It is necessary to have blogging technology which automatically figures out context information from the user in order to provide efficient micro blogging service for users in the social network service (SNS) environment. This paper proposes a context-based micro blogging service which considers the situations and emotions of users. To achieve this, a preprocessing method has been modeled based on users’ location and time context by using a granular tree, and Naive Bayes Classification has been adopted to assess a user’s behavior on the basis of a modeling context. In addition, a questionnaire has been administered to gain information about a user’s emotion in different situations and which considers location by using the Analytic Hierarchy Process (AHP). Based on this, a blog-able single sentence generation and auto blogging have been performed. The evaluation result of blogging a sentence and user’s emotional information shows 85.4% and 82.6% of accuracy, respectively; therefore, the proposed context modeling method for auto blogging is both efficient and effective.
[1]
Yao,et al.
The rise of granular computing
,
2008
.
[2]
H. Conte,et al.
Circumplex models of personality and emotions
,
1997
.
[3]
Tapio Seppänen,et al.
Bayesian approach to sensor-based context awareness
,
2003,
Personal and Ubiquitous Computing.
[4]
Zengchang Qin,et al.
Naive Bayes Classification Given Probability Estimation Trees
,
2006,
2006 5th International Conference on Machine Learning and Applications (ICMLA'06).
[5]
Sung-Bae Cho,et al.
Bayesian Inferrence and Context-Tree Matching Method for Intelligent Services in a Mobile Environment
,
2009
.
[6]
Altan Koçyigit,et al.
Granular Best Match Algorithm for Context-Aware Computing Systems
,
2006,
ICPS.