Improving social bookmark search using personalised latent variable language models

Social tagging systems have recently become very popular as a method of categorising information online and have been used to annotate a wide range of different resources. In such systems users are free to choose whatever keywords or "tags" they wish to annotate each resource, resulting in a highly personalised, unrestricted vocabulary. While this freedom of choice has several notable advantages, it does come at the cost of making searching of these systems more difficult as the vocabulary problem introduced is more pronounced than in a normal information retrieval setting. In this paper we propose to use hidden topic models as a principled way of reducing the dimensionality of this data to provide more accurate resource rankings with higher recall. We first describe Latent Dirichlet Allocation (LDA), a simple topic model and then introduce 2 extended models which can be used to personalise the results by including information about the user who made each annotation. We test these 3 models and compare them with 3 non-topic model baselines on a large data sample obtained from the Delicious social bookmarking site. Our evaluations show that our methods significantly outperform all of the baselines with the personalised models also improving significantly upon unpersonalised LDA.