Developing Unsupervised Learning Techniques for Business News Articles

Business leaders, marketers use business news to inform their day-to-day decision-making processes. This includes creating a business strategy, determining advertising tactics, planning operations, and investing in workforce development. Business news provides a global perspective on current business trends and developments, including tracking mergers and acquisitions and keeping a stakeholders informed. Keeping up with commercial ventures is crucial in today’s fast-paced business world. However, with the abundance of news websites available, it can be a daunting task to sort through the substantial volume of information. This research aims to find new ways to gather relevant news articles by embedding them into vectors through the use of in-depth research methods. The study employs unsupervised learning techniques to form clusters of these embedded vectors, with the goal of developing innovative methods for aggregating relevant news pieces. In the study, related news stories are grouped using clustering algorithms including K-means and agglomerative algorithms, as well as a variety of embedding techniques like Bag-Of-Words, TFIDF, Word2vec, and BERT.