SENTIGRADE: A SENTIMENT BASED USER PROFILING STRATEGY FOR PERSONALISATION

Journal of Engineering Science and Technology

卷:13

BIKRAM KESARI RATHA¹  SIPRA SAHOO¹ 

[1] SOA Deemed to be University, Bhubaneswar, Odisha, India;

关键词: Folksonomy data;    Personalization;    Probabilistic framework;    Recommendation;    User Profiling;   

DOI  :  

来源: DOAJ

【 摘 要 】

Nowadays, the availability of folksonomy data is increased to make importance for user profiling approaches to provide results of the retrieval data or personalized recommendation. The approach is used for detecting the preferences for users and can be able to understand the interest of the user in a better way. In this approach, the incorporation of information with numerous data which depends upon sentiment is implemented using a framework SentiGrade by User Profiles (UP) and Resource Profiles (RP) for user Personalized Search (PS). From the folksonomy data, the discovery of User Preference (UsP) is presented by a rigorous probabilistic framework and relevance method are proposed for obtaining Sentiment-Based Personalized (SBP) ranking. According to the evaluation of the approach, the proposed SBP search is compared with the existing method and uses the two datasets namely, Movielens and FMRS databases. The experimental outcome of the research proved the effectiveness of the framework and works well when compared to the existing method. Through user study, the evaluation of approaches and developed systems are made which shows that considering information such as relevance and probabilistic data in Web Personalization (WP) systems can able to offer better recommendations and provide much effective personalization services to users.

【 授权许可】

Unknown