DAO-LGBM: dual annealing optimization with light gradient boosting machine for advocates prediction in online customer engagement

【Author】 Abu-Salih, Bilal; Alotaibi, Salihah; Abukhurma, Ruba; Almiani, Muder; Aljaafari, Mohammed

【Source】CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS

【影响因子】2.303

【Abstract】Social networks have modernized the way people communicate, share information, and consume content. The widespread use of social media platforms has resulted in the creation of vast amounts of user-generated content, which can be analyzed to gain valuable insights into customer behaviour, emotions, preferences, and trends. Previous studies on online customer engagement have mainly focused on brand perspective and its socially significant elements, such as brand personality, image, reputation, and loyalty. These studies have explored how these elements influence the behavioural engagement of customers, such as their purchase intentions, word-of-mouth recommendations, and repeat purchases. However, more recent research has started to shift towards a more customer-centric perspective, which acknowledges that customer engagement is a two-way process, involving both the brand and the customer. This approach considers the role of customer experiences, emotions, topics of interest, and motivations in shaping their social engagement with the brand. This paper contributes to these endeavours by developing a consolidated framework that incorporates various facets of the customer's emotional and behavioural social content. In particular, features of online customers have been extracted using various sophisticated modules that incorporate natural language inference, topic modelling, sentiment analysis, emotion detection, and the Big-Five Personality Traits. Further, a heuristic-based feature selection (FS) strategy, Dual Annealing Optimisation (DAO), is integrated with Light Gradient Boosting Machine (LGBM) to furnish a consolidated machine learning module (DAO-LGBM) that is implemented and examined to detect advocates in online customer engagement. A thorough examination of a proposed model and its utility for detecting advocates using rigorous evaluation metrics is undertaken, reported, and discussed. These findings have substantial implications for both academic research and practical applications in social media analytics.

【Keywords】Online customer engagement; Customer advocacy; Dual annealing optimisation; Light gradient boosting machine; BERT; NLP; Feature selection

【发表时间】2024 2024 JAN 13

【收录时间】2024-01-28

【文献类型】

【主题类别】

--