Topic modelling has been a successful technique for text analysis for almost twenty years. When topic modelling met deep neural networks, there emerged a new and increasingly popular research area, neural topic models, with nearly a hundred models developed and a wide range of applications in neural language understanding such as text generation, summarisation and language models. There is a need to summarise research developments and discuss open problems and future directions. In this paper, we provide a focused yet comprehensive overview of neural topic models for interested researchers in the AI community, so as to facilitate them to navigate and innovate in this fast-growing research area. To the best of our knowledge, ours is the first review on this specific topic.
Crowdsourcing is a new way for online crowds to get involved in a company’s research and development process. Businesses can host public contests on online platforms (such as Kaggle, Topcoder, and Tongal) to seek new product ideas and technological solutions. In the contest communities, members usually have a “coopetitive” relationship: they compete against each other for the contest prize, while at the same time also cooperate with each other by sharing information and knowledge. This work investigates the effect of knowledge sharing in such crowdsourcing contests. Surprisingly, we find that the knowledge sharing process may not always help improve crowdsourcing contestants’ performance. The effectiveness of knowledge sharing is influenced by the volume, quality, and generativity of shared knowledge. Shared knowledge is only beneficial when it is of high quality or when it has high potential of being further developed collectively by the community. Meanwhile, the development process has to be diverging; narrowing the development process in one direction can restrict the community creativity and negatively influence crowdsourcing performance. Our work informs the crowdsourcing practitioners to be more cautious when they enable collaboration such as knowledge sharing for the contest community.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.