International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
Mining of Spatially Co-Located Moving Objects by Using CTMSPMINE 1
1
K.Thanga Selvi, 2E.Baby Anitha
Student/M.E(CSE), KSR College of Engineering, India. 2 AP/CSE, KSR College of Engineering, India.
1. ABSTRACT In day to day life, vehicles have become important aspects in human life where each vehicle is manufactured for a particular purpose. Co-location pattern discovery is intended towards the processing data with spatial perspectives to determine classes of spatial objects that are frequently located together. Mining colocation patterns from spatial databases may disclose the types of spatial features which are likely located as neighbours in space. In the existing system they use FPtree to mine the spatial data. In this paper, I have presented an algorithm for mining spatially co-located moving objects using spatial data mining techniques. I propose a novel algorithm for co-location pattern mining which is used to identify the vehicles movements behaviour with the aid of Cluster based Temporal Mobile Sequential Pattern Mine Algorithm. Location Based Service alignment helps in finding the similarities between vehicles. An approach for Time segmentation is provided to find the time intervals where similar vehicle characteristics exist. In the experimental evaluation the proposed mining technique produces better results in spatial vehicle moving datasets.
Keywords: Spatial data mining, Co-location patterns, Frequent Pattern Tree, Location Based Service, CTMSPMine Algorithm, Vehicle movement data.
studies and thus are related to the efforts
2. INTRODUCTION
With the wide adoption of GPS tracking
system
and
towards a sustainable earth and ecosystem.
other
An automated discovery of spatial
telecommunication technologies, massive
knowledge is required because of the fast
amounts of moving object data have been
expansion of spatial data and
collected. Moving object data could be
use of spatial databases. Nowadays, the
related to human, objects (e.g., airplanes,
spatial data mining turn out to be more
vehicles and ships), animals, and/or natural
eminent and
forces (e.g., hurricanes and tornadoes).
that abundant spatial data
Although most human and man-made
stored in spatial databases. Spatial data
object movements are closely associated
mining (Mary and Kumar, 2012) is defined
with social and economic behaviors of
as the
people and society, movements of animals
interesting and previously unknown but
and changes of natural phenomena are
probably relevant patterns from spatial
often related to ecological and climate
databases. The mining of meaningful
stimulating for the reason have been
process of drawing out the
patterns from
ISSN: 2231-2803
extensive
spatial datasets is more
http://www.ijcttjournal.org
Page406
International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
knotty than mining the analogous patterns
vehicles and temporal periods are not
from conservative numeric and categorical
considered. In this paper, we propose a
data
novel algorithm, Cluster-based Temporal
(Kumar et al., 2012a), due to the
difficulty of spatial data
types, spatial
Mobile Sequential Pattern Mine (CTMSP-
relationships and spatial autocorrelation. In
Mine), to discover the Cluster-based
various applications, spatial patterns have
Temporal
excessive demand. To determine the
(CTMSPs). Mostly, a prediction strategy is
spatial co-location patterns in
wide
proposed to predict the subsequent mobile
applications (Yoo and Bow, 2011) is the
behaviors. In CTMSP-Mine similarities
main goal of spatial data mining. For both
between users are evaluated by the
positive and negative
association rules,
proposed measure, Location-Based Service
spatial
and
de-location
Alignment (LBS-Alignment) and a time
patterns are identical. The subsets of
segmentation approach is presented to find
Boolean spatial feature types are depicted
segmenting time intervals where similar
by spatial co-location
mobile characteristics exist.
co-location
patterns and its
instances are usually positioned in close geographic proximity Spatial de-location patterns extend
the conservative spatial
associations in order to
include an
association rules in the form of A→¬B, which denotes that B will not exist nearby A. In some spatial problems,
these
association rules are well-organized in discovering
useful
and
previously
unknown concealed information and also very
advantageous
(Saranya
and
Hemalatha, 2012).
Mobile
Sequential
Patterns
Literature Survey From spatial databases numerous research works exist in the literature focus on the subject of mining spatial co-location patterns. Nowadays, the developing
methods for co-location pattern mining has drawn a great concentration in real life applications. In large spatial datasets the significant
co-location and de-location
patterns have been mined
using a Co-
location and De-location patterns Mining algorithm (CODEM), proposed by Wan et
We have proposed an innovative algorithm in this study for efficaciously drawing out the spatially colocated moving objects from the spatial databases. In the existing system colocation patterns may not be precise enough for predictions since the
al. (2008). Here, by means of k-Nearest Features (k-NF) the spatial close/separate relationships
of
co-location/de-location
patterns in spatial datasets have been analyzed. To decide the close/separation relationship between this feature and other
differentiated mobile behaviors among
ISSN: 2231-2803
of
http://www.ijcttjournal.org
Page407
International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
features the k-NF set of one feature type’s
demonstrated that the proposed Grid
instances
Clique
have
been
employed.
algorithm
was
proficient
in
Subsequently, by applying a correlation
generating all maximal cliques in the
checking operation the irrelevant patterns
SDSS data and allows the discovery of
have been filtered. In addition, a grid index
relevant co-location patterns. The FPTree
Manikandan, G. and S. Srinivasan /
is used to find the group of vehicles that
American Journal of Applied Sciences, 10
are frequently moved in a particular area
(3): 195-208, 2013 Experimental results
has been proposed by Manikandan G,
have revealed that the proposed algorithm
Srinivasan S(2013).
was very potent in mining those patterns and
3. EXISTING METHODS
its time complexity was O (n). A
technique for discovering the co-location patterns
in Sloan Digital Sky Survey
(SDSS) data has been proposed by AlNaymat (2008). 3.6 TB of data was presented in SDSS Data Release 5 (DR5). Due to the presence of such giant amount of useful data, there is a possibility for the application of data mining methods to produce
interesting
information.
The
shortage of data in an appropriate format is the main reason for the scarcity of such data mining applications in SDSS. A procedure has been given to acquire more types of
In an Existing System, we don’t use
Cluster-based
Sequential
Patterns
Temporal
Mobile
(CTMSPs).
Since
Previous studies doesn’t consider the time as a important factor. Users have some specific behaviors in specific time and also data’s not maintained in a centralized way. A
mobile
transaction
database
is
complicated since a huge amount of mobile transaction logs is produced based on the user’s mobile behaviors. The main difference between these literatures is the involved
information
of
proposed
algorithm CTMSPMine.
galaxy from an available
attributes and the data has been converted into maximal cliques of galaxies that has been then used as transactions for data
4. PROPOSED SYSTEM
4.1 Algorithm for Mining Spatially CoLocated
mining applications. The maximal cliques from giant spatial
databases have been
Moving Objects
mined by using the proposed Grid Clique
availability
algorithm. NP-Hard represents the general
telecommunication and Web technologies,
problem of mining a maximal clique from
massive amounts of object movement data
a
have been collected from various moving
graph. Experimental results have
ISSN: 2231-2803
of
With the wide
http://www.ijcttjournal.org
GPS,
wireless,
Page408
International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
object targets, such as animals, mobile
shown in Table 1 that is an input to the
devices, vehicles and climate radars.
proposed algorithm. The ultimate aim is
Analyze such data has deep implications in
to find the spatially co-located moving
many applications, e.g., ecological study,
objects that will helpful in analyzing the
traffic control, mobile
communication
different classes of vehicles moving in an
management and climatologically forecast.
identical locations. Spatial co-location
In this study, we focus of our study on
pattern,
vehicle movement data analysis and
Coach)} shows that luxury car and bus are
examine
for
moved in a same location. This will
discovery of various vehicle movement
helpful in analyzing the market value of
patterns. It is common that objects follow
particular vehicles with respect to the
some regular movement patterns.
location. In order find the mostly moved
the
mining
methods
4.2 Conversion of Spatial Data into Vehicle List
{(Endeavor
and
Pajero, A/c
vehicles in a particular area, first we should identify the each vehicle classes and their types for the here the each class
In day today life, vehicles have become
vehicles have the unique id that are
important
describes in Table 1. In the spatial
aspects in human life where
each vehicle is
manufactured for a
temporal data consists of longitude value
particular purpose. In order to find the co-
and latitude value in
location patterns, we need to identify the
location of the object. The following
all
vehicles. For our convenience, we
Table 1 describes example of input
classify the vehicles into five classes; each
representation. Here we have object id and
class has some number of vehicles and
x coordinate value and y coordinate value
each vehicle have their unique id. The
as given in the following Table 2. Based
input for the
on this Table 2 the objects are located at
proposed algorithm is a
spatial database that contains three fields
order to find the
the following graph Fig. 1.
such as, instance id (classes), the spatial information (location (x, y)) and moving object id (1, 2 and 3). Here the instances are used to classify the
vehicles, the
4.3
Incorporate
the
R-Tree
Data
Structure with the Help of MBR in Every Grid
spatial information is used to represent the
Each grid gi in the spatial is treated
location of the corresponding vehicle and
as area, in that each grid is having some
the object id is used to find out the vehicle.
number of vehicles [n(vi)gi], in order to
An illustrated example for each class is
find the co-location, we need high density
ISSN: 2231-2803
http://www.ijcttjournal.org
Page409
International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
area since we assign the minimum
patterns by considering user clusters and
bounding
is
temporal relations in LBS environments
minimum number of vehicles in each grid
simultaneously. The main contributions of
mbr. Here the R-Tree (Francis and
this work are that we propose not only a
Thambidurai, 2007) data structure is use to
novel algorithm for mining CTMSPs but
satisfy the minimum bounding rectangle
also two nonparametric techniques for
mbr. If the grid is not satisfy the mbr then
increasing the predictive precision of the
the size of the grid is get increase gi in all
mobile
rectangle
mbr
which
users’
behaviors.
side in one unit, this process is repeat up to three times to satisfy the mbr. Here mbr is necessary to find the co-location of vehicles. If the grid gi has the number of vehicles equals to zero then no need to extend the size of the grid and no need to consider the corresponding grid. We take the vehicles from grid and make the vehicles list Vlist after the grid satisfied the mbr condition and the vehicles list
An example for a mobile transaction
Vhits consist of gird id and vehicles in the
sequence. (a) Moving sequences. (b)
corresponding grid.
Service sequences.
4.4 Using CTMSPMine This
Proposed
system
we
implement new technique called CTMSPs. To mine CTMSPs, we first propose a transaction clustering algorithm named Cluster-Object-based
Smart
Cluster
Affinity Search Technique (CO-SmartCAST) that builds a cluster model for mobile transactions based on the proposed Location-Based Service Alignment (LBSAlignment) similarity measure. To our best knowledge, this is the first work on mining and prediction of mobile sequential
ISSN: 2231-2803
The time interval segmentation method
helps
to
find
various
user
behaviors in different time intervals. For example, users may request different services at different times (e.g., day or night) even in the same location. If the time interval factor is not taken into account, some behaviors may be missed during specific time intervals. To find complete mobile behavior patterns, a time interval table is required. In this section, we describe our system design. Four
http://www.ijcttjournal.org
Page410
International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
important research issues are addressed
database according to the user cluster table
here:
and the time interval table.
1.
Clustering
of
mobile
transaction
sequences. 2.
Time
We proposed a clustering algorithm CO-Smart-CAST. Before performing the
segmentation
of
mobile
CO-Smart-CAST, we have to generate a
transaction sequences.
similarity matrix S, based on the mobile
3. Discovery of CTMSPs.
transaction database. The entry Sij in
Fig. 1b shows the proposed system
matrix S represents the similarity of the
framework. Our system has an “offline”
mobile transaction sequences i and j in the
mechanism for CTMSPs mining and an
database, with the degrees in the range of
“online” engine for mobile behavior
[0, 1]. A mobile transaction sequence can
prediction.When mobile users move within
be viewed as a sequence string, where each
the vehicle network, the information which
element in the string indicates a mobile
includes time, locations, and service
transaction.
requests will be stored in the mobile
timestamps of their mobile transactions are
transaction database. Table 1 shows an
more similar. Based on this concept, we
example of mobile transaction database
specifically design the time penalty (TP)
which contains seven records. In the
and the service reward (SR) in the LBS-
offline datamining mechanism,we design
Alignment. The base similarity score is set
two techniques and the CTMSP-Mine
as 0.5. Two mobile transactions can be
algorithm to discover the knowledge. First,
aligned if their locations are the same.
we
CO-Smart-CAST
Otherwise, a location penalty is generated
algorithm to cluster the mobile transaction
to decrease their similarity score. The
sequences. In this algorithm, we propose
location
the
0:5=ðjs1jþjs2jÞ, where js1j and js2j are the
propose
the
LBS-Alignment
to
evaluate
the
when
penalty
is
and
defined
Second, we propose a GA- based time
respectively. Notice that the maximal
segmentation algorithm to find the most
number of location penalties is js1jþjs2j.
suitable time intervals. After clustering and
When two sequences are totally different,
segmentation, a user cluster table and a
their similarity score is 0
time
Segmentation of Mobile Transactions
respectively.
Third,
are we
generated,
and
s2,
the
In a mobile transaction database,
CTMSP-Mine algorithm to mine the
similar mobile behaviors exist under some
CTMSPs from the mobile transaction
certain time segments. Hence, it is
ISSN: 2231-2803
propose
s1
as
lengths
table
sequences
orders
similarity of mobile transaction sequences.
interval
of
the
http://www.ijcttjournal.org
Page411
International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
important to make suitable settings for
patterns of our proposed algorithm are
time segmentation so as to discriminate the
done here.
characteristics of mobile behaviors under different time segments. We propose a GA-based method to automatically obtain the most suitable time segmentation table with common mobile behaviors. shows the procedure
of
our
proposed
time
segmentation method, named Get Number of Time Segmenting Points (GetNTSP) algorithm. The input data are a mobile transaction database D and its time length T. The output data are the number of time segmenting points. For each item, we accumulate
the
total
number
of
occurrences at each time point.
Mining Sequence
=
Algorithm 100
By
with
Input
analyzing
the
proposed spatial co-located vehicle mining algorithm with the help of the Synthetic datasets, evaluation
we
have measures
utilized
different
with
diverse
minimum support values. We have done the analysis part and plotted as a graph by computing the generated number of colocation vehicles, execution time and the memory usage with different minimum support. We have analyzed the results using
synthetic
datasets
with
input
sequences = 100 and the plotted graphs is The effective usage of the memory for
5. RESULTS & DISCUSSIONS
In this section, we conducted a series of experiments to evaluate the
mining the co-location vehicles in the proposed algorithm is shown in Fig. 2.
performance of the proposed CTMSPMine, under various system conditions. Experiments can be divided into two parts: 1) user
clustering,
2)
time interval
segmentation. All of the experiments were implemented in c#dotNet on a 3.0 GHz machine with 2 GB of memory running Windows 7. The experimental results of the proposed algorithm for spatially co-located moving
100% 80% 60% 40%
proposed2 existing
20% 0% no of values
Fig 1:Comparision Results
objects from the spatial databases are described here. The experimental results and analysis of the spatially co-located
ISSN: 2231-2803
http://www.ijcttjournal.org
Page412
International Journal of Computer Trends and Technology (IJCTT) – volume 9 number 8 – Mar 2014
6. CONCLUSION
REFERENCES
We have presented an efficient
[1] Manikandan, G. and S. Srinivasan An Efficient Algorithm for Mining Spatially Co-located Moving Objects, American
algorithm for mining spatially co-located
Journal of Applied Sciences, 10 (3): 195-208, 2013.
moving objects which materializes spatial [2] Al-Naymat, G., 2008. Enumeration of maximal clique for
neighbour relationship and reduces the memory with the help of the well known
mining spatial co-location patterns.
Proceedings of the
IEEE/ACS International Conference on Computer Systems and Applications, Mar. 31-Apr. 4, IEEE Xplore Press, Doha, pp:
FP
Tree
mining
and
CTMSPMine
algorithm. In first we split the area by applying grids afterwards the minimum bounding rectangle is used to find groups
126-133. DOI: 10.1109/AICCSA.2008.4493526 [3] Celik, M., J.M. Kang and S. Shekhar, 2007. Zonal colocation
pattern
discovery
with
dynamic
parameters.
Proceedings of the 7th IEEE International Conference on Data Mining, Oct. 28-31, IEEE Xplore Press, Omaha, NE., pp: 433-
of vehicles by elaborate the size of the
438. DOI: 10.1109/ICDM.2007.102
grid. Makes the vehicles list of each grid
[4] Francis, F.S. and P. Thambidurai, 2007. Efficient physical
afterwards remove the de-located vehicles
organization of R-trees using node clustering. J. Comput. Sci., 3: 506-514. Huang, Y. and P. Zhang, 2006. On the Relationships
with the help of minimum support and
between Clustering and Spatial Co-location Pattern Mining.
named that list as modified vehicles list.
Proceedings of the 18th IEEE International Conference on Tools with Artificial Intelligence, Nov. 13-15, IEEE Xplore Press,
FP tree constructed for the modified
Arlington, VA., pp: 513-522. DOI: 10.1109/ICTAI.2006.91
vehicles list afterwards the co-located
Jiang, Y., L. Wang, Y. Lu and H. Chen, 2010.
vehicles are mined from it. Finally
[5] Kumar, N.S., V.S. Ramulu, K.S. Reddy, S. Kotha and M.
CTMSPmine is used to find the vehicle
Kumar, 2012b. Spatial data mining using cluster analysis. Int. J. Comput. Sci. Inform. Technol., 4: 71-77.
behaviour and the time segmenting points [6] Kim, S.K., J.H. Lee, K.H. Ryu and U. Kim, 2012. A
are used to find the interval timing of the vehicles.
We
have
carried
out
the
framework of spatial co-location pattern mining for ubiquitous GIS. Multimedia Tools Applic. DOI: 10.1007/s11042-0121007-2 Kumar, G.K., P. Premchand and T.V. Gopal, 2012c.
experimental synthetic
evaluation
datasets
and
using our
the
proposed
method leads to reduce the memory usage extremely
when
compared
with
the
Mining of spatial co-location pattern from spatial datasets. Int. J. Comput. Applic., 42: 25-30. DOI: 10.5120/5836-7994. [7] Manikandan, G. and S. Srinivasan, 2012a. Mining spatially co-located objects from vehicle moving data. Eur. J. Sci. Res., 68: 352-366.
previous algorithm. From the results, we [8] Manikandan, G. and S. Srinivasan, 2012b. Mining of spatial
ensured
that the proposed
technique
outperformed of about more than 50% of
co-location pattern implementation by FP growth. Ind. J. Comput. Sci. Eng., 3: 344-348.
previous algorithm in time and memory usage.
ISSN: 2231-2803
http://www.ijcttjournal.org
Page413