k*z
发帖数: 4704 | 1 Today, everyone is talking about big data in all kinds of industries. Data
Storage and ETL Process have significantly changed compared to RDBMS. A
bunch of new and fancy words, such as Hadoop, MapReduce, HDFS, Hive and Pig;
run onto our traditional analysts’ heads, who may not understand the
meaning of those words and feel very frustrated and out-of-date.
Fortunately, the core of data analytics still remains the same. The
fundamental business, mathematics and statistics are not changed at all. All
statistical and mathematical tools still work with those "Big Data", such
as SAS, R, etc.
If you are able to survive, succeed at work without knowing Oracle,
Teradata and DB2 while using data on them. You will fail now without knowing
Hadoop? The answer is “No”.
What you need worry about is how you can understand the data itself better
and deeper instead of the platforms the data are on.
Due to the business nature, telecommunication, banking and social media
industries are so obsessed to big data. If you work in those industries and
do not claim you are expertise of big data, they think you are a fool.
Personally, I do not completely agree. I think we can separate “Big Data”
into two stages: IT (Architecture and Infrastructure) & Analytics (business
, mathematics and statistics).
At the end of the day, one of major goals all companies are pursuing is
still revenue. One of important methods to improve revenue is doing
meaningful analytics through business, mathematics and statistics. Buying
many big servers and storing data on Hadoop cannot make more money but
lowering your EBITDA.
For a successful business, a good business idea is still the most
important thing among all factors (e.g.: LVSC starts business at Macau and
Singapore instead of expanding domestically. Doing analytics cannot help
executives make this big move). But understanding business from experiences
and making strategies through analytics can also help business run more
efficiently and better.
Regarding "Big Data", we should outsources all of semi-hardware and
hardware to IT manufactures or 3rd party solution vendors, and let them
worry about infrastructures and other related problems. All of our "old
school" analysts should focus on our own business, come up with useful ideas
and do our financial, mathematical and statistical researches on business
problems. Of course, we do need to learn how to get data out of "big data”. | d***e
发帖数: 793 | 2 who is your audience for the piece? | k*z
发帖数: 4704 | 3 Hotel and Casino Executives | z******4
发帖数: 4716 | 4 你自己写的吗，还是抄的，写的不好，太技术化了
Today, everyone is talking about big data in all kinds of industries. Data
Storage and ETL Process have significantly changed compared to RDBMS (这里改
big data 对于industry的影响，你刚说完就结束了，讲完之后，再提big data cause
tremendous impact for tranditional way of handling data过度).
后面写的比前面好，但最后不要下结论，用建议的口吻
你从听众角度出发，我希望听到什么，什么容易理解，我作为演讲者要达到的目的是什么 |
|
