Mining gps data

THE GOAL
 Discover interesting places from GPS data of 62 users.
 The data collected with Microsoft Research Asia,
in two years, 4115 days in china for the users
 3،497،750 GPS point
 43،18،284 km squared

Notations
 GPS Users: We have GPS logs of n users U = (u1,u2,………..,un).
 GPS Log: GPS points of the ith user, Gi = (gi1,gi2,………,gin).
 GPS Point: (Longitude, Latitude, Altitude, Date, Time).
 Stay Point: stay point Si is a geographical location where an individual
spent a significant amount of time.

ALGORITHM METHODS
1) Determine every user stay point
with algorithm (1).
2) Apply relational algebra operations
(Bag Union & Bag Intersection).
3) Apply Mode of Stay Points.

ALGORITHM(1)
 Algorithm1:
StayPointCalculation( GPSPoints,
ThresDistance,
ThresTime)
 Input: GPS points G={g1, g2,…., gn} ,
a distance threshold ThresDistance
and time span threshold ThresTime
Output: A set of stay points StayPts={S}

RELATIONAL ALGEBRA OPERATIONS
 Bag Union:
SPUnion= {Su1} UB{Su2} UB…………………… UB {Sun}
{ S ui} = Stay Points of User ui.
UB = Bag Union
 Bag Intersection:
SPIntersect = {S u1} ПB{Su2} ПB…………………… ПB {Sun}
{ S ui} = Stay Points of User ui.
ПB = Bag Intersection
 Mode of Stay Points: with (ThresCount)
M1 = Mode({SPUnion})
M2 = Mode({SPUnion} {M1})
M3 = Mode({SPUnion} {M1,M2})
Mn = Mode({SPUnion} {M1,M2,M3,…..,Mn-1})
SPMode={M1 , M2 , M3 ,…………, Mn}

APPLY ON DATA SET
Bag union
result

Mining gps data

More Related Content

Similar to Mining gps data (20)

Recently uploaded (20)

Mining gps data