دانلود رایگان مقاله انگلیسی وب کاوی مبتنی بر الگوریتم Kohonen تک بعدی: تحلیل وب سایت رسانه اجتماعی - اشپرینگر 2017

عنوان فارسی
وب کاوی مبتنی بر الگوریتم Kohonen تک بعدی: تحلیل وب سایت های رسانه های اجتماعی
عنوان انگلیسی
Web mining based on one-dimensional Kohonen’s algorithm: analysis of social media websites
صفحات مقاله فارسی
0
صفحات مقاله انگلیسی
5
سال انتشار
2017
نشریه
اشپرینگر - Springer
فرمت مقاله انگلیسی
PDF
نوع مقاله
ISI
نوع نگارش
مقالات پژوهشی
رفرنس
دارد
پایگاه
اسکوپوس
کد محصول
E9313
رشته های مرتبط با این مقاله
مهندسی کامپیوتر، فناوری اطلاعات
گرایش های مرتبط با این مقاله
هوش مصنوعی، الگوریتم ها و محاسبات، اینترنت و شبکه های گسترده
مجله
محاسبات عصبی و برنامه های کاربردی - Neural Computing and Applications
دانشگاه
School of Management - Hangzhou Dianzi University - Hangzhou - China
کلمات کلیدی
الگوریتم Kohonen، رسانه های اجتماعی، داده کاوی، محاسبات عصبی
doi یا شناسه دیجیتال
http://doi.org/10.1007/s00521-016-2410-9
چکیده

Abstract


One-dimensional Kohonen’s algorithm is a process of mining knowledge which finds the characteristics of social media websites as a mode from the sequence database. Social media web log records generated constantly, and user access patterns will change accordingly. This study focused on taking advantage of the dynamic characteristics of the Kohonen algorithm, delivering a fast and efficient incremental mining algorithm and testing the new developed model.

نتیجه گیری

5 Conclusion


This paper has provided insights into the different accessing patterns of social media websites by improved Kohonen’s one-dimensional neural networks. The study confirms that better algorithms on social commerce users’ mining could enhance their acceptance patterns on social media different from previously literature. Current research adds new knowledge regarding time and neuron matrix in existing Kohonen’s models. The research helps practitioners and researchers better understand the different accessing characteristics between social media providers and users. Experiments with real and synthetic data sets are considered. A comparative study of the proposed networks with fuzzy c-means methods of the literature of symbolic data analysis for interval data was performed. The comparison was based on an external index, the overall error rate of classification and the number of iterations needed. For the synthetic data sets, these measures were estimated by the Monte Carlo simulation method. Continued research, development and evaluation are required to provide further understanding about other potential factors that may have an impact on the acceptance of social media services in colleges and to provide useful guidelines for marketers and product designers. The results pointed out that networks introduced in this paper outperformed the methods for these synthetic and real interval data sets regarding these clustering quality measures used.


بدون دیدگاه