Recently, the third CIO Summit of the Bank of China (BOC) was held in Shanghai in 2021, bringing together executives and CIOs from the financial, technological and Internet sectors to discuss and share with them on the theme of "Banking Era 4.0: Go all in on Digital Transformation", so as to share ideas on the way of the digital transformation of banking.
Magic Data Tech participated in the Summit as a data service provider, fully interpreting how conversational AI datasets can help banks to achieve digital transformation from digitized banking scenes, AI model training, and data security.
Data boosts "threshold" of customer experience in the retail banking
Different from a decade ago, bank counters nowadays are replaced by ATM, online banking, and mobile applications. An APP is used to meet customers' needs whenever and wherever. The trend towards digitalization forces banks to change their business models and redesign their channels to provide their customers with more diversified and personalized experience.
Faced with diversifying business scenes both online and offline, artificial intelligence, such as the intelligent transformation of bank outlets, service quality and compliance testing, intelligent customer services, intelligent marketing and intelligent meetings has become necessities to the digitalization of banks.
At present, banks and all other sectors are at the initial stage of being intelligent, and much progress remains to be made, particularly under the scene of intelligent customer service—robot calls are "mechanical" and "can't do anything" about heavy accents, which lead to a bad consumer experience.
In response to this pain point, Zhang Qingqing, founder and CEO of Magic Data Tech, said in the "Data Drives Finance" sharing: among the three forces driving AI, data is the "food" of algorithms, only high-precision, high-matching, tagged standardized data can be trained into high-performance models, so that can AI can talk with customers naturally. The quality of data determines the performance of intelligent system. Structured data plays a strong supporting role in improving the intelligent level and promoting digital transformation of banking industry.
Improving efficiency and reducing cost by applying the 20/80 rule
In the tide of digital transformation, most of the banks are in the trial stage of artificial intelligence, and take a cautious attitude toward investment in artificial intelligence.
Among the 116 companies surveyed by E&Y on the Maturity of Artificial Intelligence in Greater China, 53% of AI business is under trial implementation. Many executives worry about the uncertainty of the ROI, while firmly believe that intelligent transformation is an inevitable trend.
How to find a balance between the input and the effective output, and how to realize the intelligent transformation steadily are the focus of enterprise managers. For the AI industry, besides algorithm, the performance of the system is determined by the data. The quality of the data determines the performance of the system, and the quantity of the data determines the performance of the system.
How can AI developers achieve desired performance due to constraints such as budget limitation and specific application scenes?
Magic Data Tech believes that data ratio should follows the "20/80" rule, which mean with 80% of the data guaranteeing universal performance, customizing the remaining 20%.
The standardized datasets can satisfy requirement of training model when the required recognition rate is 0%-90%; when the required recognition rate is 90%-95%, besides standardized datasets, tailored datasets will be in need; when the required recognition rate is 95%-99%, fully customized datasets are highly recommended.
AI developers can use portfolio of datasets containing standardized data and customized according to specific needs, helping banks to control ROI effectively and reduce costs but improve efficiency.
Data Security as the Key to Digital Transformation
In order to realize digital transformation, banks shall value data and ensure data privacy. Data privacy protection has become an important topic for discussion at the Summit.
In recent years, problems of data security have emerged one after another in commercial banks, and many banks have been supervised due to data leakage or improper use. In the context of the artificial intelligence, data security is related to privacy, ethics, social security and other aspects, and data security has become increasingly important.
For banks, data security is the key to digital transformation. In order to ensure data security, Magic Data Tech complies with regulations of compliance, ensuring personal information are properly handled.
Datasets of Magic Data Tech are qualified for ISO/IEC 27701: 2019, ISO9001, ISO 27001 and CMMI3 criteria, especially the ISO/IEC 27701: 2019 criterion. At the same time, Magic Data Tech develops independent SaaS for data annotation, which can be privately deployed within enterprise, by doing which the whole process of data annotation can be monitored and completed within the intranet.
The digital transformation of banks is well underway, and Magic Data Tech will use massive standardized data to support the digitization of banks in Industry 4.0.
As AI research and development is moving forward both in depth and breadth, the needs for structured data grow explosively. Meanwhile, the data labeling industry is undergoing decentralization: production of structured data is shifting from large-scale third party data processing centers to scattered data end-users.
The annual INTERSPEECH, held between August 30, and September 3, 2021, is a global conference organized by International Speech Communication Association (ISCA). The INTERSPEECH 2021 is held in hybrid form, that is participants can join the conference virtually online and physically in Brno, The Czech Republic.
Annotator® 5.0, an independently developed data annotation system, was official launched by Magic Data Tech on July 8, 2021 on World Artificial Intelligence Conference (WAIC) 2021, which is a global gathering and exchange of AI innovation ideas, technologies, and applications, held on Shanghai between July 7 and July 10.
Collecting and sorting out data have always been a time-consuming and tedious procedure for AI developers and researchers. Here we list 20 sites where high-quality data is ready and free, in hope of assisting you to locate the proper dataset for your AI modal in a better way.
A batch of datasets for conversational AI were newly release on MagicHub.io, our open-source community. Let’s have a quick look.