Usefulness Analysis for the Tool's Usage: Performance

[pedropaulofb]

Considering only taxonomies that do not have inconsistencies and divergences, plot 2 line graphs (for ac and for an execution modes) with 3 lines each and with error bars (standard error?).

Plotted Lines:

• Line A: average execution time for all rules (total_time of Times csv file) of all executions of all these 3 sets of taxonomies (NUMBER_OF_EXECUTIONS_PER_DATASET_PER_PERCENTAGE).
  • A1: using the set of 20% taxonomies of smaller sizes
  • A2: using the set of 20% taxonomies of intermediate size
  • A3: using the set of 20% taxonomies of bigger size

Axis:

• Y axis: execution time in ms
• X axis: percentage of classes used as input (from PERCENTAGE_INITIAL to PERCENTAGE_FINAL, increasing PERCENTAGE_RATE)

[pedropaulofb]

Determine calculations to be done.

Reevaluate this issue after analyzing the taxonomies.csv file to be created (Tester's unibz-core/Scior-Tester#20).

[pedropaulofb]

@mozzherina, you can perform this analysis in February.