Reimagining How We Collect User Data
- By CDOTrends editors
- February 06, 2023
Collecting user data is an essential part of delivering a personalized customer experience. However, recent changes in data privacy regulation have forced many companies to find new ways to collect user data to improve the customer journey.
So a research proposal by Dongguk University in Korea to use a network effect-based reasoning model to collect meaningful and reliable user data is turning heads. The claim is that the model uses the “influence-based social exchange” and “two-step flow of information theories” to drive outcomes.
Dr. Kihwan Nam, assistant professor of management information systems at the Business School of Dongguk University, Korea, and his colleague presented a new theoretical model that improved data reliability during the collection stage. The proposal was published in the journal Expert Systems With Applications and argued the importance of using big data analysis for improving user-data quality.
“It demonstrates theoretical, technical, and practical aspects of reliable data collection, making it valuable for corporations. Not only will our proposed model enhance business intelligence, but it will also help extract meaningful information for various other applications involving big data analysis,” said Nam about the future potential of the work.
Companies are increasingly relying on business intelligence. However, the effectiveness of business intelligence depends upon the quality of user data.
Big data has a lot of noise, and only meaningful inputs can lead to meaningful outputs. Therefore, reliable user data is essential. But current ways to improve the reliability of already collected data are not yet efficient enough.
The researchers verified the effectiveness of their reasoning model by applying it to an online media content platform's recommendation system — researching user data from that platform helped them collect verifiable proof. They found that its performance on data quality and user satisfaction improved significantly by up to five times which led to an upsurge in the number of users on the platform.
Nam explained further how the model works, “The model, based on the network effect, applies both the influence-based social exchange theory and the two-step flow of information theory. In it, initially, a user evaluates and recommends a product to a group. Its recommendation range depends on the user’s influence. The user, aware of their influence, will try to increase it by getting positive feedback from the group, thereby acting as an ‘opinion leader’ and generating data that reliably indicates popular products.”
It's still early days, but it will be interesting to see how such a model will evolve as regulations governing user data become stricter.
Image credit: iStockphoto/wildpixel