Efficient Mining of User-Aware Sequential Topic and Event Patterns in Time Series Data

P.Manvitha,   G. Sunil Vijaya Kumar

Abstract Recently, micro-blogs such as Twitter are attracting more and more attentions all over the world. Micro blog messages are real-time, spontaneous reports of what the users are feeling, thinking and doing, so reflect users’ characteristics and statuses. However, the real intentions of users for publishing these messages are hard to reveal directly from individual messages but both content information and temporal relations of messages are required for analysis, especially for abnormal behaviors without prior knowledge. If illegal behaviors are involved, detecting and monitoring them is particularly significant for social security surveillance. We can verify the activities of students to analyze the growth and the economy level to understand the behavior of student. For example, a student may opt for part time while studying to manage his economy needs this prevent the students financial problems and helps for better studies. Below are the various activities of student those are used for analysis. (1).Higher Education (2).Training (3).Joblessness (4).Employment (5).Further education (6).School. In this paper, we are proposing a system , an efficient mining of User-Aware Sequential Topic and Events Pattern for time series data The proposed system will be used for time series data sets directly connected to source micro-blog applications (like twitter and Facebook) so it fetches the real time streaming time series data It will be optimized for large data sets with parallel processing The proposed system will be used to get the event level information that can be used for identifying the abnormal events those are performed illegally Keywords: Web mining, sequential patterns, document streams, rare events, pattern-growth, dynamic programming

Title:	Efficient Mining of User-Aware Sequential Topic and Event Patterns in Time Series Data
Author Name:	P.Manvitha, G. Sunil Vijaya Kumar
Abstract:	Abstract Recently, micro-blogs such as Twitter are attracting more and more attentions all over the world. Micro blog messages are real-time, spontaneous reports of what the users are feeling, thinking and doing, so reflect users’ characteristics and statuses. However, the real intentions of users for publishing these messages are hard to reveal directly from individual messages but both content information and temporal relations of messages are required for analysis, especially for abnormal behaviors without prior knowledge. If illegal behaviors are involved, detecting and monitoring them is particularly significant for social security surveillance. We can verify the activities of students to analyze the growth and the economy level to understand the behavior of student. For example, a student may opt for part time while studying to manage his economy needs this prevent the students financial problems and helps for better studies. Below are the various activities of student those are used for analysis. (1).Higher Education (2).Training (3).Joblessness (4).Employment (5).Further education (6).School. In this paper, we are proposing a system , an efficient mining of User-Aware Sequential Topic and Events Pattern for time series data The proposed system will be used for time series data sets directly connected to source micro-blog applications (like twitter and Facebook) so it fetches the real time streaming time series data It will be optimized for large data sets with parallel processing The proposed system will be used to get the event level information that can be used for identifying the abnormal events those are performed illegally Keywords: Web mining, sequential patterns, document streams, rare events, pattern-growth, dynamic programming
Cite this article:	P.Manvitha, G. Sunil Vijaya Kumar , " Efficient Mining of User-Aware Sequential Topic and Event Patterns in Time Series Data " , International Journal of Emerging Trends & Technology in Computer Science (IJETTCS), Volume 6, Issue 4, July - August 2017 , pp. 201-211 , ISSN 2278-6856.
Full Text [PDF] Home