Topics generated by topic models are typi cally presented as a list of topic terms. Au tomatic topic labelling is the task of gener ating a succinct label that summarises the theme or subject of a topic, with the in tention of reducing the cognitive load of end-users when interpreting these topics. Traditionally, topic label systems focus on a single label modality, e.g. textual labels. In this work we propose a multimodal ap proach to topic labelling using a simple feedforward neural network. Given a topic and a candidate image or textual label, our method automatically generates a rating for the label, relative to the topic. Ex periments show that this multimodal ap proach outperforms single-modality topic labelling systems.
展开▼
