SWRL-IQ User Manual
Contents
1. To install this Java attachment first compile it into a jar file On Windows enter the following commands in the main SwrlIq java directory javac testi java jar cf hello jar test class On Linux or Mac use these commands javac test java jar cf hello jar test class Next copy hello jar to the Swrllq plugin directory under Prot g Restart Prot g Now enter the query swrlex callJavaStaticMethod res test BuiltinTest helloWorld John Doe This should return one result Tres Hello John Doe This is all it takes to create your own Java Attachments 8 3 Limitations and Caveats Java Attachments have a number of limitations when compared to predicates defined using SWRL rules Generally speaking Java Attachments do not have a well defined semantics and make things more procedural and less declarative They cannot be used backwards and all the input arguments must be fully instantiated before the attachment is called The user must understand the operational semantics of the reasoner to make sure this happens and rules containing Java Attachment calls must be written with the atoms in a certain order Java Attachments will not work with OWL SWRL reasoners other than SWRL IQ since they are our own extension Despite all this for some purposes they can be quite useful 29 SWRL IQ User Manual 9 ACKNOWLEDGMENTS The work described in this paper was carried out at the SRI facilities in2. Trace All Debug C Reload KB OWL SWRL Answer s wine sugar vin MariettaCabernetSauvignon vin Dry vin LongridgeMerlot vin Dry vin FormanChardonnay vin Dry vin CorbansDryWhiteRiesling vin OffDry vin ClosDeLaPoussieSancerre vin OffDry Show SPARQL XML Save as CSV Fig 1 Wine sugar query File Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help Deb B wada 9S Bala lt b gt prot g SWRL Query INSTANCE BROWSER For Class swrlu Query Asserted Instances xe o Ga 4 InPriceRange Oooo Ha hasPrice wine price withinRange price 10 50 Submit query Trace First Trace All Debug C Reload KB OWL SWRL Answer s wine price in ChateauMorgonBeaujolais 25 0 in PageMillWineryCabernetSauvignon 24 0 in MariettaCabernetSauvignon 20 0 in FormanChardonnay 39 0 in VentanaCheninBlanc 22 0 v Show SPARQL XML Save as CSV Fig 2 Wine price range query SWRL IQ User Manual Solutions Y 29 Solution 0 depth 2 hasPrice vin ChateauMorgonBeaujolais 25 0 assert Y withinRange 25 0 10 50 rule Within Range Rule swrlb greaterThanOrEqual 25 0 10 comp swrlb lessThanOrEqual 25 0 50 comp Y amp Solution 1 depth 2 hasPrice vin FormanChardonnay 39 0 assert Y withinRange 39 0 10 50 rule Within Range Rule swrlb greaterThanO
3. min is the smaller of the elements x and y according to the ordering predicate p Example swrlu minimum min 4 3 swrlb lessThanOrEqual returns x 3 The rules defining this predicate are discussed in Section 6 min list x 1 p Satisfied iff x is the smallest element in the non empty list 1 according to the ordering predicate p Example swrlu min list x rdf List 4 17 3 20 swrlb lessThanOrEqual returns 7X 3 There are also the corresponding predicates maximum max x y p 26 SWRL IQ User Manual max list x 1 p and these convenient shortcuts for numerical values min num min x y min num list x 1 max num max x y max num list x 1 sort sl 1 p Satisfied iff s1 is the version of the list 1 sorted in ascending order using the ordering predicate p Example Sorting numbers swrlu sort sl rdf List 4 17 3 20 swrlb lessThanOrEqual returns x 3 4 17 20 Example Sorting wines by their price First we define a predicate wineCheaperThan using a SWRL rule hasPrice w1 p1 hasPrice w2 p2 swrlb lessThanOrEqual pi p2 wineCheaperThan w1 w2 Then we can run the query swrlex allKnown wines wine hasPrice wine price A swrlu sort sortedWines wines wineCheaperThan which will return a list of wines that have a price sorted by price in the sortedWines variable 27 SWRL IQ User Manual 8 JAVA ATTACHMENTS Like many reasoning systems
4. Menlo Park CA and was funded by the U S Department of Defense TRMC Test and Evaluation Science and Technology T amp E S amp T Test and Evaluation Science and Technology Program under NST Test Technology Area prime contract N68936 07 C 0013 The authors are grateful for this support and would like to thank Gil Torres NAVAIR for his leadership of the Netcentric System Test NST technology area to which the ANSC project belongs We would also like to acknowledge ODUSD R RTPP Training Transformation for its sponsorship of the associated ONISTT project Reg Ford Mark Johnson and Stijn Heymans at SRI provided invaluable feedback on this manual and query testing Finally David Warren and Terrance Swift provided expert guidance and bug fixes for XSB Prolog and Miguel Calejo did the same for InterProlog 30
5. and dateTime Yes date time dateTime types can not be entered directly in a query Other SWRL time built ins swrlb resolveURI swrlb anyURI No swrlb swrlb swrlb swrlb swrlb swrlb listConcat listIntersection listSubtraction member swrlb length first swrlb rest sublist swrlb empty Yes Table 3 Summary of support for SWRL built ins 20 SWRL IQ User Manual Query Bindings returned Constraints returned swrlb equal x 3 7x 3 none swrlb notEqual x 3 7x Var0 Var0 3 swrlb lessThan x 3 A 7x Var0 y Varl Var0 lt 3 0 Varl Var0 lt 0 0 swrlb lessThan y x swrlb lessThan0rEqual x 3 A 7x 3 none swrlb greaterThan rEqual x 3 swrlb pow _sq x 2 A 7x 1 4142135623730951 none swrlb multiply 10 _sq 5 7x 1 4142135623730951 swrlb pow _sq x 2 A x 1 4142135623730951 none swrlb multiply 10 _sq 5 A swrlb greaterThanOrEqual x 0 Table 4 Examples of SWRL queries with constraints Query Bindings returned swrlb listConcat l rdf List 1 2 rdf List 3 4 71 1 2 3 4 swrlb length 2 1 A swrlb first 1 71 A 71 1 2 swrlb rest r 1 A swrlb first 2 r swrlb length 2 1 A swrlb member 1 1 7 L Var0 1 Var1 1 swrlb first 1 1 1 1 Var0 Table 5 Examples of SWRL queries with lists returns dt 2011 11 11T11 11 11 Note
6. appear in Prot g To test that the query engine works create a new query individual by clicking the create instance button in the instance browser on the top left of the SWRL IQ tab Then enter a simple query into the query text field e g swrlb add x 3 2 An answer to the query should appear on the bottom of the tab Note that the first query always takes a little while This is because the entire knowledge base has to be loaded into the Prolog back end Subsequent queries are much faster This is discussed in more detail later in this manual If this does not work look in the console window to see if there is any error message or Java exception If you need help solving the problem send email to swrl iq users lists esd sri com 10 SWRL IQ User Manual 3 BASIC USAGE SWRL IQ is a plugin to Prot g and uses some of the same user interface elements We recommend that users who are new to Prot g first familiarize themselves with it The Prot g web site contains a number of guides and tutorialf In addition learning how to use Prot g s SWRL rules talf will help in understanding our SWRL query editor 3 1 Queries The SWRL IQ user interface has three main parts The query instance browser the query editor and the query results panel see Figure 10 The query instance browser is a standard Prot g component Most importantly it allows you to create new query individuals and to select among existing ones Re nam
7. ii wa YS AEB lt p prot g Metadata food_planning owl OWLClasses BW Properties 4 Individuals Forms SWRLRules SWRL IQ SWRL Query swrlex callJavaStaticMethod r java lang Double toHexString 14 432 For Class swrlu Query oxg omo 22 Asserted Instances 4 Java ttachment El B8 Submit query TraceFirst Trace All Debug C Reload KB OWL SWRL Answer s r l Ox1 cdd2f1a9fbe77p3 Show SPARQL XML Save as CSV Fig 9 Java attachment SWRL IQ User Manual 2 INSTALLATION 2 1 Unzipping the plugin SWRL IQ is distributed in binary form as a zipped archive The first step of installing the tool is to unzip this archive This will create a folder named SwrlIq which contains this manual some sample ontologies a sample java attachment and a swrliq plugin zip file Unzip swrliq plugin zip into the Prot g plugins directory This creates a com sri swrltab directory with a few files in it 2 2 Installing XSB SWRL IQ needs a working installation of XSB Prolog to work The main XSB site does not supply sufficiently up to date binaries so we provide instructions below for building it from the source code We also provide binaries for Windows at http www onistt org display SWRLIQ If you get those you can skip sections through 2 3 Getting the Source Code First
8. pad query 2 http protege stanford edu doc users html 11 SWRL IQ User Manual in the query editor can be deleted before saving the OWL file so that it does not get saved unintentionally although saving it is not harmful other than making the OWL file slightly larger The query editor works almost exactly like the Prot g SWRL rule editor In terms of the user interface the main difference is the lack of the implication symbol In other words editing a query is much like editing the body of a SWRL rule The main purpose of a query is to find values for all the variables in the query such that the query is satisfied with respect to the current ontology rules and facts in the knowledge base Once a query is ready to run click the Submit query button the Trace First Trace All and Debug buttons are discussed in Section 3 2 There are three types of possible answers to the query SUCCEEDED FAILED or a table of variable bindings The first type can only happen when there are no variables in the query or when all variables are anonymous see section 3 1 2 The variable bindings table shows one row in the table for each solution to the query The table header shows the variable names and the values in each column are the different bindings for the variable in question Double clicking on a value in a cell containing an RDF resource opens up the standard Prot g individual edi
9. we d Po SEO Reque 19 EMITTE 21 6 SWEL Extensions aiso e iida ran eii oi Va e o wee e e qe per VEN eae Gaba Ed det 23 C SWRI Utilities rrr C ree a td a IUS ile Ga DD REA 26 B Java Attachrnents i ooi oe ete ptas ben nade ber bbb et et deer E E dei 28 8 1 Argument Conversions eizkue e het e A A VR D ERA T ER 28 8 2 Writing and Installing a Java Attachment seseeeeeeeee II 29 8 3 limitations and Caveats covemcrncrinic mars Ie a ud nade REY ra dS 29 v E ean ee aaa eee et 30 LIST OF TABLES iE Sine oie A eb eet SARE oa a ee BE due eat duel d ae 14 be A ate Sar al ETT 15 Poe E ones a cse d Ebr Up B or AA 20 MEME NAO 21 DIETE 21 SWRL IQ User Manual 6 OWL Java type conversions sese hh LIST OF FIGURES I Wine sugar quety 2x eiui pen ele ey a e eeu bre ia 2 Wine price range query sseeesee eR RR IRR AI 5 Query result in html using an xsl transformation 6 Query result saved to CSV and graphe All known regions query Sort wines by price query Java attachment 11 Saving XML Ulicisni ica rimar ara 10 Wine sugar query amp Query UI sssssssesseese ro E NRFNNOOTK WY bw SWRL IQ User Manual 1 OVERVIEW SWRL IQ Semantic Web Rule Language Inference and Query tool is a plugin for Prot g 3 x that allows users to edit save and submit queries to an underlying inference engine based on XSB Prolog The tool has a number of features that distingu
10. 4 Wine Properties Query result in Protege Example using a predicate as an argument swrlex allKnown _listOfWinePricePairs rdf List wine price hasPrice vine 7 price swrlu sort sortedListOfWines _listOfWinePricePairs winePairCheaperThan produces a list of the lists of all the wines and their prices sorted by price as shown in Figure S hover the mouse pointer over the result to see the blue box tool tip The sort predicate is discussed in Section 7 It takes a predicate here winePairCheaperThan as an argument and sorts the list according to the ordering defined by that predicate Example using a Java attachment swrlex callJavaStaticMethod r java lang Double toHexString 14 432 returns the result of executing the given Java method as shown in Figure 9 Any static method from the standard Java library can be used and users can also call their own code See Section 8 for more detail SWRL IQ User Manual query_result html 2 A A PR file Swrliq html query_result htmi Q7 Google o IReference 1452 zn FormanCabernetSauvignon T D NS E StonleighSauvignonBlanc nt Fig 5 Query result in html using an xsl transformation MariettaCabernetSauvignon RReference 14 SWRL IQ User Manual o B D E F G H maker color grape sugar pice Juri label d vin ChateauLa vin Red vin CabernetS vin Dry 1452 http www wi http sa3 wine vin Selaks vin W
11. SWRL IQ User Manual January 16 2012 Version 1 0 SWRL IQ User Manual Prepared by Daniel Elenius Susanne Riehemann SWRL IQ User Manual Table of Contents UELUT 1 1 1 Project Home Page and Mailing List llle 1 I TT T 1 buo dide ubi reb pb A e 8 8 8 Pc 8 RES 8 A T 8 A Windowsi 45x arden arden ce inan d dome aer op eder A deena eee NOR EEA EA 8 2 0 2 64 bit Windows iae REIR ER AT I REEL Vua dA Pe ed ead d yt 9 RH HM 9 A A A A AAA A id 9 E A A SU re E Tia dd 9 ZR 10 Rare iin sende eda due due Ais A s e dera ho ert opa le o ADR eie tas 10 erm 11 A A A heed bbe eee eu eae eed ooo Rede 11 9 1 1 Pscudo Bil las acia ads e ale bate dix yea hates ae HUS eh dle wae Sau A 12 3 1 2 Anonymous Variables 0 eee eee em 13 23 2 Tracing and Debugging xxx bois we Hen gea Eo i RE Ra A RES gd 13 AA EEDUTTTTTTUTTTTS 13 3 22 Debugging cv er PEE REOR EG hominy me aR REPRESENT S RD IE 14 3 3 Semantics and Limitationsi o oo oooocoorrocoororrror eee ee teen eee teen eens 14 ole ee AS 17 ee AA en NOU Edgy lide 17 ss ecc debo dier CR UR ER RP ag 18 A TE 18 5b SWRI Buit Ins sioe ee Ry vee ei x RE YR EYE TUE RENE A OH RA dcs 19 5 1 Constraints aia ai e REA Dede hohe de qw Nene gts 19 AA A es adi O83 duda eee eee ird b EE Gada Vinee dpud E m oe oed 19 A redet ed E UL DLL EMEN e I E E 19 Led hive pda d ees duce Feb e oe
12. SWRL IQ has a mechanism for calling arbitrary programming code from within the reasoner This is traditionally called procedural attachments In our implementation the code to be called has to be written in Java We call our mechanism Java Attachments Most computations can be performed in SWRL especially with our extensions see Section 6 One might wonder why an external procedural attachment mechanism is needed Indeed SWRL rules are in many ways more powerful and flexible see Section 8 3 below However there are situations where it does not make sense to use SWRL rules One such case is when a large and complicated function has already been implemented in programming code and perhaps certified Re implementing and re certifying in SWRL may not be feasible Indeed if the source code is not available and the algorithm is not known there is no choice in the matter The hook to the Java attachments mechanism is the following built in predicate defined in swrl extensions owl callJavaStaticMethod ret cls mthd arg1 arg2 ret is the return value cls is a string representing the class including the full package name mthd is a string for the static method to call and arg1 arg2 are the arguments to the method Example swrlex callJavaStaticMethod r java lang Math random returns Tr a random number from 0 0 to 1 0 As seen here any static method of the standard Java library can be used However user define
13. als can be typed directly in a query by enclosing them in double quotes For example Example swrlb upperCase 7 s foo returns s FOO 5 4 Dates Times and Durations These can be used both to generate and to disassemble the corresponding units Example swrlb dateTime dt 2011 11 11 11 11 11 tz 10 nttp www w3 org Submission SWRL 8 19 SWRL IQ User Manual Built in predicate Support Constr Notes swrlb equal swrlb notEqual Yes Yes Only for numerical datatypes swrlb swrlb swrlb swrlb lessThan lessThan0rEqual greaterThan greaterThan rEqual Yes Yes swrlb swrlb swrlb swrlb swrlb swrlb add swrlb subtract multiply swrlb divide pow swrlb unaryPlus unaryPlus swrlb unaryMinus abs swrlb sin swrlb cos tan Yes Yes swrlm sqrt swrlm log Yes Yes swrlb swrlb swrlb integerDivide swrlb mod ceiling swrlb floor round Yes No swrlb roundHalfToEven swrlb booleanNot Yes swrlb swrlb swrlb swrlb swrlb swrlb swrlb swrlb stringConcat stringLength startsWith swrlb endsWith upperCase swrlb lowerCase contains containsIgnoreCase stringEquallgnoreCase matches swrlb replace Yes swrlb swrlb swrlb swrlb swrlb normalizeSpace translate substringBefore substringAfter tokenize swrlb swrlb date swrlb time
14. d or third party code can also be called as explained below 8 1 Argument Conversions Java and OWL SWRL have vastly different type systems Java Attachments support neither all Java types nor all OWL types In fact we support what might be considered the intersection of the two viz the basic types of both with some obvious conversions Table 8 1 summarizes the supported types for the two sides and their conversions SWRL OWL type Java type xsd int xsd integer etc int xsd float xsd double xsd decimal etc double xsd boolean boolean xsd string String rdf List Object Table 6 OWL Java type conversions If integers floats etc are inside a rdf List they are mapped to the corresponding boxed Java type i e Integer Double or Boolean Note that there is no way to pass RDF resources individuals classes and properties to Java Attachments 28 SWRL IQ User Manual 8 2 Writing and Installing a Java Attachment Creating a new Java Attachment involves creating a new class with the new method to call compiling it and putting the jar file in the com sri swrltab directory under Prot g s plugins directory As an example see the java test BuiltinTest java file distributed with SWRL IQ Its full contents are shown below package test builtins public class BuiltinTest public static String helloWorld String firstname String lastname return Hello firstname lastname
15. e difference with the first example for swrlb pow in Table 3 In that case we got two different solutions When we use allKnown we get all the solutions gathered into a list The x parameter can also be a list of variables as in the wine example from Section 1 swrlex allKnown _listOfWinePricePairs rdf List _wine _price hasPrice _wine _price When used in this way the result is a list of lists of bindings of all the variables in the variable list If the goal contains additional variables that are not in the second parameter they are treated existentially For example swrlex allKnown wines wine hasPrice wine price in the food planning owl ontology returns one list containing all the wines that have some value for the hasPrice property This built in significantly extends the power of the language reasoner It provides a kind of local closed world assumption As the name of this predicate implies it returns all known solutions to a query The open world assumption of OWL and SWRL means that there could be additional solutions firstMatch res list p x1 x2 Returns the first element in the list that satisfies p x1 x2 The variable res is normally one of the arguments x1 x2 Example swrlex firstMatch x rdf List 1 2 3 4 swrlb greaterThan x 2 returns x 3 callJavaStaticMethod ret cls mthd arg1 arg2 This predicate is the hook to the Java Attachment mechanism Thi
16. e ordering predicate as less than or equal The minimum predicate is defined by the following two rules in swrl utilities owl and also documented in Section 7 swrlex apply p x y minimum x x y p swrlex apply p y x gt minimum y x y p The minimum predicate can now be used to determine the minimum of any types of numbers or individuals as long as the appropriate ordering predicate can be supplied This in itself may not seem that useful but the value of this predicate becomes clear when we consider that it is the basis for the more 24 SWRL IQ User Manual complex predicates min list and sort The latter can be used to sort a list of arbitrary types of RDF resources for example to sort a list of wines according to their price by passing a properly defined predicate wineCheaperThan as the parameter p Note that if the two values are equal according to p i e both p x y and p y x are true then we get two solutions This is useful for example the sort predicate produces all valid sortings of the equal elements All of these predicates can be used on plain numbers by passing swrlb lessThanOrEqual as the parameter p Section 7 contains examples of this allKnown res x p x1 x2 Returns a list res containing all values for x such that p x1 x2 is satisfied Example All solutions to x 4 swrlex allKnown res x swrlb pow 4 x 2 returns x 2 0 2 0 Note th
17. ers and strings Note that the exact representation may be changed e g between scientific and decimal notations Example swrlex string_number 5 0e 3 x returns x 0 005 Example swrlex string_number x 0 00000001 returns x 1 000000e 08 string resource str res Takes two argument a string and an RDF resource i e an OWL individual class etc Satisfied when the string is the full name of the resource including the full URI Like string number this predicate can be used in both directions Example swrlex string resource s owl Thing returns s http www w3 org 2002 07 owl Thing value s o Satisfied when rdf value s 7 0 is satisfied This predicate is needed because SWRL does not accept RDF predicates in queries This lets us call the rdf value predicate from a SWRL rule or query apply p x1 x2 Satisfied when the atom with the predicate p and arguments x1 x2 is satisfied This predicate lets us use predicates as arguments and have variables that bind to predicates This gives us a limited higher order logic which can be quite useful sometimes Note that queries using this predicate may take a very long time if p is not instantiated i e if you ask for all predicates that satisfy given arguments As an example consider a predicate minimum min x y p which returns the smallest value min out of two values x and y according to some ordering predicate p Think of th
18. h reference manual covering all the tool s features All example queries can be seen and run in SwrlIqg owl food food planning owl Example query that uses only information directly expressed in OWL vin Wine wine vin hasSugar wine sugar returns all the wines and their hasSugar properties Dry OffDry or Sweet as shown in Figure 1 Example query using a pseudo builtin defined by a SWRL rule hasPrice wine price withinRange price 10 50 returns all the wines that have an average price between 10 and 50 according to http www wine searcher com as in Figure 2 This query uses the pseudo builtin see Section 3 1 1 3 argument predicate withinRange defined by the SWRL rule swrlb greaterThanO0rEqual x y A swrlb lessThan0rEqual x z withinRange x y z Example trace You can see how the rule is used when you click the Trace All button as shown in Figure 3 There is more information about tracing in Section 3 2 SWRL IQ User Manual File Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help DCM 835 ma Ys 4p S prot g Metadata food planning owl OWLClasses WW Properties 4 Individuals Forms SWRLRules SWRL IQ SWRL Query INSTANCE BROWSER uec Cuin Vin hasSugar wine sugar For Class swrlu Query Asserted Instances x S ome Sa Bee 4 WineSugar gt B8 Submit query Trace First
19. hite vin Sauvignon vin Dry 16 http www wi http static3 wi al vin Forman vin Red vin CabernetS vin Dry 79 http www wi http static3 wi aC vin Marietta vin Red vin CabernetS vin Dry 20 http www wi http static3 wi 9 vin Stonleigh vin White vin Sauvignon vin Dry 4 http www wi http static3 wi RE Ww UJ UJ UJ UJ UJ SST Ll Ll d d d et steel det e b y Fig 6 Query result saved to CSV and graphed File Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help DOG x88 mi ov AAA lt b gt prot g IS MA s Forms SWRL Rules SWRL IQ SWRL Quei swriex allknown regions region vin locatedin vin BancroftChardonnay region For Class swrlu Query Asserted Instances x S o d Bee 4 allknown_locatedin E ETE SES y 8B Submit query l Trace First Trace All Debug t C Reload KB OWL SWRL Answer s regions vin CaliforniaRegion vin NapaRegion vin USRegion Show SPARQL XML Save as CSV Fig 7 All known regions query SWRL IQ User Manual File Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help Dee B wa 29 AO ao prot g 0 Metadata food_planning owl OWLClasses BW Properties 4 Individuals Forms SWRLRules SWRL IQ SWRL Que swrlex allKnown listOfWinePricePairs rdf List wine pr
20. ice hasPrice wine price swrlu sort sortedListOfWines listOfWinePricePairs winePairCheaperThan For Class swrlu Query x Asserted Instances 4 InPriceRange 4 InRegion 4 Java amp ttachment 4 sortWinesByPrice 4 WineSugar omo gn amp 4 allKnown locatedln HA C E ioc Submit query TraceFirst Trace All Debug Reload KB OWL SWRL Answer s sortedListOfWines vin StonleighSauvignonBlanc 14 0 vin SelaksSauvignonBlanc 16 0 vin MariettaCabernetSauvignon 20 0 vin MountEdenVineyardEdnaValleyChardonnay 21 0 vin VentanaCheninBl vin StonleighSauvignonBlanc 14 0 vin SelaksSauvignonBlanc 16 0 vin MariettaCabernetSauvignon 20 0 vin MountEdenVineyardEdnaValleyChardonnay 21 0 vin VentanaCheninBlanc 22 0 vin PageMillWineryCabernetSauvignon 24 0 vin ChateauMorgonBeaujolais 25 0 vin FormanChardonnay 39 0 vin FormanCabernetSauvignon 79 0 vin WhitehallLaneCabernetFranc 93 0 vin ChateauDYchemSauterne 634 0 vin ChateauMargaux 769 0 vin ChateauLafiteRothschildPauillac 1452 0 vin ChateauChevalBlancStEmilion 91743 0 Show SPARQL XML Save as CSV Fig 8 Sort wines by price query File Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help DG E
21. ild run makexsb64 bat You should now be able to run XSB config x64 pc windows bin xsb 2 5 3 Cygwin Compiling under Cygwin basically works the same as compiling under Linux Mac with one caveat The space in the C Program Files directory which usually is part of the JAVA HOME environment variable can cause some problems To solve this introduce extra escaped quotes before running configure e g export JAVA_HOME C Program Files Java jdk1 6 0_24 Note that Cygwin only supports 32 bits 2 6 Compiling CLPR Go to XSB packages Start XSB e g bin xsb or bin xsb Enter the command clpr Exit XSB by using the command halt Go to XSB packages clpar Start XSB again e g bin xsb or bin xsb Enter these commands clpr make make clpr 2 7 Setting the XSB environment variable SWRL IQ needs to know where to find XSB This is done by setting up an XSB environment variable pointing to the XSB executable However the variable must not point to the XSB bin xsb or XSB bin xsb bat file This file is a script that in turn calls the real executable which is located in XSB config ARCH bin xsb where ARCH is an architecture dependent directory name For example on a 64 bit Linux system the architecture directory is usually called x86 64 unknown linux gnu To set up the XSB environment variable on Windows 7 1 Click on the Windows button in the lower left corner 2 In the search field type e
22. in SelaksSauvigno vin Selaks lt feinding www W3 0rg TR 2003 PR owl guide 20031209 w http www wine se http static3 wine vin FormanCaberne vin Forman lt binding name maker gt http www wine se http static3 wine i gt http www w3 TR 2003 PR owl guide 20031209 lvin MariettaCabern vin Marietta binding WS org TR ER owi quide http www wine se http static3 wine vin StonleighSauvig vin Stonleigh lt bindin name color hd Sy SS mne Save Show SPARQL XML Save as CSV Fig 11 Saving XML T http www w3 org TR rdf sparql XMLres 17 SWRL IQ User Manual 4 1 1 Post Processing XML Query Results with XSLT The XML query results can be post processed with XSLT to produce output in other formats such as HTML To try this out yourself save the output of the WineProperties query from food_planning owl into your SurlIq html directory and name it query_result xml this file name is hard coded into the query result html file provided Then look at the query_result html page in an XSL capable web browser such as Firefox It should look as in Figure 5 in the Overview Section The query_result xsl stylesheet takes the query output and formats it as a table stripping prefixes like vin showing jpg URLs as inline images making other URLs clickable and formatting decimals as prices in dollars This is jus
23. ing a query currently requires double clicking on a query individual which brings up the standard Prot g individual editor For editing the actual query content use the query editor to the right of the query instance browser Example a simple query vin Wine wine A vin hasSugar wine sugar This query returns all the wines and their hasSugar properties Dry OffDry Sweet as shown in Figure 10 Eile Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help gt cr I lt Q prot g Metadata food planning owl E OWLClasses MA Properties Individuals Forms SWRL Rules SWRL IQ SWRL Quen INSTANCE BROWSER vin hasSugar wine vin Wine wine sugar swrlu Query 4 WineSugar e uery Instance Browser Submit query Trace First Trace All Debug Reload KB Query Editor OWL SWRL Answer s sugar vin MariettaCabernetSauvignon vin LongridgeMerlot vin FormanChardonnay vin CorbansDryWhiteRiesling vin ClosDeLaPoussieSancerre Query Results Panel vin Dry vin Dry vin Dry vin OffDry vin OffDry Show SPARQL XML Save as CSV Fig 10 Wine sugar query amp Query UI Note that you have to have a query individual to edit before you can change the content in the query editor Sometimes you just want to try different queries without saving them This is most easily done by just creating one scratch pad query individual for all such queries The content of the scratch
24. ions as shown in Figure 5 For more details see Section 4 1 Example of saving results to CSV and graphing Query results can be saved to a csv comma separated values file and graphed in your favorite spreadsheet software as shown in Figure 6 See Section 4 2 for more detail Example using reversible math Many math builtins generate numerical constraints that are combined and if possible simplified away later This makes them reversible and able to solve systems of equations For example instead of using swrlb pow x 3 2 the usual way to determine what the result of 3 is it can also be used to compute square roots by providing the output argument and leaving one of the input arguments as a variable The query swrlb pow 9 3 x returns x 2 0 See Section 5 1 for more detail Example using allKnown swrlex allKnown regions region vin locatedIn vin BancroftChardonnay _region produces a list of all regions satisfying the transitive vin locatedIn property as shown in Figure 7 The allKnown predicate is used to aggregate query results into a list and is described in more detail in Section 6 Also note the use of the anonymous variable region the values of which we are not interested in see Section 3 132 SWRL IQ User Manual File Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help Dek BA ma pipi ap prot g Metadata food_planning owl OWLClasses BW Pr
25. ishes it from other query and reasoning tools such as the Prot g SQWRLQueryTab supported by the Jess rule engine Goal oriented backward chaining Prolog style reasoning as opposed to the forward chaining paradigm used by Jess and the Prot g SWRL Bridge framework Constraint solving based on CLP R Constraint Logic Programming with Reals This allows for more declarative and powerful rules and queries Saving queries to XML or CSV format the SQWRLQueryTab saves to CSV only Tracing and debugging inference results Exporting query results in different formats No dependency on proprietary or closed source components Uses XSB Prolog which is freely available under the LGPL license SWRL IQ also has powerful SWRL extensions and a Java procedural attachment mechanism similar to what the Jess query tab supports 1 1 Project Home Page and Mailing List The SWRL IQ home page is http www onistt org display swrliq The user mailing list is swrl iq users lists esd sri com If you wish to subscribe send a message to swrl iq users requestQlists esd sri com with the word subscribe as subject You will receive an auto response that includes a link to confirm the subscription You will then receive an email welcoming you to the mailing list 1 2 A Quick Tour The rest of the Overview section provides a quick tour of the tool to illustrate some of its capabilities The following sections provide a more thoroug
26. nctional property symprop P P is a symmetric property transprop P P is a transitive property invfuncprop P P is an inverse functional property Class expressions class name the named class itself union Cl Cn Union class int C1 Cn Intersection class enum I1 In Enumerated class compl C Complement class all P C allValuesFrom on property P class C some P C someValuesFrom on property P class C has P I hasValuesFrom on property P individual I maxCard P N maxCardinality on property P value N minCard P N minCardinality on property P value N card P N cardinality on property P value N Table 2 Justifications in the trace and debug trees 15 SWRL IQ User Manual SWRL IQ is a sound but not complete reasoner with regard to the OWL language In other words all query answers are correct but there could be additional answers that the reasoner is not able to infer This is because the reasoner sacrifices support for some OWL constructs in order to maintain computational efficiency Indeed SWRL IQ could be considered a pure SWRL reasoner in the sense that all of the OWL axioms that are supported could also be written as SWRL rules For exampld the OWL axiom SubClassOf Wine restriction hasMaker allValuesFrom Winery can be written as a SWRL rule Wine x hasMaker x y Winery y More precisely the fragment of OWL that is supported is called DLP Description Logic Programs SWRL IQ does n
27. ns are different for different solutions The meaning of an atom node along with its justification and its child atom nodes is that the atom was proved using the inference rule given in the justification and if applicable the child atoms which in turn have their own justifications and so on The possible justifications inference rules are shown in Table 3 2 1l Note that the class expressions can occur wherever a C Cl or Cn argument is indicated This also means that the justifications can be nested to an arbitrary depth since class expressions can be nested in such a way 3 2 2 Debugging Debug trees show attempted proofs rather than actual proofs and add two types of nodes to the proof trees FAILED and ABORTED nodes FAILED nodes occur as a child of an atom that was attempted but could not be proved ABORT nodes occur subsequent to FAIL nodes because sub proofs where one part has failed cannot succeed so the sub proof attempt is aborted Note that debug trees can have a very large number of solutions because there can be a very large number of ways to attempt to prove something Often many proof attempts will seem meaningless to the human user and failures will occur in uninteresting places If the debug output is overwhelming we recommend breaking the query down into smaller parts to find the interesting failure 3 3 Semantics and Limitations The SWRL IQ reasoner differs significantly from ty
28. nvironment 3 Click on Edit the system environment variables SWRL IQ User Manual 4 Click the Environment Variables button 5 Click one the the New buttons depending on whether you want to make this a system wide setting or a user setting 6 Enter XSP as the variable name and the appropriate executable as discussed above as the variable value On Linux and Mac one simple way to set up the XSB variable is to edit the run protege sh script in the main Prot g folder Add a line export XSB PATH before the last line of the file where PATH is the appropriate path to the executable For example the whole line might be export XSB homes elenius XSB config x86 64 unknown linux gnu bin xsb 2 8 Importing SWRL Utilities SWRL IQ comes with a few owl files One of them is swrl utilities owl This file needs to be imported in order to use SWRL IQ If it is not the SWRL IQ tab will show a text reminding the user of this It contains the definition of the swrlu Query class which is used to save queries It also provides a small library of useful predicates see Section 7 and imports swrl extensions owl which provides a number of very useful new SWRL built ins see Section 6 2 9 Verifying the Installation SWRL IQ is a so called Tab Widget To use it the widget must first be enabled under the Project Configure menu item Checking the box for SwrlIQTab should make a new tab with that title
29. oding of mathematics and other functional expressions and this always involves the introduction of new variables The solution to this is a feature called Anonymous Variables By prefixing a variable name with _ underscore it will not be reported in the query result Thus Example x 5 64 8 swrlb multiply _z 5 6 A swrlb add x _z 3 returns only x 33 0 3 2 Tracing and Debugging The tracing and debugging features of SWRL IQ are used for explanation of reasoning results They answer the questions why or why not did a result happen The user interface for both is similar The debug view adds a few elements that the trace view does not have so we start by describing the trace view 3 2 1 Tracing Tracing is used to explain how and why a query succeeded It can be used for example to learn why an unexpected result happened or to check that some SWRL rules got used in the intended way There are two versions of the tracer Trace First and Trace All The former returns the first proof of a given query The latter returns all proofs of the query There are often multiple ways to reach the same result When many parts of a proof each have multiple proofs the combination of possible proofs for the overall query can become very large and cause the tracing to take a very long time In such cases the tracing can be aborted using the Stop button and re tried with the Trace First feature The trace window c
30. ontains a tree showing one or more proofs of the given query as was shown in Figure 8 Each node in the solutions tree has a context menu that can be accessed by right clicking the node The context menu allows you to expand all nodes below the current one copy the current node to the clipboard or copy the current node and all its children to the clipboard The copy to clipboard options can be useful for doing text searches on the proof trees when the proofs are very large or numerous In the future we intend to integrate search and filtering capabilities into the user interface in a more direct way Each Solution node shows the depth of its proof which can sometimes be interesting This corresponds to the depth of the tree starting at the solution node 13 SWRL IQ User Manual The Solution nodes have one or more child nodes which we call atom nodes One atom node for each atom in the query The atom nodes consist of an atom a justification and possibly a number of child atom nodes The atom nodes are color coded as shown in Table Element Color Class Yellow Object Property Blue Datatype Property Literal Green SWRL Builtin Variable Red Individual Purple Other RDF Resource Pink Table 1 Color coding in the trace and debug trees In tracing mode each atom will have all its variables instantiated In debug mode this is not always the case for failed sub proofs more on this below The variable instantiatio
31. operties 4 Individuals Forms SWRLRules SWRL IQ SWRL Query vin Wine wine vin hasMaker wine maker vin hasColor wine color vin madeFromGrape wine grape For Class swrlu Query vin hasSugar wine sugar hasPrice wine price priceReferenceURL wine url bottlelmageURL wine label Asserted Instances 94x96 4 Java amp ttachment a 4 SortWinesByPrice 4 WineProperties 4 WineSugar 4 allKnown locatedln HA ET le js Submit query Trace First Trace All Debug C Reload KB OWL SWRL Answer s wine maker color grape sugar price url label in ChateauLafiteRothschildPauillac vin ChateauLafiteRothschild vin Red vin CabernetSauvignonGrape vin Dry 1452 0 http www wine searcher http sa3 wine searcher vin SelaksSauvignonBlanc vin Selaks vin White vin SauvignonBlancGrape vin Dry 16 0 http www wine searcher http static3 wine search in FormanCabernetSauvignon vin Forman vin Red vin CabernetSauvignonGrape vin Dry 79 0 http www wine searcher http static3 wine search vin MariettaCabernetSauvignon vin Marietta vin Red vin CabernetSauvignonGrape vin Dry 20 0 http www wine searcher http static3 wine search vin StonleighSauvignonBlanc vin Stonleigh vin White vin SauvignonBlancGrape vin Dry 14 0 http www wine searcher http static3 wine search Show SPARQL XML Save as CSV Fig
32. or the variable bindings where sri is a prefix for the namespace http www onistt org sparql When the SPARQL XML result window is showing the Save button allows you to save the result to a file as shown in Figure 11 File Edit Project OWL Reasoning Code Tools BioPortal Window Collaboration Help DG E 86 lt PoP ar prot g Metadata food_planning owl OWLClasses HA Properties Individuals Z Forms SWRL Rules SWRL IQ SWRL Query vin WineCwine vin hasMaker wine maker vin hasColor wine color vin madeFromGrape wi For Class swrlu Query vin hasSugar wine sugar hasPrice wine price priceReferenceURL wine url bottlelmageUl SPARQL XML Results Asserted Instances lt xml versionz 1 0 4 SortWinesByPrice lt sparql xmlns http www w3 org 2005 sparql results 4 WineProperties Ame sri http www onistt org sparql gt 4 WineSugar lt variable names wine lt variable namez maker lt variable name color gt lt variable names grape lt variable names sugar lt variable name price gt lt variable name url gt lt variable name label gt OWL SWRL Answer s lt head gt wine maker results url label i 7 i 3 result lvin ChateauLafiteR vin ChateauLafiteR binding name wine gt http www wine se http sa3 wine se lv
33. ot support OWL class descriptions inside SWRL rules and Prot g does not even let you enter those so most people will not miss this However defined classes can be used to achieve the same effect as long as the class definition is within the DLP fragment Other than this limitation arbitrary SWRL rules are supported 5 This example is taken from the OWL Web Ontology Language Guide ihttp www w3 org TR 2004 REC owl guide 20040210 http www2003 org cdrom papers refereed p117 p117 grosof html 16 SWRL IQ User Manual 4 EXPORTING AND PROCESSING QUERY RESULTS SWRL IQ can export query results in two different formats These results can be further processed or consumed by other tools We describe the two export formats along with some example post processing below 4 1 SPARQL Query XML Results After a query has been submitted and results appear in SWRL IQ clicking the Show SPARQL XML pops up a new window that shows the query result in the SPARQL XML Result forma It may seem odd that we export using a SPARQL format as we do not support SPARQL as a query language However the motivation is that a exporting to some XML format can come in handy since it allows us to process the result using XML tools and b the SPARQL query result format already exists and there is no reason to define a new format We make one addition to the SPARQL XML format In order to allow lists in query results we add a new sri list element f
34. pical OWL reasoners such as Pellet and Fact Most OWL reasoners are Description Logic reasoners which are intended primarily for so called T box or terminological reasoning For example they can be used to infer that class A is a subclass of class B These reasoners have no or limited support for SWRL rules SWRL IQ on the other hand focuses on so called A box or assertional reasoning i e reasoning about facts in the knowledge base while still taking into account terminological axioms in the ontology SWRL IQ supports complex built ins see Section 5 advanced SWRL extensions see Section 6 and user defined procedural attachments defined in Java see Section 8 Some of these e g disj and differentFrom will currently not appear due to limitations of the reasoner See Section 14 SWRL IQ User Manual Justification entry Meaning Axioms assert comp rule R subcls C1 C2 eqvcls C2 C2 disj C2 C2 subprop P1 P2 eqvprop P1 P2 dom P C rng P C invprop P1 P2 sameAs 11 12 differentFrom 11 12 Asserted fact Computed as a built in SWRL rule with name R Cl is a sub class of C2 C1 and C2 are equivalent classes Cl and C2 are disjoint classes Pl is a sub property of P2 P1 and P2 are equivalent properties Property P has domain C Property P has range C P1 and P2 are inverse properties I1 is the same individual as 12 I1 is a different individual than I2 funcprop P P is a fu
35. rEqual 39 0 10 comp swrlb lessThanOrEqual 39 0 50 comp Y z9 Solution 2 depth 2 hasPrice vin MariettaCabernetSauvignon 20 0 assert Y withinRange 20 0 10 50 rule Within Range Rule swrlb greaterThanOrEqual 20 0 10 comp swrlb lessThanOrEqual 20 0 50 comp Y Solution 3 depth 2 hasPrice vin VentanaCheninBlanc 22 0 assert Y withinRange 22 0 10 50 rule Within Range Rule swrlb greaterThanOrEqual 22 0 10 comp swrlb lessThanOrEqual 22 0 50 comp Y z9 Solution 4 depth 2 hasPrice vin MountEdenVineyardEdnaValleyChardonnay 20 0 assert Y withinRange 20 0 10 50 rule Within Range Rule swrlb greaterThanOrEqual 20 0 10 comp swrib lessThanOrEqual 20 0 50 comp Y Solution 5 depth 2 hasPrice vin PageMillWineryCabernetSauvignon 24 0 assert Y withinRange 24 0 10 50 rule Within Range Rule swrlb greaterThanOrEqual 24 0 10 comp swrlb lessThanOrEqual 24 0 50 comp Fig 3 Price range query trace Example of saving results to XML and transforming to HTML vin Wine wine A vin hasMaker wine maker A vin hasColor wine color A vin madeFromGrape wine grape A vin hasSugar wine sugar A hasPrice wine price A priceReferenceURL wine url A bottleImageURL wine label returns various properties of the wines as show in in figure 4 The query results can be saved to an XML file and transformed to HTML or other formats using XSL transformat
36. rong number of arguments Thus you must make sure to consistently use the same number of arguments Of course your new predicate isn t really built in i e it is not implemented by the reasoning engine You have to create the rules that define the semantics of the new predicate Hence we can refer to these predicates as pseudo built ins Being able to use more than two arguments is extremely useful When you are limited to two arguments one of which is often a return value you essentially can only create functions of one argument The other solution is to use a list as an argument and then pack the real arguments into this list However this becomes very cumbersome as most rules need lots of additional content to construct and deconstruct these lists One word of caution Using pseudo builtins is a non standard use of SWRL Ontologies containing them may not load or work properly in other ontology tools 12 SWRL IQ User Manual 3 1 2 Anonymous Variables In many queries there is a need to use intermediate variables whose bindings are not interesting to the user Normally the bindings for all variables are reported in the query result Example x 5 64 8 swrlb multiply z 5 6 A swrlb add x z 3 returns x 33 0 Tz 30 0 However we re not interested in z This is also indicated by the fact that it does not even appear in the mathematical notation for the query SWRL provides a relational enc
37. ry result One consequence of this is that these built ins can be used backwards and in effect to solve systems of equations This often makes rules that use these built ins more declarative and powerful The constraint mechanism is best understood through examples see Table 4 Note that the built ins integerDivide mod ceiling floor and round cannot be used backwards for obvious reasons The results are not uniquely defined in the reverse direction These predicates do not generate constraints If the input arguments to these predicates are not instantiated the query will simply fail 5 2 Lists A few things are worth pointing out when it comes to our implementation of list operations rdf Lists can be created on the fly in SWRL queries using the syntax rdf List m1 m2 where mi m2 are the list members The list built ins can be used to generate lists not just to check properties or retrieve values of existing lists The reasoner can operate on and return partially instantiated lists e g a list of three elements where the second element is unknown Lists in the variable bindings returned by queries are presented in a different syntax using square brackets for brevity as in Prolog i e m1 m2 If the tail of the list is unspecified i e the list has unknown size the Prolog bar syntax is used e g m1 m2 tail These features are illustrated in Table 5 5 3 Strings String liter
38. s is described in detail in Section 8 25 SWRL IQ User Manual 7 SWRL UTILITIES Besides containing the ontological elements for saving queries as discussed in Section 3 swrl utilities owl also contains a number of predicates defined in SWRL rules that are generally useful The rules for these predicates can also serve as nice small examples of how to write SWRL rules Remember to use the prefix for the SWRL Utilities ontology when using these predicates in queries outside of that file e g swrlu nth rather than just nth disjoint With 11 12 Satisfied if the two lists 11 and 11 do not have any common elements intersectsWith 11 12 Satisfied iff the two lists have at least one element in common subset Of 711 712 Satisfied iff the list 711 is a subset of the list 712 nth n 1 res Satisfied iff res is the nth element of the list 71 starting at n 0 for the first element first x 1 second x 1 third x 1 fourth x 1 Satisfied iff x is the first second third fourth element in the list 1 Use nth for higher indices or define additional short cut predicates like these logn x y n Satisfied iff x is the base n logarithm of y log10 x y Satisfied iff x is the base 10 logarithm of y i e just a short cut for logn dB x y Satisfied iff x is the decibel value of y notMember el 1 Satisfied iff el is not a member of the list 1 minimum min x y p Satisfied iff
39. t a simple and not very robust example that can be modified for your own purposes 4 2 CSV Query Results The Save as CSV button allows you to save query results to a comma separated values CSV file The file will have the variable names in the first row and the bindings in subsequent rows similarly to the query results table in the main SWRL IQ tab The main usefulness of CSV files is that they can be opened directly in Excel and other spreadsheet applications and are automatically rendered into cells appropriately This can be used to generate graphs export to other formats etc An example of this was shown in Figure 6 in the Overview Section XSL Transformations http www w3 org TR xslt Alternatively you could create an html page that reads in path names dynamically and or calls an XSLT Processor like SAXON instead of relying on the browser 18 SWRL IQ User Manual 5 SWRL BUILT INS SWRL IQ supports most of the standard SWRL Built ind Table 3 summarizes our level of support for these built ins with more detailed notes and examples in the following sections We do not repeat the standard definitions of all the predicates here 5 1 Constraints Many of the built ins can generate numerical constraints and are implemented to do so in a very flexible way Constraints are maintained throughout the query and simplified as much as possible If any constraints cannot be simplified away they are returned in the que
40. that Prot g does not have a way to store timezone information the last argument Also note that there is no way to directly write a xsd date dateTime or xsd time value in a SWRL query or rule E g the query swrlb dateTime 2011 11 11T11 11 11 y m d h m s tz is not syntactically correct you will get an invalid literal error in the SWRL query editor If you put quotes around the first argument the query is syntactically correct but the first argument is treated as a plain string and therefore the query fails However a date time term can be created dynamically using these predicates as shown above If a value of one of these types exists in the ontology e g as the value of a datatype property assertion it can be retrieved by an appropriate query and the different parts of the date time value can then be extracted using these predicates The time arithmetic built ins are not implemented mainly because they operate on non standard XSD data types that cannot actually be used in Prot g 5 5 Prot g Extensions Prot g adds a number of SWRL built indH SWRL IQ does not support most of these The ones we do support are described below Note that you need to import the relevant ontology i e rdfb owl or swrlm owl respectively to use these built ins rdfb hasLabel s 1 Satisfied iff rdfs label s 7 1 is satisfied This predicate is needed because SWRL does not accept RDF predicates in queries This lets
41. tor for that resource double clicking on numbers and strings does nothing Clicking on a column header sorts the table alphabetically according to the entries in that column Sometimes a query or trace debug attempt takes a very long time In fact with recursive SWRL rules it is possible to create situations where the reasoning engine can run forever In such cases the Stop button can be used to abort what the reasoner is doing and make it available for new queries The Show SPARQL XML and Save as CSV buttons are discussed in Section 4 3 1 1 Pseudo Built Ins There is a hidden and possibly unintentional feature of Prot g that makes SWRL rules much more powerful Regular SWRL supports only unary and binary predicates except for the SWRL built ins see Section 5 which can take an arbitrary number of arguments This is because OWL classes and properties are used for user defined predicates However in Prot g there is a way to create new predicates with an arbitrary number of arguments Here s how to do it 1 In Project gt Configure gt Options check Display Hidden Frames 2 In the Individuals tab select the swrlb Builtin class 3 Create a new individual of this class and give it the name you want for your new predicate You can now use your new predicate in SWRL rules and queries Note that the arity is not defined and the SWRL rule and query editors will not display an error when you use the w
42. ts and is always satisfied This is useful when creating queries because SWRL requires all variables that are in the rule head to also be in the rule body However there is often a need to have extra variables in the rule head e g when there are several rules that define a predicate and some of the arguments are not used in some of the rules Inserting an ignore atom into the rule body accomplishes this without unwanted side effects Example A predicate that adds a number to all numbers in a list swrlb empty 12 A swrlb empty 11 A swrlex ignore n gt addN 11 7 12 7 n swrlb first f2 12 A swrlb add f1 f2 n A swrlb first f1 11 A swrlb rest r2 12 A swrlb rest ri 11 A addN r1 r2 n gt addN 11 7 12 n In the first rule ignore is needed since the variable n would otherwise appear in the rule head but not in the rule body The ignore predicate is not useful in queries pi x Takes one argument which is unified with the constant 7 as a decimal up to the precision that the underlying Prolog engine supports Example swrlex pi pi A swrlb divide _cf pi 180 A swrlb multiply radians 90 7 cf returns pi 3 141592653589793 radians 1 5707963267948966 23 SWRL IQ User Manual string number str num Takes two arguments a string and a number Satisfied when the string is a recognized string representation of the number Can be used in both directions to convert between numb
43. us call the rdfs label predicate from a SWRL rule or query SWRL IQ User Manual swrlm log s 1 Satisfied iff the first argument is equal to the natural logarithm base e of the second argument swrlm sqrt s 1 Satisfied iff the first argument is equal to the square root of the second argument This can also be achieved with the swrlb pow built in 22 SWRL IQ User Manual 6 SWRL EXTENSIONS SWRL IQ comes with an swrl extensions owl ontology which contains a number of new SWRL built ins These built ins all have dedicated implementations in the Prolog reasoner core Remember to use the prefix for the SWRL Extensions ontology when using these predicates in queries outside of that file e g swrlex pi rather than just pi Keep in mind that these extensions are non standard and will not work in other reasoning engines besides SWRL IQ unless others decide to implements the same extensions The examples below are all kept very simple and can be executed with no ontology loaded other than swrl extensions owl The first several predicates ignore pi string_number string_resource and value are relatively simple utility predicates The last few predicates apply allKnown firstMatch and callJavaStaticMethod offer more radical extensions to SWRL They go beyond the semantics of SWRL and should be used sparingly if a pristine model theory is considered critical ignore x1 x2 Takes an arbitrary number of argumen
44. you need to get the XSB source code either the latest stable release at http xsb sourceforge net downloads downloads html or the latest version from CVS To get it from CVS enter the commands cvs d pserver anonymousOxsb cvs sourceforge net cvsroot xsb login Hit enter when asked for password Then cvs z3 d pserver anonymousOxsb cvs sourceforge net cvsroot xsb co P XSB This should give you a XSB directory 2 4 Compiling XSB on Mac Linux In a shell window go into the XSB build directory and enter configure makexsb You should now be able to run XSB bin xsb If there are errors first make sure you have the latest versions of the build tools e g Xcode on Mac 2 5 Compiling XSB on Windows First you need to install Microsoft Visual C Express which is freely available at http www microsoft com express vc Then install the Microsoft Windows SDK from http msdn microsoft com en us windows bb980924 aspx Next follow the instructions below depending on the type of machine that you are compiling for 2 5 1 32 bit Windows In a command window run C Program Files x86 Microsoft Visual Studio 10 0 VC BIN vcvars32 bat Then in XSB build run nttp xsb sourceforge net SWRL IQ User Manual makexsb bat You should now be able to run XSB config x86 pc windows bin xsb 2 5 2 64 bit Windows In command window run C Program Files Microsoft SDKs Windows v7 1 bin SetEnv cmd Then in XSB bu
Download Pdf Manuals
Related Search