What is Taobao data mining?

Question 1: Taobao data mining is completely impossible. If this were the case, Taobao would have spam text messages or Want Want messages all over the place. Taobao would not be able to reveal customers' private information to you.

Question 2: What are the commonly used data mining data sources in e-commerce? 1. Traffic 1. Search traffic tool: Search Diagnostic Assistant

A-Basic conditions: No violation, can be found in "Seller Workbench" - "Search" "Diagnostic Assistant"-"Baby Diagnosis" check.

B―Relevance: Category attribute relevance, title keyword relevance. C - Popularity points: whether it is a window recommendation, whether it has consumer protection, DSR score, Alipay usage rate, Wangwang effect speed, and the time difference between auction and delivery.

D-Pictures: Many sellers often ignore the optimization of pictures when optimizing the main search traffic. However, the gap in picture click-through rates directly affects the final search traffic. Buyers do not come in directly by searching, but are attracted by pictures, so optimizing pictures is very important. It is recommended to use a through train to test images (the method will be introduced below).

E―Price and sales volume: Products with similar sales volume have more display opportunities with higher prices; products with the same price have more display opportunities with higher sales volume. When checking this indicator, we mainly check the gap between ourselves and our direct competitors, especially the gap in sales volume within 7 days, to make adjustments.

F - Title optimization: use more long-tail words when sales are relatively low, use more general words and central words when sales are high, and test repeatedly to get the maximum search traffic conversion rate .

2. Paid traffic tools: data reports and store checks for each paid tool.

―Taoke: To diagnose Taoke, you only need to look at the sales and commission gap between yourself and your competitors.

2. Conversion 1. Conversion rate tool: Store Check

A- Inner pages: first look at sales, secondly look at evaluation quality, and then look at single product conversion rate and page dwell time and inquiry rate. If there are no basic sales and poor reviews, the conversion rate cannot be good. After the two prerequisites are solved, let’s see if the single product conversion rate, page dwell time and inquiry rate are not lower than the industry average (or the products sold well in the store). If it is lower than that, then optimize the USP selling points, logical sequence (whether they are all centered around the USP), diversified display content, and display methods one by one.

B-Depth of visit: Since 80% of customers enter the store from the inner page, we mainly optimize the positions where the inner page can divert traffic, namely store promotion, baby page association, and baby page sidebar. , Optimize the end of the store. Then optimize the homepage.

C-Payment rate: whether it is above 80.

D-Marketing activities: Regularly holding marketing activities can increase conversion rates.

E-Customer service inquiry conversion rate: whether it is at least the industry average. Viewing tools: third-party tools such as StoreChacha. Optimization method: Establish standard answers to every customer question. 2. DSR tool: Taobao DSR score calculator. Optimization methods: a. Upgrade of Taobao’s original services (7 days to 30 days without reason, 3 days of delivery to 24 hours of delivery, etc.); b. Innovation of Taobao’s unavailable services (centered on the contact points between customers and merchants) Innovation, such as SNS, games). 3. CRM CRM mainly checks the proportion of old customers, the conversion rate of old customers, the second purchase rate, and the ROI of customer group short color mails. Tools: Seller Workbench - Member Relationship Management, Shuyun, Kedao and other third-party software. Optimization method: Establish groups of old customers and create different privileges for old customers based on the groups. Higher-level customers have higher-level privileges.

Question 3: Why are data mining salaries so high? Now is the era of big data, and it is necessary to mine the relationship between data and draw some rules. For example, when you are shopping online, Taobao uses mining technology to discover your behavioral preferences, and will recommend items you like when you browse Taobao-related web pages.

Question 4: The difference between big data and data mining. Data mining requires a lot of interdisciplinary knowledge such as artificial intelligence, databases, machine language and statistical analysis knowledge. Furthermore, the emergence of data mining requires conditions. The first condition: massive data; the second condition: the ability of computer technology to process large amounts of data; the third condition: the storage and computing capabilities of computers; the fourth condition: The development of interdisciplinary subjects.

Big data is only a condition for the success of data mining.

Question 5: What do data mining engineers generally do? Job Responsibilities:

1. Based on your understanding of the industry and the company's business, independently undertake complex analysis tasks and formulate analysis reports;

2. Relevant analysis directions include: user behavior Analysis, advertising click analysis, business logic related and competitive environment related;

3. Based on changes in business logic, design corresponding analysis models and support the development of business analysis work.

Job requirements:

1. More than 2 years of experience in industry modeling;

2. Bachelor degree or above in mathematics, statistics, computer, physics and other related majors Graduation;

3. Proficient in basic science and data mining technology, especially regression models and decision tree models.

4. Proficient in various types of data analysis tools such as SPSS Clementine/SAS EM, and able to produce professional analysis reports;

5. Have actual data mining projects in a certain industry such as finance, communications or the Internet Experience and a deep understanding of this industry;

6. Passionate about the Internet field, strong learning and interpersonal skills, influence and persuasion skills, and like challenging work.

Question 6: Which one is more promising, big data or data mining? Big data includes data mining. Data mining is one of the branches of big data and is also the foundation. If you are studying the direction of BI, data mining is Basics, the two are closely related. The concept of data mining came out relatively early. You should know the allusion of beer and diapers. Data mining has been used in early data warehouse modeling, and big data has been a hot trend in recent years. Very good, the future is the era of big data. Currently, many large enterprises are doing big data (such as solution providers: IBM, ORACLE, SAP, EMC, Huawei, etc.; self-research: Taobao, Tencent, etc.; Party A : Mobile, telecommunications, etc.) The career prospects are still very good. Big data content is very rich, including hadoop, stream processing, distributed, NAS/SAN, etc., which will be of great help to your future development. My suggestion is big data. Hope it will be adopted.

Question 7: How to use big data mining to correspond to e-commerce. Data mining can discover the uniqueness and personality of e-commerce customers, necessary and accidental knowledge, independent and related knowledge, reality and Predictive knowledge, etc. After analysis, all this knowledge can make statistical and correct analysis of customers' consumption behavior such as psychology, ability, motivation, demand, potential, etc., and provide managers with decision-making basis. The specific applications are as follows:

1. Application of classification and prediction methods in e-commerce

In e-commerce activities, classification is a very important task and is also the most widely used technology. . The purpose of classification is to construct a classification function or classification model, usually called a classifier. Classifier construction methods usually include statistical methods, machine learning methods, neural network methods, etc. These methods can map the data in the database to a given category for prediction, that is, using historical data records to automatically derive a generalized description of the given data to predict future data.

2. Application of clustering method in e-commerce

Clustering is to classify a group of individuals into several categories based on the principle of similarity. For e-commerce, customer clustering can provide strong support for market segmentation theory.

The purpose of market segmentation is to make the distance between individuals belonging to the same category as small as possible, and the distance between individuals of different categories as large as possible. By extracting clustered customer characteristics, e-commerce websites can provide customers with Personalized service.

3. Application of data extraction methods in e-commerce

The purpose of data extraction is to condense the data and give its compact description, such as summation value, average value, Variance values, other statistical values, or graphic representations such as histograms and pie charts. More importantly, he discusses data summary from the perspective of data generalization. Data generalization is a process of abstracting the most original and basic information data from low level to high level. Multidimensional data analysis methods and attribute-oriented induction methods can be used. In e-commerce activities, dimensional data analysis method is used for data extraction, which is aimed at the customer data warehouse in e-commerce activities. Aggregation operations such as sum, total, average, maximum, and minimum are often used in data analysis. Such operations require a particularly large amount of calculations. The results of the aggregation operations can be pre-calculated and stored for use in decision support systems. .

4. Application of association rules in e-commerce

The management department can collect and store a large amount of sales data and customer information, analyze these historical data and discover association rules. For example, it analyzes the purchasing behavior of online customers to help managers plan the market and determine the type, price, quality, etc. of goods. Generally, there are two types of association rules: meaningful association rules and generalized association rules. Meaningful association rules are rules that satisfy the minimum support and minimum credibility. Minimum support, which represents the minimum level that a group of objects needs to satisfy in a statistical sense, such as the number of customers in e-commerce activities, customer spending power, consumption patterns, etc. The latter is the minimum reliability of association rules specified by the user. The second is the generalization rule. This kind of rule is more practical because there is a hierarchical relationship between the research objects. For example, bread and cakes belong to the pastry category, and pastries belong to the food category. With the hierarchical relationship, it can help to discover more interesting things. rules of meaning.

5. Optimize corporate resources

Saving costs is the key to corporate profitability. Based on data mining technology, we can grasp enterprise resource information in real time, comprehensively and accurately. By analyzing historical financial data, inventory data and transaction data, we can discover the key points of enterprise resource consumption and the input-output ratio of main activities, thereby providing enterprise resources Optimizing configuration provides the basis for decision-making, such as reducing inventory, increasing inventory turnover rate, improving capital utilization, etc. Through Web data mining, business information can be quickly extracted, allowing enterprises to accurately grasp market dynamics, greatly improving their ability to respond to market changes and innovating, allowing enterprises to maximize the use of human resources, material resources and information resources, and coordinate reasonably The relationship between internal and external resources of the enterprise produces the best economic benefits. Promote the scientific, informatization and intelligence of enterprise development.

6. Manage customer data

As the "customer-centric" business philosophy continues to gain popularity, analyzing customers, understanding customers, and guiding customer needs have become the key to business operations. important topic.

Based on data mining technology, enterprises will maximize the use of customer resources to carry out analysis and prediction of customer behavior...gt;gt;

Question 8: Approximate data mining services for R language generation programming How much does it cost? I searched for the "Big Data Tribe" store name on Taobao and the price is 20 yuan. The reviews are pretty good? It is judged based on the difficulty and workload of the data service. You have to send the specific requirements to the buyer of Yibao, and he will judge with you. Generally, the prices on Yibao are units of measurement, and they are actually 20 yuan. Multiples,

Question 9: What is the promotion system for Taobao store operations? Promotion system for Taobao store operations:

1. E-commerce strategic planning

Based on data mining, through 360-degree research on the market, competitors, consumers, and the company itself Through insight analysis, we plan out the company's overall e-commerce model, overall strategic goals, development stages, investment and expected benefits, etc. to clarify ideas and clarify directions.

And decompose the project functions to form a project progress control Gantt chart, and implement the detailed strategic implementation plan that is subdivided into executable, supervisory, and controllable.

2. Store Planning and Decoration

Based on the overall analysis and planning, we will establish first-class Taobao store planners and first-class UI designers. Through the overall structure, column division, and The overall integrated planning and design of process experience and visual style highlights the store's brand temperament and customer shopping experience.

3. Product Planning

Use USP (Unique Selling Proposition) to plan a comprehensive system of FABE model brand planning, combine industry characteristics and Taobao shopping network cultural characteristics, and use both perceptual and rational methods. With the organic integration of ideas, we plan and design the product baby page with the most sales power, thereby effectively increasing the product conversion rate.

Based on data mining, through matrix planning and pricing system planning of star hot-selling products, golden bull profit products, and impact products, a complete product width and product portfolio are formed to achieve a balanced and unified hot sales and profits. And solve the problem of conflict between online and offline channels.

4. Product Promotion Operation

Use Taobao’s various promotional activities to plan various themed activities for creative stores and related sales, cross-sales and other means to bring products to life and enhance users Stickiness, increase unit price per customer, create hot-selling products, and ultimately achieve a sales leap.

5. Promotion and operation

The BRIC Taobao promotion and operation system is centered on the introduction of target traffic, and adopts methods such as free promotion within the Taobao site, tool advertising promotion within the Taobao site, and auxiliary promotion across the entire network. The system solves Taobao store traffic problems and brings a large number of effective target purchasing customers to the store. We insist on achieving the maximum promotion effect with minimum investment under the guidance of the strategy. We will never blindly follow traffic, let alone promote ineffective traffic, and achieve Double effect of sales and brand promotion.

6. Customer service sales

Customer service sales is the key link to achieve sales, the last step, and has a core position. BRICS will carry out standardized and systematic operations from the four levels of business, culture, management and training to achieve a streamlined and replicable sales and customer service system.

7. Data Analysis

Data mining and analysis are the most obvious differences between e-commerce and traditional offline commerce. The data of e-commerce is accurate and real-time. The basis of Taobao's operating system is data mining and analysis.

Through horizontal, vertical and cross-sectional analysis of various data, we can formulate strategies, improve promotion effects, and increase store conversion rates, thereby increasing the ROI of the entire store and maximizing corporate profits.

The above views on Taobao project operations are just a brief analysis of my personal suggestions from the system level. Jinzhu believes that Taobao e-commerce operations should be based on data mining and focus on improving store conversion rates. Starting from strategic planning, online store planning, product planning, product promotion, Taobao promotion, customer service sales, data analysis, etc., only by building a system can you win!

Question 10: Data analysis is "finding a needle in a haystack" "Has Alibaba data mined? Introduction: How does big data generate value? Is big data omnipotent? Where are the application boundaries? Everyone seems to have a vague idea of ??these questions, but there is never a unified answer. Today's discussion on "big data" has reached a peak. Data is the future and has unequivocally become the center of the future new strategic development of Internet companies. What is big data? How does big data generate value? Is big data omnipotent? Where are the application boundaries? Everyone seems to have a vague idea of ??these questions, but there is never a unified answer. When it comes to big data, the first one to bear the brunt should be the Alibaba system, which has been working in the data ocean for a long time and has derived financial lending business. Jack Ma merged the two core businesses of the group, Alibaba Finance and Alipay, to form Alibaba Small and Micro Finance, and placed the most popular successor Peng Lei at the helm of Alibaba Small and Micro Finance. This shows that Jack Ma attaches great importance to the future data battlefield. As the data platform of Alibaba Small and Micro Financial Services Group under preparation, the person in charge, Feng Chunpei, also has unique insights into data. He told the author that the current domestic discussion of big data focuses more on the technical direction, that is, "how to precipitate data." There is less thinking about application. How does data generate value? This needs to start from the nature of big data. Online data is big data. To understand what big data is, you first need to know what kind of data is useful. According to Feng Chunpei’s understanding, any behavior itself will generate data, but only online data can be deposited and utilized. "For example, without going through Taobao, people's transaction behavior originally generates data offline, but this transaction behavior is private. Except for the buyer and seller, other people do not know my transaction behavior, and both parties to the transaction are also anonymous. From the nature of the data, it cannot be precipitated, and from the source, there is no method to effectively collect it. "What is big data? Feng Chunpei’s understanding seems to be closer to the essence: “The essence of owning data is that you have a more comprehensive and clear understanding of the world, these people, these companies, and this era. You can understand these People's needs, you can understand any changes in the world." You can understand that if you are an in-depth user of Alibaba (such as a Taobao seller), they have enough data about you and will evaluate your credit. More comprehensively, this data can not only play a role in the financial field, such as helping you get more convenient loans from Ali Microcredit, but also reflect your credit status in life, "For example, on a blind date, how do you prove your income? You get When you pay Alipay bills, the girl saw that you spent 1 million in a year. If you say that your credit is good and you pay off your credit cards on time every month, it is much more useful than just talking about it. " Data is the means of production. If data? Just as auxiliary reference information, so much energy needs to be invested. In terms of production factors, what role does data play? Feng Chunpei’s definition is “means of production”. "The name of our department is 'Business Intelligence Department'. Data is more like an auxiliary decision-making for business. As a "consultant", now we have to gradually integrate this data into our business and product processes. Inside, data and business are like two gears that can rotate together. As our mining and understanding of data become stronger and stronger, eventually data can not only generate value, but also directly generate products, such as some data from Alibaba Finance. , we define it as a means of production.” This is what Alibaba will do in the future, turning data into means of production. Unlike traditional production materials, data can be used unlimited times, and the more it is used, the richer it becomes.

Alibaba has recently made frequent moves in the mobile Internet market. In the future, it may be possible to integrate data and present various user information in a panoramic view. Even in a completely unfamiliar city, with the help of this service, you can know where is nearby. The store supports Alipay payment. Which netizen on Weibo just stopped at a nearby coffee shop. Data analysis is "finding a needle in a haystack". The same problem exists with most Internet products. The data generated by the Internet may be forged, and it is also disordered and fragmented. Regarding this point, Feng Chunpei also made no secret of it, “Of course it is possible to falsify short-term data, and it is also entirely possible to falsify data using specific dimensions. However, because our business is based on long-term data for tracking and analysis, the adopted dimensions It is also wider, and the cost and difficulty of falsifying data will become larger and larger. According to our current credit model, the benefits of falsifying data are unlikely to cover the costs, so we can basically judge that data...gt; gt;