LinkedIn - by Ethan Ning
LinkedIn (https://www.linkedin.com/) is a business and employment-oriented service that operates via websites and mobile apps. Founded on December 28, 2002, and launched on May 5, 2003, it is mainly used for professional networking, including employers posting jobs and job seekers posting their CVs. As of March 2019, LinkedIn had 610 million registered members in 200 countries.
Our host, Dr. Gao Yan is a CUHK alumnus of IE department graduated in 2006, who is now serving as the senior software engineer in LinkedIn in the data segments. Before the sharing, we were shown around the campus of LinkedIn, and we are truly amazed by the high welfare standard of it. Dr. Yan proudly said that they have the “best kitchen around south Bay Area”， which in my opinion is of reliability.
Dr. Yan has work experiences both in mainland China and the US, which gives him many insights about the different working culture within these two countries. So at the beginning of his sharing, he gave us some comparison between working in mainland China and in the US. One thing he’d mentioned about mainland is that China is never short of young people working either in start-ups or in giant internet companies with passion and excitation, which give the original thrust to the rapid growth of its economic progress. For the reason above, nowadays, it is indeed a suitable choice for young people to work in China. But personally, he does enjoy more about the work-life balance as an engineer in Silicon Valley. He also suggested that cloud computing might be one of the most promising fields for the future development of software engineering.
After that, Dr. Yan provided us with some detailed explanation about how LinkedIn generate profit by utilizing their profile data. The revenue mainly comes from the recommendation information provided to professional human resource department of giant companies to make sure that these companies can reach out to the right people. Other methods of profit generation include sales, advertisement and personal premium account.
Dr. Yan’s team is working on the machine learning based recommendation system on moments and notifications. Not only the content itself but also the time to push these contents are optimized through well-trained machine learning model. But other than the machine learning model, they also have manually selected contents like a daily rundown, which is a public topic selected by the administrator.
When talking about the machine learning methodology used through his career, we wondered since Dr. Yao’s Master and Ph.D. degrees are both in the network field, why he suddenly switched to machine learning. Dr. Yan explained that it was the solid transferable technological know-how like mathematics that helped him complete this shift without having a too steep learning curve. This is indeed an inspiring point offered by him, which give us some clear guideline for the future study.