Introduction part
1. Intelligent Fishery
The micro sensor is placed on the fish, which is intended to track fish positions at any given time and to use sonar to detect the fish density and when and where to cast the net.
2. Intelligent Motion
Rugby players have a GPS tracking system that allows the fake-face to know the situation of each player
3. Intelligent Health Care
The patient's physical condition is detected by cloud services such as Ginger.io and applications that can be installed through a consumer phone. Proteus Company produces a size of only a grain of sand "can swallow" sensor, can track the patient's medication time and manner
4. Smart Home
To grow and talk.
5. Intelligent Love
Get the model through Big data analysis, and then match all the people through the model
6. Intelligent Childcare
For example, the latest generation of urine is not wet can be analyzed urine samples to remind parents of the current baby's physical condition
Business Intelligence1. Wal-Mart
For example, if you buy a barbecue-related product from Wal-Mart and your location is within a 3-mile radius of the bureau's Wal-Mart store, and the sun shines, Wal-Mart will send a barbecue detergent coupon to your phone
2. Unstructured data
Research tells us that less than 20% of the information the company actually uses to make decisions. Moreover, these 20% are just traditional structured KPIs or financial data. If the non-institutional data can be used, then the value will be beyond imagination.
Develop an intelligent strategy
When you start out with a strategy, you understand what you need and then decide what data your phone will answer.
Identify what information is needed, what information is available and what is used, and then build the smallest and most direct database in time
One sentence may often be heard "well, it's best not to lose anything, maybe in the future." But this sentence needs to distinguish the situation, for example, in the data, the data is a statement cycle, and let the ever-increasing information to you pressure, it is better to forget the database five years ago, but more importantly, step back, ask yourself what is the most important. Then, use the resulting conclusions to guide action and select data.
Smart Strategy Board in practiceA, the purpose panel
Strategic purpose and ultimately want to make the company what
B, Customer panel
There are two areas to consider----the target market and value proposition, simply to clarify why customers are buying from you.
C, Finance panel
Increase the company's revenue
D, Operation panel
What are the best internal links for fostering core competencies----? (Product/service development, creating needs, meeting needs, regulatory and social relations, etc.)
E, resource panel
What kinds of IT systems and data do you need to accomplish strategic goals, and what kind of people can help with such an IT system and what tools can get such data?
F, Competition and risk panel
You should think about what challenges and risks you will face in implementing your strategy.
Metrics and DataUnstructured and semi-structured data
Represents so it is not easy to insert columns, rows, and fields and data. Examples include: photos and graphic images, video recordings, URLs, text, or posts such as e-mail, PDFs, blogs, social media, and PowerPoint. There are some structures in this data that can be used for analysis, but there is a lack of a rigorous data model structure. For example, posts on Facebook can be categorized by author, data, length, or even emotion, but the content is often unstructured.
A lot of the data is affecting us.
For example, reading data, what we read, how long we read it, whether we skipped it, where we made the comment, and where we highlighted it. This information is useful for both authors and publishers. This allows other authors to adjust the content to shorten or improve the content of a particular section, thus enabling readers to better read the experience. In addition, authors and publishers may identify the interests of readers from the passages in which books are often highlighted, judging the trends of new topics.
For example, after the 2013 Typhoon swept the Philippines, volunteers received millions of photos, text messages and videos on average every day. By finding out exactly where the changes were coming from in the Philippines (using photographs or GPS sensors for text distribution), volunteers made an online map, and when the disaster relief teams came, they didn't have to take the time to know which disaster was the worst.
Data for the sensor
It seems like every book of this kind always comes up ...
The internet of things has created a lot of possibilities for the future.
If an early cancellation occurs, your computer will automatically be set up for a later job, and your coffee machine will be postponed to a new wake-up time. Your refrigerator will know what's inside it, and you won't have to do anything, and your refrigerator will automatically place orders on the web to fill in the inventory.
decomposition of Big data
confirm the data you need to measure (there is a logic level)A, internal structure data, which is easiest to find and most easily analyzed
b, internal semi-structured data
c, Internal unstructured data
D, externally structured data
E, externally unstructured data
For example, can you think about the following data to help you complete your project? A, survey and research data
B, Sales and event data
C, eyeball tracking data
D. Operational data requirements
E. Supply data
F, financial data
G, financial transaction data
H, Stock market data
I, cash flow data
J, recruitment data, job search data
Using data analysis techniquesText analysis methods, typical text analysis methods include
1. Text categorization
Put some structure into the text so that it can be used for analysis and querying, such as spam categorization
2. Text Collection
Text phones allow you to automate a large number of knowledge bases in your phone text and turn them into meaningful quick access information or filtered themes or categories
3. Concept Extraction
It is now used by many law firms in the business world, where hundreds of thousands of legal documents are required to deal with them. The concept extraction method can lead to the price of the target tracking file which is likely to be the most relevant to the new situation, which will save a lot of valuable time to locate the previous data.
4. Emotional analysis
Also become a point-of-view, which strives to extract subjective views or emotions from text, such as sentiment analysis using TWITRRATR tools for Starbucks, for example, sports drinks company Gatorade uses a series of topics to test the conversation of a blog and to show how much of the conversations in the blogosphere are burning. The company also carries out a specific emotional analysis of important topics, products, and campaigns. He also tracks vocabulary related to brands (including competitors, athletes, and the theme of sports nutrition). Basically, Gatorade company knows what people around the world say about their freezing and the company's products.
5. Text Summary
This text analysis tool allows you to use a computer program to automatically summarize the contents of a file to preserve the most important content in the source file. This automatic induction has two methods: extraction and summarization.
Video and image analysis technology1, Google's nametag, the user only need to look at the stranger through the NameTag app in Google Glasses, will immediately get a message containing the names of strangers, occupations and personal information on Facebook exactly match
2. Some companies not only analyze individual images, but also analyze changes through a set of images. For example, you can find that the picture you are uploading now is fat or thin, happy or sad.
Joint Analysis Technology
Combine the various analytical techniques to use.
What does it mean to be liked?
The likes that others give you on Facebook can be used to reveal many of the intimate details, personality traits and preferences associated with you that you can't share with them in other situations. A study at the University of Cambridge and the Microsoft Research Laboratory showed us how patterns like the likes of Facebook can automatically and accurately predict a large number of highly sensitive personal traits. The study eventually showed that the point. Very small or almost no, usually, a praise can produce accurate predictions. Like science and Mozart, you know, you have a high IQ.
After getting your personality traits, you can also do more interesting things, such as the ability to push some goods or books.
At last
When analyzing data, be honest and transparent about the purpose and means of information use---especially for customers
Add value to the people who provide the data, and remember that it is fair and worthwhile to do so.
Visualization of data1, Column chart
2. Line drawing
3. Pie chart
4. Scatter map
many new and old tools can visualize data, but visualization is not a goal. Pretty icons can still be meaningless. The information we package should be able to tell a story. The best way to tell a story can be through the following ways1. Display Map
If you need to show a series of data sets that are cluttered, your story will be more concise and clear if the data is presented in a superimposed manner on the map. You can refer to Polymaps.
2. Display text
The most common emotion display based on word cloud
3. Display data
D3.js can display data in multiple styles
3. Show behavior and emotions
Now we can know where the hotspot of a website is and what shape a song is
4. Display Association
successful data visualization and elements of the infographic1. Determine your target first case
2. Customized Data visualization scheme
3. Visualize a clear label or title for the data
4, strong data visualization and your strategy to link up
5. Choose your display chart wisely
6, use the title to highlight the focus
7, in the appropriate to add text description
priorities and programme of Action1. Better understanding and positioning of customers
2. Improve Business Process optimization
3, improve people's health and happiness level
4. Enhance security and reduce fraud
5, encourage the improvement of enterprise and employee performance
6. Conditions for improving the workplace and other infrastructure
Conclusion1. Don't focus on data from the start. If you do, you will find yourself in a state of being full of choices. The realization is to be aware of the strategy, to figure out what you need to know, why you should know it, and to link these dead ends with your long-term strategy and short-term tactical goals.
2, clear their own goals, you can start to find the data to achieve these goals. If there are three sources of data that can help you solve your problem, focus on the kind of data that you already have or are most likely to get.
3, remember, must be easy to get from the structure of good internal data, relative to the poor structure of external data, the former more easily analyzed.
4, the data will be considered to consider the analysis
5, after the analysis, it is necessary to consider how to show the data
6, finally in determining whether to achieve their own goals to change the contents of the previous steps
Smart Large Data Smart Guidelines (reading notes)
