Title: Research on personalised short video push on social media platforms based on affinity propagation clustering

Authors: Miao Wang

Addresses: School of Culture, Tourism and International Education, Henan Polytechnic Institute, Nanyang, 473000, China

Abstract: Personalised short video push on social media platforms can help enterprises improve user experience, competitiveness, and marketing effectiveness. This article proposes a personalised short video push method on social media platforms based on affinity propagation clustering. By determining the attractiveness, belonging, and reference between data points, the optimal clustering centre of the affinity propagation clustering algorithm is selected to achieve user behaviour data collection. Based on the data collection results, Markov matrix is used to extract user sentiment labels, combined with sentiment labels and XGBoost model to predict user personalised preferences. The i Expand algorithm is used to determine user interest vectors and generate recommendation lists, achieving personalised short video push on social media platforms. The experimental results show that the maximum push accuracy of this method is 97%, the maximum time consumption is 97.4 ms, and the maximum satisfaction with push results is 98.6.

Keywords: affinity propagation clustering; social media platforms; short video; personalised; push; sentiment labels; XGBoost model; i Expand algorithm.

DOI: 10.1504/IJWBC.2024.142480

International Journal of Web Based Communities, 2024 Vol.20 No.3/4, pp.263 - 277

Received: 22 May 2023
Accepted: 10 Oct 2023
Published online: 04 Nov 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article