Citation Context

...words: Time-aware Search Task; Search Personalisation; Latent Topics 1. INTRODUCTION Recently, search personalisation has been an active research area and attracted increasing attention in literature =-=[1, 2, 4, 5]-=-. Previous research has shown that mining and modelling search task, which represents an atomic user information need [2], helps improve the performance of web search personalisation [2, 5]. In the co...

Citation Context

...words: Time-aware Search Task; Search Personalisation; Latent Topics 1. INTRODUCTION Recently, search personalisation has been an active research area and attracted increasing attention in literature =-=[1, 2, 4, 5]-=-. Previous research has shown that mining and modelling search task, which represents an atomic user information need [2], helps improve the performance of web search personalisation [2, 5]. In the co...

Citation Context

...words: Time-aware Search Task; Search Personalisation; Latent Topics 1. INTRODUCTION Recently, search personalisation has been an active research area and attracted increasing attention in literature =-=[1, 2, 4, 5]-=-. Previous research has shown that mining and modelling search task, which represents an atomic user information need [2], helps improve the performance of web search personalisation [2, 5]. In the co...

Citation Context

...ines: - LongTerm, ShortTerm are similar to our proposed reranking model. However, instead of modelling search tasks, these methods construct long-term1 and short-term2 user profiles respectively (see =-=[3]-=- for more detail). - StaticTask is the non-temporal search task modelling method (that is, the decay parameter α = 1). Overall Performance We evaluate our proposed method by comparing the original ran...