Model bloggers' interests based on forgetting mechanism

Abstract

Blogs have been expanded at an incredible speed in recent years. Plentiful personal information makes blogs a popular way mining user profiles. In this paper, we propose a novel bloggers' interests modeling approach based on forgetting mechanism. A new forgetting function is introduced to track interest drift. Based on that, the Short Term Interest Models (STIM) and Long Term Interest Models (LTIM) are constructed to describe bloggers' short-term and long-term interests. The experiments show that both models can identify bloggers' preferences well respectively.

Personalized Ranking Algorithm Based on User Interest Modeling

“…In addition, we adopt [6] to evaluate the personalized ranking algorithm we propose, defined as follows: (9) where is the order of the web page in the search result. If the web page is clicked by the user, then , otherwise .…”

“…As a result, we use a triplet by introducing the time factor to represent each dimension of the user interest model, where is the latest time that is updated. We introduce a forgetting factor to simulate the user interest attenuation [6]:…”

“…As a result, we set the value of nouns to be 1.5 and other words 1. BM25 is improved based on this and the relevance between query Q and web page D is now calculated as: (6) where is the value of . Generally, users are always interested in the information of a certain field.…”

A Microblog User Interest Model Based on Participated and Not Participated Data

“…Aiming at the problem of user interest drift and user profile optimization, the paper [5] uses classification errorrate to track user interest drift and optimizes time window size automatically with user profile. The paper [6] proposed a user interest model based on forgetting mechanism. Short Term Interest Models (STIM)and Long Term Interest Models (LTIM) describe user's interests with weak and strong stability individually.…”

“…User interest models based on forgetting mechanism follow the ideas that human interests wane as time goes by, forgetting speed slows down gradually and the accumulated interests become more stable [6]. As given in Equation 6 and 7, t denotes the micro blog's time attribute, T means the current time and F(t, T) denotes the function satisfying the features of Ebbinghaus forgetting curve [10].…”