Discovery of Collocation Episodes in Spatiotemporal Data
read more
Citations
CONSTAnT – A Conceptual Data Model for Semantic Trajectories of Moving Objects
DB-SMoT: A direction-based spatio-temporal clustering method
Mining frequent trajectory patterns in spatial-temporal databases
ST‐DMQL: A Semantic Trajectory Data Mining Query Language
Mobility Data Management and Exploration
References
Mining Sequential Patterns: Generalizations and Performance Improvements
Discovery of Frequent Episodes in Event Sequences
Levelwise Search and Borders of Theories in KnowledgeDiscovery
Discovering Spatial Co-location Patterns: A Summary of Results
An Evaluation of Non-Equijoin Algorithms
Related Papers (5)
Frequently Asked Questions (15)
Q2. What is the simplest way to join two ITLists?
The band-join is a straightforward extension of the merge join algorithm that replaces the equality condition by a maximum difference constraint (maximum time difference W in their example).
Q3. how many windows are used to count c?
Function slide window is used to count |Ic| (the number of windows that contain valid instances of c) for each candidate c ∈ C from a transformed sequence Sfi ∈ Sfr .
Q4. What is the method for finding close feature pairs?
To summarize, for finding close feature pairs, the hash-based technique is much faster than the linear scan method, whereas for discovering collocation episodes from feature sets, the Apriori method with the counting optimization technique performs best.
Q5. What are the default values of the data generation parameters?
The default values of the data generation parameters are n = 500, m = 2000, w = 2, W = 20, |F|= 40, = 7, = 1% and min sup = 0.03.
Q6. What is the naive method for the computation of Sfi for each Si?
A naive method for the computation of Sfi for each Si, is to scan all the other sequences in order to identify the windows and feature-sets in each Sfi .
Q7. How many times does a pattern with a reference type appear in a database?
Problem Definition: Given a database of trajectories S1, · · · ,Sn of n moving objects, each with type(oi) ∈ F , discover all the frequent spatiotemporal collocations, subject to , a closeness duration window length w, a maximum pattern window length W , and a frequency threshold min sup∈ [0, 1).
Q8. How many windows are there to count?
Since only three windows, [5, 13), [6, 14), [9, 17), correspond to the event of a feature-set entering the sliding window, and two windows, [11, 19), [12, 20), correspond to the event that a feature-set leaves the window, the authors just need to examine these five windows.
Q9. What is the definition of a spatial collocation pattern?
A spatiotemporal collocation pattern (or episode) P is an ordered list of spatiotemporal collocation units: g1g2 · · · g where ∩ i=1(gi.V ) = ∅.
Q10. What are the parameters used to generate trajectories?
Given these parameters, the authors generate n trajectories, each of which is assigned to a type in F while making sure that the generated trajectories instantiate collocation episodes.
Q11. What is the definition of sliding window counting?
Sliding window counting for event episodes has also been proposed in [5], however, the valid instances in their case are more difficult to count, because of the constraint that one collocation unit instance should end before the beginning of the next one (see condition (ii) in Definition 6).
Q12. What is the process of combining two ITLists?
For counting the instances of a longer candidate pattern P (procedure MJ count cand), the authors slide a W -window along the two ITLists of the two subpatterns P1 and P2 that generate P , and merge-join the lists to create ITListfr(P ).
Q13. How many times can a vulture move near to a deer?
An exemplary collocation episode for this application could be “Once the authors detect that a puma is moving close to a deer for 1 minute, the authors expect that a vulture will also move near to this deer in 3 minutes with high probability.
Q14. What is the cost of examining a window?
As a result, the cost of examining a feature-set sequence Sfi becomes proportional to |S f i |, instead of the number of window positions (which normally is much larger).
Q15. What are the two methods used to extract the collocation episodes?
the authors provide two collocation episode mining algorithms (one Apriori-based approach and one that is based on the vertical mining paradigm) and some pruning techniques to improve the mining efficiency.