What is the difference between BAT's Internet big data application? In terms of data types, Tencent's data is the most comprehensive, which is completely related to its Internet business. The most prominent are social data and game data. Among them, the core of social data is relationship chain data, interactive data between users, text, pictures and video content generated by users; Game data mainly includes large-scale online game data, web game data and mobile game data. The core of game data is the active behavior data and payment behavior data of the game. The biggest feature of Tencent's data is various user behaviors and entertainment data based on social interaction. The most prominent thing in Ali is e-commerce data, especially the data of users browsing, searching, clicking, collecting and purchasing goods on Taobao and Tmall. The biggest feature of data is that users funnel data from browsing to paying. Baidu's data are mainly keywords searched by users, web pages, pictures and video data crawled by crawlers. Baidu's data is characterized by reflecting users' interests and needs more directly by searching keywords, and Baidu's data is more unstructured data.
Data Application Scenarios of Baidu, Alibaba and Tencent
The data application scenarios of Baidu, Alibaba and Tencent all have the same system, which is divided into seven layers, representing different levels of data value application scenarios of enterprises, forming a data value pyramid of enterprise operations:
(1) database platform layer. The bottom layer of the pyramid is also the basic layer of the whole pyramid. If the basic layer is not well built, it is difficult for the application layer above to play an effective role in enterprise operation. The technical goal of this layer is to realize effective data storage, calculation and quality management. The business goal is to string all the user (customer) data of the enterprise with a unique ID, including portraits (such as gender, age, etc. ), the behavior and hobbies of users (customers), so as to achieve the purpose of comprehensively understanding users (customers);
(2) Business operation monitoring layer. The first thing this layer should do is to build the key data system of business operation. On this basis, the data products developed by intelligent model can monitor the changes of key data, and quickly locate the reasons for data changes through various analysis models to assist operational decision-making.
(3) User/customer experience optimization layer. This layer mainly monitors and optimizes the user/customer experience through data. Both structured data and unstructured data (such as text) are used to monitor the experience. The former is more realized by using various models or tools of user (customer) experience monitoring, while the latter is more to find negative word-of-mouth by monitoring the words of Weibo, forums and internal customer feedback systems, so as to optimize products or services in time;
(4) Refine the operation and marketing layer. This layer mainly drives the refined operation and marketing of the business through data. It can be mainly divided into four aspects: first, build a user-based data extraction and operation tool to facilitate operators and marketers to extract customers through crowd orientation, so as to conduct marketing or operation activities for customers; Secondly, improve customers' response to activities through data mining; Thirdly, customer life cycle management is carried out by means of data mining; Fourthly, personalized recommendation algorithm is mainly used to recommend different goods or products based on different interests and needs of users, so as to maximize the efficiency and effect of promoting resources, such as personalized recommendation of Taobao goods;
(5) Data external service and market communication. Data external service generally serves the customers or users of Internet enterprises. For example, Baidu serves its advertisers by providing Baidu public opinion, Baidu spokesperson and Baidu index. Taobao serves customers through data Rubik's Cube, Taobao Smart and cloud products; Tencent serves its openers through Tencent Analytics and Tencent Cloud Analytics. In terms of market communication, it is mainly realized by interesting data information maps and data visualization products (such as taobao index, Baidu Index and Baidu Spring Festival travel rush Migration Map).
(6) Management analysis. Mainly through the statistics of big data by analysts, weekly, monthly and quarterly reports of experience analysis are formed to analyze users' operation and revenue completion, find problems and optimize business strategies.
(7) Strategic analysis. In this respect, it is necessary to combine internal big data to form a data view of the decision-making level, and also to combine external data, especially various competitive intelligence monitoring data and foreign trend research data, to assist the decision-making level in strategic analysis.
Although Baidu, Alibaba and Tencent have the same characteristics in the data value application system operated by enterprises, their overall big data development strategies are also significantly different due to different business models and data assets.
Baidu's Big Data Strategy
The most important source of Baidu's big data is nearly one trillion webpage data collected from more than 100 countries through crawlers, and the amount of data is at EB level. Baidu's data is very diverse. The data it collects includes unstructured or semi-structured data, including web data, videos and pictures, as well as structured data, such as users' click behavior data and advertisers' payment behavior data.
Baidu big data mainly serves three types of people: one is netizens, who make their search more accurate through big data and natural language processing technology; The second category is advertisers. Through big data, advertisers' advertisements have a high degree of matching with search keywords, or have a high degree of matching with the content of web pages that netizens are watching; The third category is Baidu big data engine, which is also a key promotion, focusing on serving enterprises with a certain data scale in traditional industries.
Baidu's big data engine represents the trend of openness and cooperation in data service capabilities of Internet companies. Baidu Big Data Engine consists of the following three aspects:
Open Cloud: Baidu's large-scale distributed computing and ultra-large-scale storage cloud. Open Cloud Big Data is open to infrastructure and hardware capabilities. In the past, Baidu cloud was mainly for developers, and the big data engine open cloud was for "big developers" with big data storage and processing needs. According to relevant personnel of Baidu, Baidu Open Cloud also has the characteristics of high CPU utilization, high flexibility and low cost. Baidu is the first large-scale commercial ARM server company in the world. The ARM architecture is characterized by low energy consumption and high storage density. At the same time, Baidu is the first company to apply GPU (graphics processor) in the field of machine learning, achieving the purpose of energy saving.
Data Factory: The data factory is Baidu's software ability to organize massive data, which is similar to the function of database software, except that the data factory is used to process TB-level or even larger data. Baidu data factory supports very large-scale heterogeneous data query, SQL-like and more complex query statements, and supports a variety of query business scenarios. At the same time, Baidu data factory will also carry the concurrent query and scanning of TB-level large tables, and the low concurrency of large queries will reach 100 GB per second.
Baidu brain: Baidu brain has opened up Baidu's previous capabilities in artificial intelligence, mainly large-scale machine learning ability and deep learning ability. Previously, it was used for voice, image, text recognition, natural language and semantic understanding, and was opened to intelligent hardware through platforms such as Baidu. Now these capabilities will be used for intelligent analysis, learning, processing and utilization of big data and opening to the outside world.
Baidu packaged infrastructure capabilities, software system capabilities and intelligent algorithm technology. After opening up through the big data engine, industries with big data can access their own data to this engine for processing. From the architecture point of view, enterprises or organizations can also choose only one of the three sets to use, such as storing data in their own cloud, but using some intelligent algorithms of Baidu brain or storing data in Baidu cloud and writing their own algorithms.
The role of Baidu big data engine
We can specifically look at the role of Baidu's big data engine from two aspects:
(1) For * * * institutions: If the transportation department has big data in places such as car networking, internet of things, road network monitoring, ship networking, terminal station monitoring, etc. If these data are combined with Baidu's search records, network-wide data and LBS data, intelligent path planning and capacity management can be realized by using the big data capability of Baidu's big data engine; The health department has the data of national influenza statutory report, sentinel surveillance and pathogenic surveillance of influenza-like cases. If we combine Baidu's search records with the data of the whole network, we can make flu prediction and vaccination guidance.
(2) For enterprises: Many enterprises also have huge amounts of big data, but their ability to process and mine big data is relatively weak. If Baidu Big Data Engine is applied, massive data can be stored reliably at low cost, and intelligent value can be mined from shallow to deep. For example, at the Baidu Technology Open Day in April, 20 14, China Ping An introduced how to use Baidu's big data capabilities to enhance the understanding and prediction of consumers, segment customer groups, and formulate personalized products and marketing programs.
Alibaba's Big Data Strategy
The overall development direction of Alibaba's big data is the development of DT(data technology) data era with the purpose of activating productivity. Alibaba's future big data will be composed of "data opening based on cloud computing+big data instrumental application":
(1) Open data based on cloud computing. Cloud computing enables small and medium-sized enterprises to obtain data storage and data processing services on Alibaba Cloud, and also to build their own data applications. Cloud computing is the foundation of data openness. Cloud computing can provide a data working platform for data developers all over the world. Ali's distributed storage platform and the algorithm tools on this platform can be better used by data developers. At the same time, Alibaba also needs to desensitize the data, so that the business definition and each label of the data are clear enough, so that data developers all over the world can start data thinking on the Alibaba platform, so that data used by * * *, consumers and industries can be used. After the opening of Ali Big Data, online and offline data can be connected in series, and everyone is a data provider and a data user.
(2) In the application of big data, Ma Yun identified two policies in the whole data application:
The first policy: from IT to DT (data technology), DT is the power to ignite and stimulate the whole data, which is used by management, society, sales, manufacturing and consumer credit. As already analyzed, Alibaba's data assets are mainly based on e-commerce. Among them, Taobao and Tmall generate rich and diverse data every day, and Alibaba has precipitated various data including transactions, finance and life services. These data can help Alibaba to carry out digital operations (as shown below).
Another of its most important applications is the financial field-microfinance. In the field of micro-finance enterprise financing. Because banks can't grasp the real business data of small and micro enterprises, not only many enterprises can't get loans, but also the whole judgment process is too long because of the lack of data types. Ali decides whether he can issue loans and the amount of loans through various data such as transactions, credit and SNS in his e-commerce data.
The second policy: Let Alibaba's data and tools become China's commercial infrastructure. Alibaba has begun to transform. Ali will change from directly facing consumers to supporting online merchants to face consumers. Ali will develop more tools to help the growth of online merchants according to the existing operation and data experience, so that online merchants can better serve consumers with the best tools and services. As Ma Yun said, "I believe that no online merchant doesn't want to have his own customers, and no online merchant doesn't want to know whether the customer experience is good or bad, and how to have these customers for a long time. We believe that a country's economy should be left to entrepreneurs, and we believe that the future economy of Taobao merchants should be left to online merchants to decide, not us to make decisions. "
Tencent's Big Data Strategy
Tencent's big data is currently more aimed at the internal operations of Tencent enterprises. Compared with Ali and Baidu, data openness is not high. So for Tencent, we mainly focus on the application scenarios and services of Tencent big data in service enterprises.
More than 90% of Tencent's data is managed centrally, and the data is concentrated in the data platform department. The data of more than 100 products are centrally managed and stored in the data warehouse (TDW) independently developed by Tencent. Tencent Big Data can be divided into four levels from different aspects of data application, including data analysis, data mining, data management and data visualization:
(1) The data analysis layer consists of four products: self-help analysis, user portrait, real-time multidimensional analysis and change intelligent positioning tool. Self-help analysis can help non-technical personnel to realize data statistics and display through simple condition configuration; User portrait is a crowd portrait aimed at a certain user group or users of a certain business to realize automation; Real-time multidimensional analysis tools can realize real-time multidimensional subdivision of an indicator, which is convenient for analysts to analyze an indicator from different angles; The intelligent location tool of data change realizes the intelligent location of data change.
(2) The product applications at the data mining level include: accurate advertising system, personalized recommendation engine for users and customer life cycle management. Accurate advertising systems, such as Guangdiantong, based on the massive data of Tencent's social platform, realize accurate advertising through accurate recommendation algorithm and intelligent targeted promotion; According to the interests and preferences of each user, the user personalized recommendation engine realizes the personalized recommendation requirements of products through personalized recommendation algorithms (collaborative filtering, content-based recommendation, graph algorithm, Bayesian, etc.). ); Based on big data, customer lifecycle management system carries out data mining according to different life cycles of users/customers, establishes prediction, early warning and user characteristic models, and carries out refined operation and marketing according to different life cycle characteristics of users/customers.
(3) On the data management level, there are TDW (Tencent data warehouse), TDBank (database), metadata management platform, task scheduling system and data monitoring. This level is mainly to realize efficient centralized storage of data, data business index definition management, data quality management, timely scheduling and calculation of computing tasks, monitoring and alarm of data problems.
(4) On the data visualization level, there are tools such as self-help report tools, Tencent Compass, Tencent Analysis and Tencent Cloud Analysis. Self-service report tool can realize self-service report with relatively simple structure and logic. Tencent compass is divided into internal version and external version. The internal version is an efficient reporting tool for Tencent's internal users (product managers, operators and technicians, etc.). ), while the external version is a reporting tool for Tencent partners such as developers. Tencent Analytics is a web page analysis tool to help website owners analyze their websites in all directions. Tencent Cloud Analysis is an analysis tool to help application developers make decisions and optimize operations.
Generally speaking, Baidu, Alibaba and Tencent all have big data, and use the data of the three Internet giants to optimize the operational effect of their own businesses. From this perspective, their data value application scenarios are similar. However, due to different business and business models, their data assets are different, and their future big data strategies are also different. Especially from the perspective of the openness and cooperation of big data, Baidu and Alibaba are relatively more open. For Internet companies that attach importance to the openness and cooperation of big data, what they are most looking forward to is to exchange more data with more traditional industries through the strategy of big data openness, so as to better enrich their online and offline data, form online and offline data collaboration, and expand new business models, such as intelligent hardware and big data health.
What is the difference between bat's Internet big data application? This has to be analyzed from the respective genes of BAT. Baidu mainly searches for products, so big data is mainly used for Baidu's search, making the search more accurate and matching; Alibaba is dominated by e-commerce, so big data will be Alibaba's main user goods; Tencent is mainly social, so big data may be more applied to social network analysis for Tencent. The main purpose of big data is prediction, so the similarity between BAT and big data is to provide more accurate service and marketing through the analysis of users.
See how Baidu, Ali and Tencent use the Internet big data application. Ali has a data cube to provide sellers with charging services.
What's the difference between "Internet" and "all space" in Baidu? "Internet"
and
"All space"
Internet refers to all the information on the Internet.
For Baidu,
Mainly Chinese information
All space
It means all users of Baidu.
Built Baidu space
(blog+photo album+message board)
Obviously, looking for the latter
Does not include blogs outside Baidu space.
How to acquire and apply Internet big data is a large amount of high-speed and changeable information, which needs new processing methods to promote stronger decision-making ability, insight and optimization. Big data provides unprecedented space and potential for enterprises to gain deeper and more comprehensive insights.
With the help of big data and related technologies, targeted marketing can be carried out for customers with different behavioral characteristics, even from "recommending a product to a suitable customer" to "recommending a suitable product to a certain customer", thus focusing more on customers and carrying out personalized precision marketing.
Precision marketing in the era of big data refers to obtaining the preferences and behavior preferences of objects through big data, and carrying out different marketing for different objects. The core of big data precision marketing can be summarized as several key words: users, needs, recognition and experience.
Yimei SoftCom launched data cloud service, which continued Yimei's business philosophy of customer service, customer marketing and customer management, and provided customers with data-level services such as data verification and precision marketing through huge consumption data resources. Simply put, it is to provide data verification and data screening services for enterprises.
-
What is the application prospect of Internet big data training? Don't worry, if you learn well, you will have a good prospect. {Variable 9}
What's the difference between big data and small data? 1. Re-prediction of big data and re-interpretation of small data; 2. Big data is rediscovered and small data is empirical; 3. Big data is relatable, and small data is causal; 4. Big data focuses on the whole, and small data is resampled; 5. Big data is re-perceived, and small data is re-accurate.
What's the difference between an enterprise data center and an Internet data center? DCCI Internet Data Center (DCCI &；; Data platform, service provider for measuring, analyzing and optimizing interactive marketing. Based on panel software, code embedding, massive data mining, semantic information processing and other leading technical means, websites and applications are carried out. ...
Internet data center: idc, which mainly stores network data (website+data+download site, etc.). ). Covering a wide range, any regular enterprise or small and medium-sized webmaster can choose.
Enterprise data center: it is more targeted and can be part of the Internet data center.