-
Notifications
You must be signed in to change notification settings - Fork 0
jumbrich/semqtree
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
How to use the benchmark suite: 1) Overview about available benchmarks #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu org.semanticweb.bench.Menu <benchmark> where <benchmark> one of StmtInsert Benchmark the time to insert X stmt Lookups Benchmark the lookup time and quality for BGPs WebFetch TBD GenerateBGP Randomly select BGPs from data 2)GenerateBGPs #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu GenerateBGPs usage: parameters: -h print help -i <arg> input location -if <arg> input format (idx, dir) -is <arg> number of estimated input stmts (estimation) -o <arg> output file -p <arg> Pattern for BGP (-1,0). -1 = variable, 0 = select, sperated by ",", e.g. only predicates (-1,0,-1) -s <arg> number of patterns Example # Generate a sampling of 100 items for the BGP class ?var <p> <o> # The sampling should be taken from files in the folder "input/linked-data" which contains around 110000 stms (this # is important to get a equal distributed sampling) #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu GenerateBGPs -i input/linked-data/ -if dir -is 110000 -s 100 -p -1,0,0 -o input/100_po_bgps 3)StmtInsert #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu StmtInsert usage: parameters: -b <arg> number of buckets [default 1000] -f <arg> fanout value [default 20] -h print help -i <arg> input location -if <arg> input format (idx, dir) -o <arg> output location -s serialise the create qtree (flag) The output folder will contain time and used-memory measures and the relevant gnuplot files to generate some graphs Example #Insert all stmts from files in input/linked-data into a QTree with bucket size 10000 and fanout of 30, add the end serialise the QTree into the output folder #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu StmtInsert -i input/linked-data/ -o tmp -f 30 -b 10000 -if dir -s 4)Lookups #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu LookupsMissing required argument i/o/if! -d <arg> original data location -h print help -i <arg> qtree location -if <arg> input format of the original data(idx, dir) -l <arg> file containing new line seperated basic graph patterns -o <arg> output location Output folder contains the time and quality measure for each BGP lookup and the average for all lookups Example #Test the 100 random selected BGPs for the created QTree. The data to verify how many sources were really required are taken from input/linked-data #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu Lookups -i tmp/_qtrees/qtree4all--hprefix_b10000_f30.ser -if dir -d input/linked-data/ -l input/100_spo_bgps -o tmp/
About
Automatically exported from code.google.com/p/semqtree
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published