Thanks for reading!
如果你喜歡我的文章，歡迎follow我
Hive account@rayccy！
If you like my post, follow me Hive account@rayccy!
