Accelerate Big Data Intellectual Application
Create New Profit Model of Enterprise
Big data collection and data mining technology have been becoming hot field of research in recent years, numerous of multi-national companies, such as Intel, Google and Alibaba, have invested enormous resources to dig and analyze big data to modify business strategy. Before applying big data, what preparation do we need?
Q: Apart from establishing platform and owning tools, how dose enterprise smoothly import and effectively apply big data?
A: Cloudera’s professional big data consultant Steven Totman claimed that the good enterprise internal application of big data contains three steps:
1. Enterprise must cherish data as asset and build up data-oriented corporate culture
He believes general enterprises are familiar with their hardware equipment assets but not paying attention to problem-solving capability of data, being ignorant to value bought by data.
2. Establish appropriate team and usage of tool and technique
Currently numerous enterprises rely on data analysts’ analysis result to adjust marketing strategy; however, they are often limited by algorithm model which used part of data for organizing data, analyzing and establishing analytic prediction, easily leading to inaccurate analysis result and rise of information security risk. They must continuously adjust analysis model and method to find an effective model which is capable to cope with complex problem of enterprise. Yet, specialization is necessary, team must combine effort of talents such as data analyst, data engineer and structural analysis engineer, flexibly allocate project to promote internal use of big data. Using analytic technology to process enormous data of enterprise, no longer require 1 – 3 quarter to plan development, but only spend 1 – 3 months to schedule.
3. Save human resources and time by using machine learning
Steven Totman mentioned case studies in various industries, through collecting feedback of data, exact status and deterioration stage of parts in production process can be learned and thus capable to schedule and know when to change unsuitable parts. He believes sole data is useless and only transforms data integration capability as information is meaningful. He uses snow storm in New York as example to illustrate the impact to daily life of citizen from prioritizing amount of salt usage by road through collecting ten million per day for analysis and research.
Steven Totman’s idea is that in future, all currently unsolvable problems can be solved by applying data.
Q: What problem will be confronted when enterprise applying big data?
What capability is needed to solve it?
A: Wang Yun, Consultant of Data Analytic Research Department claimed that the first issues are enterprise’s interpretation on analytic tool.
Concept of enterprise urgently need to change
Utilizing big data to build up model is a continuous process. It is not an end when enterprise built a model, but a new commencement – using a tool which can be continuously improved.
Analytic tool in the past only focus on existing situation, generally less improvement will be made after its production. However, data model can be continuously improved according to current data, assisting enterprise to make decision based on higher degree of understanding reality.
Therefore, enterprise must continuously input new data, testing existed assumption and prediction ability of model with new data, to understand capability of data model.
‘It’s a continuous process.’ Wang Yun believes that enterprise should concern data quality and improve data model consistently to form better foundation of creating model.
(source: How does Big Data model assist enterprise to boost business & develop a business strategy)
Want to build up platform to be well-prepared for era of data? Contact MACRO immediately
Phone: 2802 2222