Download Promoting Usage of Location-based Services, an Approach Based

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
Transcript
Promoting Usage of Location-based
Services, an Approach Based on
Intimacy Theory and Data Mining
Techniques
1. Introduction
• With the development of technology, nowadays, smartphones
have become an integral part of people daily lives.
• Accordingly, variously kinds of applications that provides
location-based services (LBS) have been developed in todays’
mobile service market.
• On the contrary, people’s interests towards LBS did not
increase correspondently, even shows a continuous decease
instead [
Trends of people’s interests on LBS
2. Materials and Methods
2.1 Data collection
• The data was collected from 220 participants in three
universities locate in Seoul, Korea.
• The data which contains 165 sample that finally used for data
analysis in this study includes 35 variables, including
demographic information, user mobile use behavior feature,
actual LBS use situation and attitudes regarding LBS usage
2.2 Test method
• Three steps were used to analysis the data: first, the Kmean clustering analysis was used to segment the mass
Smart-phone users into different groups based on their
smartphone usage characteristics and attitude variables.
• Second, decision tree analysis was conducted using SAS in
to identify the characteristics of users belong to different
clusters regarding the selected clustering criteria.
• Next, an inter & intra-clustering analysis were applied to
explore the latent antecedent factors associated with the
clustering criteria: Intimate Behavior and Intimate
Experience, as well as to generate specific association rules,
which can be used for designing suitable marketing
packages.
3. Results
3.1 Test result
• Based on the factors regarding users’ intimacy level to service
provider-intimate behavior intention and expectation to
receive the intimate experience, as well as the real usage
situation of LBS applications
• We segmented the 165 smart phone users using a clustering
algorithm. K-mean with a Newton algorithm was used for
clustering purpose, and the clustering number was be
selected as 5 in order to distinguish different clusters.
Distance between 5 clusters in terms of
selected criteria
Decision Tree Analysis-Intimacy Level for
Each Cluster
User Group Characteristics
Willingness to discl
Intention to use
User groups ose personal inform
LBS
ation
Cluster
Cluster
Cluster
Cluster
Cluster
1
2
3
4
5
Low
High
Low
Medium
High
Low
Low
High
Medium
High
One example of Association rule
analysis in cluster 2
Lift
Support Confidence( Rule: Antecedent ==> Cons
(%)
%)
equent
1.31
58.33
87.5
1.29
50
85.71
1.25
41.67
83.33
1.13
50
75
1.07
41.67
71.43
1.5
41.67
100
kaoka1 ==> Use intention AV
E2
Sns2 ==> Use intention AVE2
Used1 ==> Use intention AV
E2
recomm3 ==> Use intention
AVE2
facility2 ==> Use intention A
VE2
Heard3 ==> Use intention AV
E2
4. Discussion
• In this study, three main contributions are provided: first, a
user segmentation process was proposed from a CRM
perspective based on users intimacy level with service provide:
intimate behavior intention- users willingness to disclose
personal information and intimate experience expectationusers’ expectation to receive personalized service by using
location-based service.
• Second, regarding user’ personalized needs for locationbased services functions, different types of LBS application
were suggested to different user groups in terms of intimacy
level.