Special ability of data mining:

the data science team of Chinasoft International gathers more than 20 masters and doctors from famous universities majoring in statistics, computer and financial engineering. They are committed to carrying out the in-depth analysis on business demands in the financial, telecommunication, Internet, government, energy, transportation, medical and other industries by using data analysis and data mining technology, so as to help them make business decisions and enhance business value. Based on our own experience in the internationally recognized CRISP-DM architecture, we effectively combine business-driven factors and data-driven factors to help enterprises realize a set of data analysis and mining solutions for business understanding - data understanding - data preparation - model building - model modification - model deployment - performance monitoring. 

In the field of precision marketing, we pay attention to the tracking and service of the client’s complete life cycle, and realize the client label system construction, client segmentation, client precision marketing, client promotion, client loss warning and other subjects. These subjects are not isolated data analysis, not a cold spreadsheet in a data warehouse, but a closed-loop process from marketing design - data analysis - model prediction - marketing execution - evaluation and feedback. 

In the marketing practice of the financial industry, the precision marketing has achieved tens of times of improvement in efficiency compared with traditional marketing, which greatly saves the manpower and material resources of the banks and improves the application experience of individual customers. 

In the field of bank risk, we establish statistical models in credit risk (score card, PD, LGD, EAD), market risk (VAR calculation, stress test), and liquidity risk under the guidance of Basel Agreement III to realize the risk measurement. Meanwhile, we promote this set of theoretical approach to securities, insurance, consumption loans and other fields. As a part of risk management, many financial institutions have established risk model laboratories in recent years to standardize the process of model development. 

During the consultation and implementation of several related projects, the team has already summarized and refined the five-dimension framework of the risk model laboratory, so as to improve the development efficiency of risk model, avoid model risks, realize comprehensive security control, and achieve the accumulation of documents and program templates. In the securities industry, by taking investors, settlement participants and the market as the main analysis objects, we constantly deepen our understanding to investors, especially natural person investors from the perspective of regulation, to realize the targeted investor education and appropriate management, and meanwhile, we find and screen a small number of violation behaviors in the market to provide the analytical support for equity, prosperity and stable development of the securities market. The technology of the team covers the traditional statistical modeling - classification, regression, clustering, dimension reduction, anomaly detection, time series, and survival analysis; covers the non-structured data analysis - text classification, subject extraction, emotion analysis; proficient in SAS, R, Python and other analysis tools, and familiar with Teradata, Oracle and other database systems; familiar with Hadoop, Spark and other big data platforms as well as Mahout, MLLib and other mining algorithm development frameworks on them.