In recent years, many applications have become more helpful, personal and informative. At the core of many of these applications are the techniques of Data Science. These techniques are used to gain knowledge from the data that is collected and archived by the host organizations. This new knowledge can be applied to many areas. For example, organizations can accurately infer user preferences and future needs, resulting in a better user experience and improved customer relations.
Specific examples can be found on many retail web sites. While the use of learning systems is pervasive it is also often so natural that users are not aware of its existence. Consider the helpful suggestions from retailers based on the user’s shopping habits. These suggestions make use of collaborative filtering and other recommender system models that are built from the shopping experiences of a large number of other users.
Less visible applications are at work in supply chain, retailers, manufacturers and service organizations. Applications are used for such disparate tasks as segmentation of markets, quality improvement and fraud detection. Using clustering helps to identify market segments allows organizations to more effectively target their customers. Classification techniques are often employed to make quality assurance more effective and less costly. The difficult problem of fraud detection is aided by using anomaly detection techniques.
Many organizations are data-rich but lack either the expertise or the available personnel to take advantage of the hidden knowledge that can be found in large data stores. A common frustration for managers is sensing that there is a solution to an identified problem but not having the in-house talent available. The practice of data science has been central to many of the recent student projects. In one project, an electric utility company wanted to improve the process of entering meter readings from images that customers sent from their smart phones. The student team used image recognition and machine learning algorithms to automatically extract the readings from the computer images. In another project, students helped a consumer goods company improve customer experience by building a chatbot to provide feedback to customer questions. They used natural language processing and machine learning algorithms to map questions to relevant answers.
Within the GVSU School of Computing and Information, we have a number of faculty members with Ph.D.s in data science related areas who are always looking for new interesting problems to work on with their students. If your organization has knowledge locked up in data, and you’re looking for some assistance in unlocking its potential, we’d be happy to have a conversation with you. Another opportunity you may be interested in considering would be our new Data Science and Analytics graduate degree program.