The development of information technology is increasingly rapid, such as social media, which has much influence. Social media is a place or media used to express and express various opinions on a topic. One example is Instagram. Instagram is a social media platform with many features, such as posting photos, videos, comments, likes, and others. The comments feature that Instagram has contained much public opinion that can be used as data. Nothing but the post on the SMB Telkom University Instagram account about the entrance to the university. In posts about the entrance to Telkom university, many Instagram users comment on the post. This can be convenient for the marketing team to get topics or discussions that most followers need from Telkom University's Instagram account. Therefore, a topic modelling of Instagram users' perceptions of comments posted on the entrance to Telkom university was carried out using the Nonnegative Matrix Factorization (NMF) method. After doing several research scenarios, the best coherent value was obtained with a coherent value of 0.60628 and the best 4 topics.

Social media has become a medium for communication between individuals and aspects of the business, including decision-making processes, brand promotion, brand marketing, and personal branding. One of them is Instagram. Using the comments feature on Instagram, users can communicate and give opinions on an upload on an Instagram account. Sentiment analysis can be done to analyze comments on the LaC (language center) Instagram account to measure student satisfaction sentiment towards Telkom university's LaC (language center) services. This study aims to analyze the sentiment or opinion of student satisfaction with the Telkom University Language Center (LaC) service on Instagram. The author also performs a classification based on positive sentiment, negative, and neutral categories using the Recurrent Neural Network (RNN) method and the Confusion Matrix measurement. From the test results on the model built to get an accuracy value of 79%.

YouTube social media is one of the popular media for all people to become a platform as a means of information and expressing opinions. Opinions can be categorized as hate if they attack something targeted. Hate speech is a behavior, word or action that is prohibited, because it causes violence to any individual and group. Expressing opinions in the form of hate speech is a problem that is still very difficult for the authorities to overcome because it is very common. Therefore, in this study a system was created to detect hate speech in the youtube comment column, using the Long Short-Term Memory and Latent Dirichlet Allocation. In this study, several methods were carried out that aimed to get the best accuracy value and carried out the topic modeling process using Latent Dirichlet Allocation to produce a total of three topics containing words that often appear in youtube comments. Based on the tests that have been obtained, the best accuracy is 0.657 or 66%.