With the prevalence of online social networking sites (OSNs) and mobile\ndevices, people are increasingly reliant on a variety of OSNs for keeping in\ntouch with family and friends, and using it as a source of information. For\nexample, a user might utilise multiple OSNs for different purposes, such as\nusing Flickr to share holiday pictures with family and friends, and Twitter to\npost short messages about their thoughts. Identifying the same user across\nmultiple OSNs is an important task as this allows us to understand the usage\npatterns of users among different OSNs, make recommendations when a user\nregisters for a new OSN, and various other useful applications. To address this\nproblem, we proposed an algorithm based on the multilayer perceptron using\nvarious types of features, namely: (i) user profile, such as name, location,\ndescription; (ii) temporal distribution of user generated content; and (iii)\nembedding based on user name, real name and description. Using a Twitter and\nFlickr dataset of users and their posting activities, we perform an empirical\nstudy on how these features affect the performance of user identification\nacross the two OSNs and discuss our main findings based on the different\nfeatures.\n