Information Synthesis
Jacob Scanlon is a consulting specialist with more than ten years of experience in the commercial sector and government agencies. He graduated from the engineering school at the University of Virginia with a Masters’ in Machine Learning and Predicting Artificial Intelligence (AI). The interviewee’s research focused on text mining, which a special case of data mining. Data analytics is the basis for developing AI technology based on compiling data, which may involve various steps, such as converting into the count of specific phrases or words so that the machine can understand them as numbers. Speaking of clients, Scanlon clarifies that some companies have an extensive IT infrastructure and funds to support it, for instance, Capital One can be noted as an organization that has several groups, including IT teams working on analytics. Factories also have a lot of machines that create products and collect raw data every second in the form of petabytes.
Data analytics impacts many industries, among which there is energy, the Internet of Things (IoT), traveling, military, insurance, safety, and so on. The technologies of the IoT are now radically changing the world, and these changes concern not only high-tech industries or infrastructure but also everyday life. For example, the interviewee mentions that a car breakage can be predicted, and its owner can be notified about the opportunity to bring it to the service in advance, which allows saving customers’ money. Another example refers to the energy company performing in South Africa that saved $1 billion due to predictive maintenance. The use of central computers and data delivery promotes controlling peripherals either directly or via service terminals.
Some companies already have essential experience with data analytics integration into their processes. In this case, as stated by Scanlon, his role is mainly associated with training engines to make them perform their specific modeling (Appendix 1). People have general knowledge about data mining, even though their companies have analytics. Nevertheless, the consultant can assist them in finding answers to the existing questions based on their products in terms of a cluster model. This is possible due to the implementation of chasing projects, one of the platforms for which is Aster Analytics that providers a suite for businesses, including modules and instruments to gain relevant insights during the whole data analytics lifecycle.
The issue of security is highlighted in the given interview as an integral part of the ongoing technologic improvements and the threat of data use by unauthorized parties. In particular, Scanlon clarifies that insurance companies have the most advanced protection mechanisms to avoid trials, yet they collect the personal medical data of their customers. There are several data regulation policies, such as the California Consumer Privacy Act or the General Data Protection Regulation (GDPR) that was elaborated by the EU. In many cases, the companies try to meet their just minimum requirements.
Interesting Points
One of the most thought-provoking points noted in the interview refers to the fact that Scanlon works on machine learning every day (see Appendix 2). It is stated that pre-processing takes a great part of AI since the machine needs a special language to understand what it should do. In other words, it is not enough to merely request one or another operation, but a set of segmentation and converting should be made to turn human words into machine-comprehensible data.
Another perspective that seems to be especially pertinent to the modern business environment and discussed in the interview is the concept of deep learning. It allows for training a model to predict the outcome of a set of input data, which can be achieved by the use of both controlled and uncontrolled methods. To train the network, one needs to submit prepared data and compare the output results generated by the machine with the results from the test data set. It is quite impressive that deep learning can be used across the spheres of interest: beginning with the Internet search by a photo as a keyword and ending with video set-ups of the road in the military area. Regardless of the ultimate goal of deep learning, the final result is a product of a math equation. Accordingly, one may suggest that such technology makes customers closer to their desired products and offers companies a way to meet and anticipate their needs.
It was surprising to discover that with the introduction of the GDPR, many companies became more careful about the data they obtain from customers. Previously, the organizations offered free services and collected data via e-mails or search requests, but the establishment of the identified regulation limited the so-called privacy for convenience phenomenon (Appendix 3). The important question asked by the residents of the Silicon Valley companies is that whether they should move to avoid complying with such regulations since currently, they have to design and follow a happy median to data storage and security.
Impact of the Interview
I have learned that today’s business environment has the opportunity to take advantage of the IT sector, namely, from AI technology. While various corporations have already introduced machines that obtain the necessary information that can be used to take more thorough decisions, small companies encounter pressure to do the same. Considering harsh competition with the market giants, smaller organizations can significantly benefit from predictive maintenance that is likely to help in anticipating customer behaviors and preferences. Along with the overall understanding of data analytics functions and positive impacts, I have also learned that governments are aware of a potential data leakage problem, and they work towards protecting users from such a threat. The discussed regulations seem to be important to make customers’ privacy protected and restrict unwanted advertising.
One may assume that data mining can train machines to recognize the attempts to access personal data without permission and prevent them. For example, the analysis of the advertising media skipped by users can be conducted to block them and make the online experience of people more comfortable. However, it is also promising to expect that businesses would be resistant to such changes since they want to sell as much as possible. Elaborating on the idea of privacy, I consider that this topic should be researched further, thus leading to more sophisticated means of protecting data. Nowadays, billions of devices exchange information with each other, and the world of the IoT grows exponentially. Mastering new technologies, the countries should prepare a legal framework and practical tools that will regulate the new information space. Thus, this interview was rather beneficial to understand the key tendencies and challenges in the field of AI and its dimensions, such as deep learning, data analytics, text mining, et cetera.
Appendix
- – What happens with the client has an established analytics team? What do they have lacking in their team?
– A lot of what I have done as a consultant is train and chase projects to perform on parallel processing clusters the companies sell. Parallel processing clusters are called Teradata specifically for analytics pooled in Aster. - – I do machine learning pretty much every day or, at least, analytics every day, that is compiling data called pre-processing, which is turning data into “features” that machines can understand.
- – Regulations are very strong, and punishments are already put in place and used effectively. … financial institutions are protected most because people sue them all the time.