Download Project (1.2 MB)
Since the beginning of mankind, global occurrences were found to be directly impacted by the emotions of a common man. However people became aware via newspapers and televisions only after the event had occurred. As technology became instant, faster, and handy, it started impacting the occurrences directly as the tweets/posts posted in the Social media platforms; such as twitter directly initiates conversations about earth shattering events. One such event taken as an experiment is the ongoing war between Russia and Ukraine. Language barrier is a real life challenge and it indirectly impacts the events because of misunderstandings. In this work, a dataset of 10014 tweets posted in 40 different languages from Feb, 21 2022 till Mar, 17 2022 are collected and an attempt to categorize those tweets into positive, negative and neutral sentiments are performed. Upon inferring the pattern using event extraction and sentiment analysis techniques that lie as a branch of natural language processing, it is astounding to note meaningful information. Also, impressive data are collected directly from a community in Africa called “Masakhane” and have been successful in analyzing the sentiments for swahili language. Between the two experiments made using ‘192.Apofasi’ - the engine based out of NLP with nltk python library, found a sentimental correlation exists between the war and the tweets. The real challenge is in gathering and preprocessing the dataset to make it a machine understandable model. As a result, 5094 neutral, 2788 positive and 2132 negative tweets are obtained. It is therefore a strong evidence that by developing and deploying a machine learning model into handheld devices, every human being will be able to understand the patterns of the current events irrespective of the language thereby opening numerous opportunities for its betterment. Keywords : Natural Language Processing, Event extraction, Sentimental analysis, Emotive tweets, Russia, Ukraine
Phu Huu Phung
Primary Advisor's Department
Stander Symposium project, College of Arts and Sciences
United Nations Sustainable Development Goals
Peace, Justice, and Strong Institutions; Quality Education
"192.Apofasi - novel global events pattern detection engine" (2022). Stander Symposium Projects. 2477.