Big data cases 41-49: Monitoring illegal alterations, Mustard index

41 Monitoring illegal alterations

In any country, "private construction" is a troublesome thing and is prone to fire. Buildings that are illegally in the house are more likely to catch fire than other buildings. The City of New York receives 25,000 complaints each year about overcrowding, but there are only 200 inspectors in the city who handle complaints. An analyst team at the Mayor's Office feels that big data can help address this gap between demand and resources. The team established a database of all the 900,000 buildings in the city and added data collected by 19 departments in the city: tax detainment records, water and electricity use anomalies, payment arrears, service cuts, ambulance use, Local crime rates, rodent complaints, etc. Next, they compared this database with the fire record of buildings ranked by severity in the past five years, hoping to find relevance. Sure enough, the type of building and the year of construction are fire-related factors. However, one unexpected result is that there is a correlation between the buildings that have access to the external brick wall construction permit and the lower incidence of serious fires.

Using all of this data, the team built a system that would help them determine which housing congestion complaints require urgent processing. The various characteristic data of the buildings they recorded are not the cause of the fire, but these data are related to the increase or decrease of fire hazards. This knowledge proved to be extremely valuable: in the past, the proportion of house patrols that issued homes was only 13%, and after adopting the new method, the proportion rose to 70%.

42 Mustard index

The officials of the Planning Department of the National Development and Reform Commission responsible for drafting the National Plan for Promoting the Healthy Development of Urbanization (2011-2020) (hereinafter referred to as the "Urbanization Plan") need to know precisely the population flow, and how to count these floating populations becomes a problem.

Mustard is a low-quality consumable, and income growth has little effect on the consumption of mustard. Under normal circumstances, the resident population of the city is basically constant for the consumption of instant food such as instant noodles and mustard. The change in sales volume is mainly caused by the floating population.

According to the officials of the National Development and Reform Commission, the sales share of Fuling mustard in various regions of the country has been reflected in the past few years, which can reflect the trend of population movement. A macroeconomic indicator called the “sucker index” was born. Officials from the Planning Department of the National Development and Reform Commission found that the sales share of Fuling mustard in southern China fell from 49% in 2007, 48% in 2008, 47.58% in 2009, and 38.50% in 2010 to 29.99% in 2011. This data shows that the population outflow rate in South China is very fast. They divided the country into two parts: the population inflow area and the population outflow area based on the “salted vegetable index”. The different population structures of the two areas will be different in policy formulation.

43 Weather bill

As the saying goes, "Is the weather unpredictable," has encountered the situation of going out to travel, important outdoor roadshows, holding weddings, etc., but was badly weathered and even caused economic losses? The world's first meteorological insurance company " The Weather Bill provides users with a variety of climate guarantees. The customer logs into the "weather bill" company website and then gives a range of temperatures or rains that they do not want to encounter during a particular time period. The Weather Bills website will query the weather forecast for the customer's designated area within 100 milliseconds, as well as the weather data for the past 30 years recorded by the US National Weather Service. By calculating the weather data, the website will give the price of the policy as the insurer. This service is not only for individual users, but some companies, such as travel agencies, are also happy to participate.

A global beverage company integrates daily weather forecast information from external partners into its requirements and inventory planning process. By analyzing three data points for temperature, precipitation and sunshine hours on a given day, the company reduced inventory in a key market in Europe while increasing forecast accuracy by approximately 5%.

44 Historical scene reproduction

Researchers at Microsoft and Technion have developed a software that predicts when and where infectious or other social problems may break out based on articles from the New York Times over the past 20 years and other online data.

The performance of the system was amazing when tested with historical data. For example, according to a 2006 report on drought in Angola, the system predicts that cholera is likely to occur in Angola. This is due to the fact that, through previous incidents, the system learned that the likelihood of cholera outbreaks will increase after several years of drought. In addition, the system once again warned of cholera in Angola based on reports of major hurricanes in Africa in early 2007. And less than a week later, reports showed that cholera did occur in Angola. In other tests, such as predictions of illness, violence, and casualties, the system's accuracy rate is 70% to 90%.

The system's information comes from the archives of the New York Times in the past 22 years, from 1986 to 2007. However, the system also uses other data on the network to understand what events pose specific social problems. These sources of information provide content that does not exist in news articles but are valuable, helping to determine causal relationships or contexts between different events. For example, the system can infer the relationship between events in Rwanda and Angola, because both countries are located in Africa, have similar GDP, and other factors are similar. According to this approach, the system believes that in predicting cholera outbreaks, the location of the country or city should be considered, how much is the country's land area, population density and GDP, and whether drought has occurred in recent years.

According to Horvitz, who is responsible for this research and development, many aspects of the world have changed in recent decades, but human nature and many aspects of the environment remain unchanged, so software can understand the patterns of events from past data. So predict what will happen in the future. He said: "I am very interested in the data before the backtracking."

The market for such predictive tools is emerging. For example, a startup called RecordedFuture predicts future events based on online forward-looking reports and other sources of information, including government intelligence. The company's CEO Christopher Ahlberg said that using "hard data" to make predictions is feasible, but there is still a long way to go from prototype systems to commercial products.

45 Nike+ sensing shoes

Nike turned into an innovative company for big data marketing with a new product called Nike+. The so-called Nike+ is a product of “Nike running shoes or wristbands + sensors”. As long as the athletes wear Nike+ running shoes, the iPod can store and display data such as sports date, time, distance and calorie consumption value. Users upload data to the Nike community and can share discussions with them. Nike and Facebook reached an agreement, the user-uploaded running status will be updated to the account in real time, friends can comment and click on a "applause" button - the magic is that you can hear friends in the music while running Applause. As runners continue to upload their own running routes, Nike has a database of the best running routes in major cities. With Nike+, Nike’s urban running activities are even better. Participants upload their running data within the specified time to see which city has accumulated a long distance. With the data uploaded by athletes, Nike has successfully established the world's largest sports online community, with more than 5 million active users, constantly uploading data every day, and Nike has established an unprecedented strong relationship with consumers. Massive data plays an irreplaceable role in Nike's understanding of user habits, product improvement, precision delivery and precision marketing.

46 Volvo Industrial Internet

At the Volvo Group, a wide range of vehicle usage information, from brakes to central locking systems, is being transmitted to Volvo Group headquarters by installing sensors and embedded CPUs in truck products. “Analysis of this data will not only help us build better cars, but also help our customers get a better experience,” said Volvo Group CIORichStrader. This data is being used to optimize production processes to enhance the customer experience and improve safety. Analysis of usage data from different customers allows the product department to identify potential product problems early and alert customers ahead of these issues. "The flaws in product design can only be revealed when there are 500,000 units sold before, but now only 1,000, we can find potential defects."

47 McKesson's Dynamic Supply Chain

McKesson, the largest pharmaceutical trader in the United States, is far ahead of most companies in applying big data, integrating advanced analytical capabilities into supply chain operations that process 2 million orders per day, and oversees more than $8 billion. Inventory. For in-transit inventory management, McKesson developed a supply chain model that provides an extremely accurate view of maintenance costs based on product line, shipping costs and even carbon emissions. According to Robert Gooby, vice president of corporate process improvement, these details give the company a more realistic view of operations at any point in time. Another area in which McKesson utilizes advanced analytics is the simulation and automated processing of physical inventory configurations within distribution centers. The ability to assess policy and supply chain changes has helped the company increase its responsiveness to its customers while reducing liquidity. Overall, McKesson's supply chain transformation has enabled the company to save more than $100 million in liquidity.

48 House of cards and the film industry

The biggest feature of "House of Cards" is that it is a "network drama" unlike the production process of TV dramas in the past. In short, not only the communication channel is Internet viewing, but the drama was originally designed as a product based on “big data”, that is, Internet viewers appreciate the taste. Netflix's success lies in its powerful recommendation system Cinematch, which is based on the basic data of user video on demand such as rating, playback, fast forward, time, location, terminal, etc., stored in the database and analyzed by data to calculate what users may like. The film and provide him with customized recommendations. To this end, they opened the Netflix Awards (click to view the winning algorithm), rewarded with a million dollars, rewarding people who can improve the accuracy of their movie recommendation algorithm by at least 10%.

The cost of film production in the future will be greatly reduced, and a thousand fans will be enough to make the film a success. Still like in the "Technical Elements": "Where the eyes gather, money will follow."

49 reviews and the restaurant industry

Many state governments in the United States are working with the restaurant review network lep to supervise the hygiene situation in the catering industry, and the results are very good. People no longer look at the restaurant from the window as before, but from the comments in the mobile app! In the localized O2O reviews in China, such as public comments, tomato points, etc., consumers can judge any business, while Merchants can also use these judgments to enhance their service capabilities and make greater efficiency optimizations in the links.

The future catering industry will be completely driven by the data generated and carried by the Internet and social media. More and more people will join the review, and the speed of survival of the fittest will be greatly accelerated.