دانلود رایگان مقاله ISI درباره داده کاوی،تقسیم بندی مصرف کننده و خوشه بندی c-means فازی clustering
دانلود رایکان مقاله انگلیسی ISI با موضوع تجزیه و تحلیل رفتارهای کاربر با کاوش مجموعه داده های شبکه ای بزرگ
عنوان فارسی مقاله:
تجزیه و تحلیل رفتارهای کاربر با کاوش مجموعه داده های شبکه ای بزرگ
عنوان انگلیسی مقاله:
Analysis of user behaviors by mining large network data sets
دانلود رایگان مقاله ISI با فرمت PDF:
مشاهده توضیحات کامل و خرید ترجمه فارسی با فرمت ورد تایپ شده:
بخشی از مقاله انگلیسی :
4. User behavior patterns
Based on the algorithm presented in Section 3, in this section,
we investigate user behavior patterns and extract salient characteristics
that indicate the relationships among users’ consumption
capacity, communication time, mobility, and locations
4.1. Communication behavior clustering
In order to study the communication behaviors of users, we
firstly divide the one million sample users into 3 levels based on
their ARPU (from high to low), then we cluster each level of users
by applying the FCM algorithm.
The proportion and total ARPU contribution of each level is
shown in Fig. 3. We know that the famous 20−80 rule in economics
indicates that 20% of customers contribute 80% of profits. However,
from Fig. 3 we can see that due to fierce competition, the gap
of total revenue between each level is narrowed. This indicates
that we cannot underrate any level of users. Applying the FCM
algorithm for each level of users, the clustering result is shown in
Fig. 4. The subgraphs in Fig. 4 present the results of different ARPU
levels, from high to low. For the sake of convenience, we define
the clusters as c1, c2 and c3 in a certain ARPU level, respectively.
The proportion of c1, c2 and c3 is illustrated on the left of each
subgraph.
From Fig. 4, we can see that for the top ARPU level, c1 has longer
outgoing and incoming calling durations in the morning than afternoon
or evening and night. This situation is reversed for c2. The
cluster c3 has sharp peak calling durations in the evening around
20:00 h. Generally, the top ARPU level users have longer outgoing
call duration than the incoming one, which indicates that they have
strong affordability.