Information Technology Reference
In-Depth Information
user groups may induce greater acceptance of decisions [13]. The proximal social
context enables relevant information exchange that may also provide some clues that
draw user's attraction. The assertiveness and achievement of contributors would also
become the essential incentives for user to collaborate with proximal others.
The remaining sections are shown as follows. In section 3, we explore both
context information and content data from leisure entertainment participants. The
TF/IDF and CTD (Category Term Descriptor) methods for leisure information are
introduced and applied for recommendatory service. We introduce a leisure enter-
tainment recommendatory e-service that is designed based on the proximal social
intelligence in section 4. The evaluation of the recommendatory methods and ma-
nagerial implications are shown in section 4, too. Finally, a conclusion and future
directions of our work are provided in Section 5.
3 Exploring the Proximal Social Intelligence
Social network intelligence reserves rich personal information according to user's
social context. If the reserved information is utilized properly, users can obtain
important information from user peers within the same social context for quality de-
cision. The appropriate utilization of this collective intelligence could leads to exten-
sive knowledge enhancement for its domain. Shops and government can utilize
those information for improving their provided product and service. Customers
could also benefit from other customer's opinions, thus form a collaborative and
healthy context environment. In the collaborative leisure recommendatory service,
users can devote their up-to-the-minute personal experience as the input of the ser-
vice. The provided personal experience are deposited in text format and stored as a
tag. By gathering personal feedbacks acquired from the proximal social context for
progressive mining technique, the leisure recommendatory service will obtain tre-
mendous quality information for user to improve the overall decision quality.
In this chapter, we provide a leisure recommendatory e-service that allows the us-
ers to provide representative description of their perceptual experience regarding the
leisure related events they encountered. Next, we use these user's perceptional de-
scriptions as the hints to introduce the target event. The leisure entertainment par-
ticipants can review certain initial concepts from others with similar social context.
The provided tags are presented according to different methods to deliver an over-
view for specific target.
Two collaborative text mining techniques are applied in this leisure recommen-
datory service. The TF-IDF (Term Frequency-Inverse Document Frequency) and
CTD (Category Term Descriptor) are utilized for extract useful personal feedback
information for user to shape their knowledge and improve the decision quality.
The two methods are elaborated as follows.
3.1 The TF-IDF Method
Term Frequency Inversed Document Frequency, or abbreviate as TF-IDF is one of
the most popular term weighting schemes in information retrieval. The concept of
Inverse Document Frequency (IDF) was proposed by Spark Jones, K in 1972 for
