چكيده شركت‎هاي فعال در حوزه سرويس‎هاي ارزش افزوده تلفن همراه، حجم بالايي از اطلاعات و داده را در بانك‎هاي اطلاعاتي خود ذخيره دارند كه شامل اطلاعات مشتركين، اطلاعات سرويس‎هاي ارزش افزوده مورد استفاده مشتركين مي‎باشند. به دليل بالابودن حجم اطلاعات، استخراج اطلاعات مفيد و مورد نظر امري مشكل و زمان بر است لذا استفاده از مفاهيم كلان داده و داده‎كاوي و بهره‎گيري از نرم افزارهاي ارائه شده جهت كار با داده، ما را در اين امر ياري مي‎كنند. در اين تحقيق كوشيده ايم با استفاده از متدولوژي CRISP، اكوسيستم هدوپ ، امكانات ارائه شده در محيط كلودرا و امكان خوشه‎بندي درنرم‎افزار كلمنتاين به كشف دانش پنهان در داده‎هاي يكي از شركت‎هاي ارسال پيامك انبوه جهت شناسايي مشتريان پرسود، شناسايي جذاب‎ترين سرويس‎هاي پيامكي براي مشتركان و سودآورترين حوزه‎هاي تامين محتوي بپردازيم. واژه‌هاي كليدي: تكنيك هاي داده كاوي- كلان داده- كلودرا- سرويس هاي ارزش افزوده- سودآوري- پيامك متني- مخابرات تلفن همراه

1396/03/16

1/1/1900 12:00:00 AM

Abstract Developing a model for classification of mobile value-added services to maximize customers’ profitability using data mining techniques BY: Alieh Boubehrezh The telecom industry is one of the first industries which has used Data Mining methods. These companies produce very high volume of data and they have a large number of customers and operate in a highly competitive market. Customer relationship management has become increasingly important because of instant competition in telecom sector, so disregarding Data Mining methods in telecom sector can lead the company to failure. Using these methods can help the company to have concentrated marketing plans and to increase the number of loyal customers in order to make profitability. In this thesis, we work on a Value Added Services (VAS) Company which send short massages to customers. After introducing the most common data mining methods which can be used in telecom sector, we tried to eliminate defective records and unimportantant colomns. After that we randomly choosed two groups of transactions in order to find the profitable customers, services and consumption patterns of profitable customers. Results can be used in future researches in order to find association rules in transactions and to predict the customers’ consumption trends. Keywords: Data mining techniques- Big Data- Cloudera- Profitability- Value added services- SMS- ICT Sector