Big Data

201Efficient Frequent Sequence Mining on Taxi Trip Records

as indicated by their high BCs. There are 2,620,636 road segments in the 17,558 short-

est paths (∼150 segments per path on average) that only take a fraction of a second to

compute BCs. While the approach is efcient and provides valuable information for

visualization purposes, it cannot be used to identify frequent sequences directly.

10.4.2 results of assoCiation rule mining on original sequenCes

Table 10.1 lists the runtimes and several other measurements output by the Sequoia

package using support values ranging from 50 to 5. Here, candidate frequent sequences

refer to sequences consist ...