IMPROVING ONLINE MEETING EFFICIENCY USING LATENT DIRICHLET ALLOCATION (LDA) AND SOCIAL NETWORK ANALYSIS (SNA) METHODS

The pandemic period can change the habits of a person and organization, where all meetings are not held face-to-face/offline but virtually, so it is uncommon for to be attended by employees who Persons in Charge (PIC) on certain meeting topics. This study aims identify trends time, day, duration within Secretariat General Ministry Finance cluster matters into several themes that further identification carried out provide recommendations units having duties related using networking analysis. uses Natural Language Processing (NLP) method with Latent Dirichlet Allocation (LDA) which conclude factors represent topics produce topic clustering Social Network Analysis (SNA) modeling Degree Centrality find closest relationship between names. unit based highest centrality value possibility attending discusses particular topic. Data used this reseacrh during April 2020 up 2022 59,891 data records. results shows result dashboard an analysis often discussed become concern. research expected leaders assign dispositions each PIC attend meeting.