A comparison of models for automating the attendance of online classes is presented in this research. The inputs to the models are the registration list and the participation list for certain days. The registration list contains name, registration number, class, section, contact number, etc. The participants list in general consists of the names of the attendees which are extracted from the platform of the online class. The aim is to match the names in the participation list to the names in the registration list appropriately. This problem leads to approximate string matching since most participants join online classes using unofficial email addresses, and thus the names mentioned in those emails may not exactly match the names on the registration list. The reasons for mismatch include substitution of part of the name by abbreviation, transposition, white space, etc. In this context, three models, namely, Longest Common Subsequence (LCS), Levenshtein, and Fuzzy Token Set (FTS) are applied to map the names appropriately. The models are tested on five real sets in the dataset containing 53 observations in total. The efficacy of these models is judged through descriptive statistics and ANOVA test. The median success matching rate of all models except FTS is at least 80%. Moreover, ANOVA test and post-hoc analyses verify that FTS has the lowest performance. The suggested approach of this paper is useful for a variety of online events, including webinars, workshops, meetings, etc.