This research focuses on studying the effect of using social media platforms on customer behavior and business objectives in Jordan. The research chose three of the most famous platforms of social media and those platforms are Facebook, Instagram, and Twitter. By using a quantitative model this research collected around 350 research questionnaires using digital surveys designed by using Google drive and distributed online on previous social media platforms. The research analysis process is executed by using AMOS software, and this comprised structural equations (SEM) modeling and regression analyses. The research study output found strong effects and relationships between social media platforms engagement and a variety of consumer expenditure variables, including brand loyalty, product suggestions, and purchase decisions. Important components of each platform, including user interaction, influencer endorsements, and content relevancy, were also identified by the study as having a direct effect on consumer behavior. The research also showed how businesses may achieve their marketing objectives, boost customer engagement, and enhance their reputation by utilizing social media strategies. The significance of social media website platforms in shaping consumer behavior and propelling business success is highlighted by these studies. To reach their goals and have the most influence on customer decisions, businesses should think about using platform-specific features and investing in targeted social media advertising.

Nonhomogeneous Continuous-Time Markov Chain (NH-CTMC) is a stochastic process that can be used to model problems where the future state depends only on the current state and is independent of the past. The transition intensity in NH-CTMC is not constant but is a function of time. In this paper, NH-CTMC is employed to model information dissemination on Twitter, where transitions occur only from followee groups to follower groups. Information is considered spread on Twitter when followers retweet posts from their followees. The tweet-retweet process on Twitter satisfies the Markov property, as a retweet from a follower depends only on the tweet posted just before by the corresponding followee. The probability of a tweet spreading is determined by the transition intensity, assumed to be a Sigmoid function whose parameters are estimated using Maximum Likelihood Estimation (MLE). This method is applied to Twitter data from Indonesia related to discussions on Covid-19 vaccination. The results indicate that information about Covid-19 vaccination on Twitter spreads rapidly from followees to followers in the first 20 hours, and then slows down after 40 hours. The NH-CTMC model outperforms the Homogeneous Continuous-Time Markov Chain (H-CTMC) approach, where the transition intensity (tweet spreading intensity) is assumed to be constant.

In this paper, a homogeneous continuous time Markov chain (CTMC) is used to model information diffusion or dissemination, also to determine influencers on Twitter dynamically. The tweeting process can be modeled with a homogeneous CTMC since the properties of Markov chains are fulfilled. In this case, the tweets that are received by followers only depend on the tweets from the previous followers. Knowledge Discovery in Database (KDD) in Data Mining is used to be research methodology including pre-processing, data mining process using homogeneous CTMC, and post-processing to get the influencers using visualization that predicts the number of affected users. We assume the number of affected users follows a logarithmic function. Our study examines the Indonesian Twitter data users with tweets about covid19 vaccination resulted in dynamic influencer rankings over time. From these results, it can also be seen that the users with the highest number of followers are not necessarily the top influencer.