Введення в спеціальність / Додаткова література / data mining with ontologies

preliminary validatiOn stUdy

In this section, we use the example from seismology domain and the ontology defined in section domain knowledge with ontologies to describe system functionality. The system performs a validation test before the data mining process, checking if the user defined parameters make sense. For example, a user could ask the system toperformtheApriorialgorithmtofindassociationsbetweenthe“magnitude”andthe“date”of an earthquake. As mentioned in Section 3.2, this association is not acceptable by the seismology

domain and thus the system will suggest the user to change the parameters. If the user does not specify the attributes that he/she wants to search for associations, the system will perform the data mining algorithm using all attributes but, when generating the frequent itemsets, it will discard itemsets that contain values from attributes not related in the ontology. In this way, the time consuming phase of frequent itemset generation will be improved and no irrelative association rules will be generated. Of course, this is not always desirable as some interesting rules might not be generated. In this case the user should decide for

these rules. So, it is given as option to the user eithertoenablethesystemtoautomaticallydiscard them or just to mark the “noisy”ones for further evaluation. In the latter case, the user decides which rules are interesting and should be stored to the pattern base.

Another case is when a user is posing a query to the pattern base to retrieve patterns for example, “fetch association rule patterns that contain both

‘seasonand‘depth’attributesandthesupportofthe rule is greater than 0.3.” Such rules are not valid according to the domain knowledge and thus, the systemnotifiestheuserthatitisratherimpossible to find rules like those in the pattern base.

In our first experiments, we ran the Apriori data mining algorithm implemented in WEKA

(Witten&Frank,2005)toextractsomeassociation rulesusingrealmacroseismicdatacollectedbythe Greek Institute of Geodynamics (Seismo-Surfer). Attributes such as earthquake depth, intensity, site, date, and season of the year are some of the attributes of the table that contains 10336 tuples for the earthquake events during the 20th century. Table 1 lists 25 out of 70 rules extracted by Apriori confidencethreshold=30%andsupportthreshold

= 10% are listed in Table 1.

Out of these 25 rules, the domain expert marked only five rules (ids 1, 8, 14, 15, 17) as interesting

and all others as “noisy” because they describe a correlation between attributes/classes that is meaningless in the domain of seismology. The systemneedsathresholdparametertobedefined inordertomarksomerulesas“noisy.”Thisthreshold is the maximum path distance from the main

“earthtremor”node/class.Whenthisthresholdis defined, the system retrieves the subgraph of the ontologydefinedbythe“earthtremor”nodeand all the nodes with path distance less or equal to the threshold. Every rule that has attributes belonging to that subgraph, will be considered interesting while all others will be marked as “noisy.”

Trying to detect a reasonable threshold in order for the system to retrieve the rules that will match the expert’s evaluation, we varied threshold value from 1 to 5 and computed the rules marked as “noisy” by the system. This is illustrated in Figure 9.

According to this experiment we conclude that with threshold 3, the system matches expert choices. As such, this threshold can be used by the user for the next running of Apriori or can even bestoredasmetadataforthespecificdatasetand

KDD process for future data mining.

With this procedure, we can measure the percentage of the rules that will be marked as

“noisy” by the system and by the expert, but we

do not know if these are the same rules that is, if the rules marked by the system are the same with the rules marked by the domain expert (precision). While in our particular experiment we had a perfect match, it is not sure that we will have a perfect match each time.

COnClUsiOn and fUtUre wOrk

In this chapter, we proposed a framework consisting of a PBMS that uses ontologies to improve data mining tasks, to evaluate extracted patterns and to improve querying over pattern base. The PBMS interacts with the data mining engine to discoverpatternsandstorestheminXMLformat according to the presented pattern model in the pattern base. Users can pose queries over the pattern base and the data sources. Ontology is used to evaluate the patterns extracted from the data mining process, and to validate the queries the user is posing.

The main idea is that the system is independent from the data mining engine and the ontology. So, according to the application domain, the pattern types and the ontology have to be defined. The PBMS has also the proper design for defining complexpatternsandforcomparingpatternsusing similarity measures based on the structure and the measure component of the patterns.

The proposed framework provides domain experts with a powerful tool that can help them to better manage and evaluate the patterns extracted from data mining algorithms. We aim in improving and enhancing the KDD process. Domain knowledge and knowledge representation techniques can help both in reducing the required time to run a data mining algorithm and in evaluating the extracted patterns. It is clear though, that due to the complexity of ontologies and the lack of standards on ontology creation, this incorporation is not an easy task. We have listed the theoretical and technical problems that should be faced.

Both PBMS and ontologies are areas of recent research and their applications could be many.

Apart from geosciences, every field that has a well defined ontology can use the integrated framework to improve the KDD process. For example in the domain of B2B marketplaces, finding associations between products is more efficientwhenusingthehierarchiesdefinedinthe product ontology. Although there is not currently universally accepted product ontology, efforts are made to integrate different product ontologies

(Omelayenko, 2000) towards this end.

In order to be able to use ontologies in KDD process and to have the results available to domain experts,ontologieshavetobedefinedinacommon way. There are a lot of efforts for ontology matching (Doan et al., 2003) and ontology integration

(Cui et al., 2002; Pinto & Martins, 2001), and this illustrates the need for an ontology creation standard. In this way, exchange and comparison of ontologies describing different domains could be possible.Untilnow,onlyseveraldomainspecific ontologies and tools have been developed.

Integrating ontologies to the data mining process is not an easy task and a lot of issues have to be addressed. Things are complicated due to the fact that scientists and companies create ontologies according to their needs instead of adopting a universal ontology. There is a large number of ontology languages most of them designed for the semantic web like RDF (Beckett, 2004), SHOE (Luke & Heflin, 2000), DAML, DAML+OIL (Harmelen, Patel-Schneider, & Horrocks, 2001), OWL (McGuinness, & Harmelen, 2005). New ontologies are constructed for various fields and applications without centralized guidance and common agreement. This is getting even more complex as recent studies have indicated semantic and syntactic conflicts between these languages, especially between DAML+OIL and OWL(Horrocks&Patel-Schneider,2003;Patel- Schneiderand&Fensel,2002).Therefore,building a system that uses ontologies in the data mining

process requires choosing a specific ontology language to support.

Another important theoretical issue concerns the evaluation of various pattern types using ontologies.Itisveryhardtodefinegeneralrulesthat applytoallpatterntypes.Themostpopularpattern typesfromdataminingfieldareassociationrules, clusters, decision trees, neural networks, and time series.Wehavedefinedfiltersforassociationrule miningbutdependingontheapplicationfiltersfor each pattern type separately have to be defined in order to build a system to support the majority of pattern types. Furthermore, we investigate the precision issue regarding the patterns that system marks as “noisy.”

Our framework is currently under development. Extended experiments and expert evaluation of the results on real case studies with association rule and decision tree mining are to be conducted. Early results have been positively evaluated by seismologists. Future work includes definingfiltersfordecisiontreesandclusterspatterns for seismological data as well as applying the framework to other domains.

aCknOwledgment

Research supported by the General Secretariat for Research and Technology of the Greek Ministry of Development under a PENED’2003 grant.

fUtUre researCh direCtiOns

Ontology-assisted pattern management is an emerging research field that finds a wide area of applications. Market basket analysis is one of the most popular data mining applications. In this application, data mining algorithms are used to discover associations between products in order to (a) assist decisions on the management and

promotionofproductsofferedand(b)makerecommendations to customers with respect to products others than those already bought. Clearly, when dealingwithalargemarket,findingassociations between products may be problematic if different product categories are involved (say, electronics and food). Using product hierarchies could result in more focused associations but an ontology that would filter the data mining results would give more accurate and sophisticated results / suggestions.

Constrained clustering is another application ofontology-filtereddatamining.Constraintsare commonly defined as must-link and cannot-link relations between data in order to optimize the creation of consistent clusters. These constraints could be implemented through a domain ontology to enable the must-link and cannot-link relations. In general, data mining algorithms have to be modifiedinordertoinvolveanontologyfiltering mechanism to evaluate the generated patterns of various types.

Last but not least, the flow of data coming from low-cost modern sensing technologies and wireless telecommunication devices (mobile and ubiquitous communications) enables novel research fields related with the management of timeand space-related data. Traditional knowledgediscoverytechniquesareextendedinorderto implementappropriateanalyticssoastotransform raw data into useful knowledge. Consequently, patternmanagementtechnologyshouldbeadapted to satisfy the new requirement: management of patterns produced by spatio-temporal algorithms. Due to the new kind of data, the role of ontologies in the evaluation phase of KDD should be revisited.

Considering the applications described above, the integration of pattern management and ontologiesisapromisingfieldofapplications,raisinga lot of research issues that have to be addressed.

referenCes

Beckett, D. (2004). RDF/XML Syntax Specification (Revised), W3C Recommendation. http:// www.w3.org/TR/rdf-syntax-grammar/

Catania, B., & Maddalena, A. (2006). Pattern

Management: Practice and challenges. In J.

Darmont & O. Boussaid (Eds.), Processing and managing complex data for decision support. Idea Group Publishing.

CataniaB.,Maddalena,A.,&Mazza,M.(2005). PSYCHO: A prototype system for pattern management. In Proceedings of the International Conference on Very Large Data Bases 2005.

ChenX.,Zhou,X.,Scherl,R.,&Geller,J.(2003).

Using an interest ontology for improved support in rule mining. In DaWaK 2003 (pp. 320-329).

CINQ (Consortium on Discovering Knowledge with Inductive Queries). (2001). http://www. cinq-project.org

Cui, Z., Jones, D., & O’Brien, P. (2002). Semantic B2B Integration: Issues in Ontology-based

Approaches. ACM SIGMOD Record archive SPECIAL ISSUE: Data management issues in electronic commerce table of contents, 31(1), 43-48.

CWM (Common Warehouse Model) (2001). homepage. http://www.omg.org/cwm

Doan,A.,Madhavan,J.,Domingos,P.,&Halevy, A.(2003).Ontologymatching:Amachinelearning Approach. In S. Staab & R. Studer (Eds),

Handbookonontologiesininformationsystems.

Springer-Velag.

Fayyad,U.M.,Piatetsky-Shapiro,G.,&Smyth,P.

(1996).Fromdataminingtoknowledgediscovery, an overview. In U. Fayyad, G. Piatetsky-Shapiro,

P. Smyth, & R. Uthurusamy (Eds.), Advances in knowledgediscoveryanddatamining(pp. 1-30). Menlo Park, CA: AAAI/MIT Press.

Fayyad, U., Haussler, D., & Stolorz, P. (1996). Mining scientific data. Communications of the ACM, 39(11), 51-57.

Freitas, A.A. (1999). On rule interestingness measures. Knowledge-Based Systems, 12(5-6), 309-315.

Gruber, T.R. (1993). A translation approach to portable ontologies. Knowledge Acquisition,

5(2), 199-220.

Guarino, N., & Giaretta, P. (1995). Ontologies and knowledge bases: Towards a terminological clarification. In N. Mars (Ed.), Towards very largeknowledgebases:Knowledgebuildingand knowledgesharing(pp.25-32).Amsterdam:IOS

Press.

Harmelen,F.V.,Patel-Schneider,P.F.,&Horrocks,

I. (2001). Reference Description of the DAML+ OILOntologyMarkupLangauge. Retrieved from http://www.daml.org/2001/03/daml+oil-index. html

Horrocks,I.,&Patel-Schneider,P.F.(2003).Three theses of representation in the Semantic Qeb. In

ProceedingsoftheTwelfthInternationalConference on World Wide Web.

Hotho, A., Maedche, A., Staab, S., & Zacharias, V. (2002, April 26-27). On knowledgeable unsupervised text mining. In ProceedingsoftheDaimlerChrysler Workshop on Text Mining, Ulm.

ISO SQL/MM Part 6 (2001). http://www.sql99.org/SC32/WG4/Progression_Documents/ FCD/fcd-datamining-2001-05.pdf

Kotsifakos, E., Ntoutsi, I., & Theodoridis, Y.

(2005).Databasesupportfordataminingpatterns. In Proceedings of PCI’05 (pp. 14-24). Springer

Verlag.

Liu,B.,Hsu,W.,Chen,S.,&Ma,Y.(2000).Analyzingthesubjectiveinterestingnessofassociation rules. IEEE Intelligent Systems, 15(5), 47-55.

Luke,S.,&Heflin,J.(2000).SHOE1.01Proposed Specification, SHOE Project. Retrieved from http://www.cs.umd.edu/projects/plus/SHOE/spec. htm

Maedche, A., Motik, B., Stojanovic, L., Studer,

R., & Volz, R. (2003). Ontologies for enterprise knowledge management. IEEE Intelligent Systems, 18(2), 26-33.

McGuinness, D.L., & Harmelen, F.V. (2005).

OWLWebontologylanguageoverview.Retrieved Feburary 2005, from http://www.w3.org/TR/owlfeatures/

Niles, I., & Pease, A. (2001). Toward a Standard Upper Ontology. In Proceedings of the 2nd International Conference on Formal Ontology in Information Systems (FOIS2001).

Niles, I., & Terry, A. (2004). The MILO: A general-purpose, mid-level ontology. In 2004 International Conference on Information and Knowledge Engineering (IKE’04).

Omelayenko, B. (2000). Integration of product ontologies for B2B marketplaces: A preview.

ACM SIGecom Exchanges, 2(1), 19-25.

Padmanabhan, B., & Tuzhilin, A. (1998). A be- lief-driven method for discovering unexpected patterns. In Proceedings of the International Conference on Knowledge Discovery and Data Mining (pp. 94-100).

PANDA (Patterns for Next-generation Database Systems). (2001). Project homepage. http://dke. cti.gr/panda

Patel-Schneiderand, P.F., & Fensel, D. (2002).

Layering the Semantic Web: Problems and directions. In Proceedings of the 1st International Semantic Web Conference. Springer.

Piatetsky-Shapiro, G. (1991). Discovery, analysis, and presentation of strong rules. In G. Piatetsky-

Shapiro & W.J. Frawley (Eds.), Knowledge dis-

covery in databases (pp. 229-248). Cambridge, MA: AAAI/MIT Press.

Piatetsky-Shapiro, G. (2000). Knowledge discovery in databases: 10 years after. SIGKDD Explorations, 1(2), 59-61.

Piatetsky-Shapiro, G. & Matheus, C.J. (1994).

The interestingness of deviations. In Proceedings of KDD-94: AAAI-94 Knowledge Discovery in Databases Workshop (pp. 25-36). AAAI Press.

Pinto,H.S.,&Martins,J.P.(2001)Amethodology for ontology integration. In Proceedingsofthe1st

International conference on Knowledge Captur.

(pp. 131-138).

Pohle, C., & Spiliopoulou, M. (2002). Building and exploiting ad hoc concept hierarchies for web loganalysis.InY.Kambayashi,W.Winiwarter,&

M. Arikawa (Eds.), Proceedings of the 4th International Conference on Data Warehousing and KnowledgeDiscovery,DaWaK2002,Vol.2454of Lecture Notes in Computer Science (pp. 83-93).

Aix en Provence, France: Springer-Verlag.

Pohle, C. (2003) Integrating and updating domain knowledge with data mining. VLDB PhD

Workshop.

Rizzi,S.,Bertino,E.,Catania,B.,Golfarelli,M.,

Halkidi, M., Terrovitis, M., et al. (2003). Towards a logical model for patterns. In Proceedings of ER’03 Conference.

Seismo-Surfer. A WebGIS application for integrating, visualizing and analyzing seismic data. http://www.seismo.gr

Silberschatz, A. & Tuzhilin, A. (1996). What makes patterns interesting. In knowledge discovery systems. IEEE Transactions on Knowledge and Data Engineering, 8(6), 970-974.

Theodoridis, Y., Marketos, G., & Kalogeras,

I.S. (2004). Collecting and mining seismic data in Greek territory: The seismo-surfer tool. In

Proceedingsofthe7th PanhellenicGeographical ConferenceoftheHellenicGeographicalAssociation (HGA’04), Mytilene, Lesvos, Greece.

Witten, I.H., & Frank, E. (2005). Data Mining: Practicalmachinelearningtoolsandtechniques, (2nd ed.). San Francisco: Morgan Kaufmann.

additiOnal reading

Batagelj, V., & Ferligoj, A. (1998). Constrained clustering problems. IFCS’98, Rome.

Bradley,P.S.,Bennett,K.P.,&Demiriz,A.(2000).

Constrained k-means clustering (Technical Report MSR-TR-2000-65). Microsoft Research, Redmond.

Charest,M.,Delisle,S.,Cervantes,O.,&Shen,Y. (2006,Oct).InvitedPaper: Intelligentdatamining assistance via cbr and ontologies. InProceedings ofthe17th InternationalConferenceonDatabase and Expert Systems Applications (pp. 593-597).

Chen, A., & McLeod, D. (2005, May 25-28). Se- mantics-based similarity decisions for ontologies. InProceedingsofthe7thInternationalConference on Enterprise Information Systems, Miami.

Choi, N., Song, I.Y., & Han, H. (2006). A survey on ontology mapping. ACM SIGMOD Record archive 35(3), 34-41.

Cody, W.F., Kreulen, J.T., Krishna, V., & Spangler, W.S. (2002). The integration of business intelligence and knowledge management. IBM Systems Journal.

Fensel, D. (2001). Ontologies: Silver bullet for knowledgemanagementandelectroniccommerce

(1st ed.). Springer.

Fonseca, F., Davis, C., & Camara, G. (2003).

Bridging ontologies and conceptual schemas in geographic information integration. Geoinformatica, 7(4), 355-378.

Fu, Y., & Han, J. (1995, Dec). Metarule-guided mining of association rules in relational databases. In Proceedings of KDOOD, Singapore (pp. 39-46).

Ganti V., Gehrke, J., Ramakrishnan, R., & Loh,

W. (1999). A framework for measuring changes in data characteristics. In ProceedingsofPODS’99, Philadelphia.

Gordon,A.D.(1973).Classificationinthepresence of constraints. Biometrics, 29, 821-827.

Gordon, A.D. (1996). A survey of constrained classification. Computational Statistics & Data Analysis, 21(1), 17-29.

Han, J. (1995). Mining knowledge at multiple concept levels. In Proceedings of CIKM, Maryland (pp. 19-24).

Han,J.,&Fu,Y.(1996).Explorationofthepower of attribute-oriented induction in data mining. In U.M. Fayyad, G. Piatetsky-Shapiro, P. Smyth,

& R. Uthurusamy (Eds.), Advances in knowledge discovery and data mining (pp. 399-421). AAAI/MIT Press.

Jurisica, I., Mylopoulos, J., & Yu, E. (2004). Ontologies for knowledge management: An information systems perspective. Knowledge and Information Systems, 6(4), 380-401.

Li,T.,Zhu,S.,&Ogihara,M.(2003).Anewdistributed data mining model based on similarity. In Proceedings of ACM-SAC’03.

Parekh,V.,Gwo,J.P.,&Finin,T.(2004).Ontology based semantic metadata for geoscience data. In

Proceedings of the International Conference of Information and Knowledge Engineering.

Parthasarathy, S., & Ogihara, M. (2000). Clustering distributed homogeneous datasets. In

Proceedings of PKDD’00, Lyon, France.

Philips, J., & Buchanan, B.G. (2001). Ontologyguided knowledge discovery in databases. In

Proceedings of the 1st International Conference on Knowledge Capture (pp. 123-130). Victoria,

British Columbia, Canada.

Wagstaff K., Cardie C., Rogers S., & Schroedl

S. (2001). Constrained K-means clustering with background knowledge. In Proceedings of the EighteenthInternationalConferenceonMachine

Learning (pp. 577-584). Williams College, MA: Morgan Kaufmann.

Wang, X., Chan, C.W., & Hamilton, H.J. (2002).

Design of knowledge-based systems with the on- tology-domain-system approach. In Proceedings of the 14th International Conference on Software Engineering and Knowledge Engineering (pp. 233-236).

Achard, F., & Barillot, E. (1997). Ubiquitous distributed objects with CORBA. Pacific Symposium Biocomputing. London, World Scientific.

Achard, F., & Dessen, P. (1998). GenXref VI: Automatic generation of links between two heterogeneous databases.

Bioinformatics, 14, 20-24.

Agirre,E.,Ansa,O.,Hovy,E.,&Martinez,D.(2000).Enriching very large ontologies using the WWW. In Proceedings of ECAI Workshop on Ontology Learning.

Agrawal,R.,&Srikant,R.(1994).Fastalgorithmsformining association rules in large databases. In J.B. Bocca, M.

Jarke, & C. Zaniolo (Eds.), International Conference on VeryLargeDatabases(Vol.20,pp.487-499).SanFrancisco:

Morgan Kaufmann Publishers.

Agrawal, R., & Srikant, R. (1995). Mining Sequential Patterns. In Proceedings of the 11th International Conference on Data Engineering (pp. 3-14).

Agrawal, R., Imielinski, T., & Swami, A. (1993). Database mining: A performance perspective. IEEETransactionson Knowledge and Data Engineering, 5(6), 914-925.

Agrawal, R., Imielinski, T., & Swami, A. (1993). Mining association rules between sets of items in large databases.

In P. Buneman & S. Jajodia (Eds.), ACM SIGMOD International Conference on Management of Data (Vol.20,pp.

207-216). New York: ACM Press.

Alani,H.,Kim,S.,Millard,D.E.,Weal,M.J.,Hall,W.,Lewis, P.H., et al. (2002). Automatic ontology-based knowledge extraction and tailored biography generation from the Web. Technical Report 02-049.

Alba, J.W., & Hasher, L. (1983). Is memory schematic?

Psychological Bulletin, 93, 203-231.

Anand, S.S., Bell, D.A., & Hughes, J.G. (1995, November

29-December 02, 1995). The role of domain knowledge in datamining.Paper presented at the Conference on Information and Knowledge Management, Baltimore.

Andersen, T.J. (2005). The performance effect of com- puter-mediatedcommunicationanddecentralizedstrategic decision making. Journal of Business Research, 58(8), 1059-1067.

Annesley,C.(2005).Banks’pushtoupgradehamperedby complexity of legacy systems. Computer Weekly, 14.

ANSI/NISO. (1993). Guidelines for the construction, format, and management of monolingual thesauri. National

Information Standards Organization.

Antonioletti, M., Krause, A., Paton, N. W., Eisenberg, A., Laws, S., Malaika, S., et al. (2006). The WS-DAI family of specifications for web service data access and integration.

ACM SIGMOD Record, 35(1), 48-55.

Appice, A., Berardi, M., Ceci, M., & Malerba, D. (2005). Mining and filtering multilevel spatial association rules with ARES. In M. Hacid, N. V. Murray, Z. W. Ras, & S.

Tsumoto (Eds.), Foundations of Intelligent Systems, 15th International Symposium ISMIS. Vol. 3488 (pp. 342-353). Berlin: Springer.

Apte, C., Damerau, F., & Weiss, S.M. (1994). Automated learning of decision rules for text categorization. ACM Transactions on Information Systems, 12(3), 233-251.

Compilation of References

Arkin, A. (2002). Business process modeling language

(BPML). Specification. BPMI.org.

Arkin, A., Askary, S., Bloch, B., Curbera, F., Goland, Y., Kartha, N., et al. (2005). Web services business process executionlanguageversion2.0.wsbpel-specificationdraft- 01, OASIS.

Arnoux, M., Lechevallier, Y., Tanasa, D., Trousse, B., & Verde, R. (2003). Automatic clustering for the Web usage mining. In D. Petcu, D. Zaharie, V. Negru, & T. Jebeleanu

(Ed.). Proceedings of the 5th International Workshop on SymbolicandNumericAlgorithmsforScientificComputing (SYNASC03) (pp. 54-66). Editura Mirton, Timisoara.

Ashforth, B.E., & Humphrey, R.H. (1997). The ubiquity and potency of labeling in organizations. Organization Science, 8(1), 43-58.

Aussenac-Gilles, N., & Mothe, J. (2004). Ontologies as background knowledge to explore document collections. In Proceedings of RIAO (pp. 129-142)

Aussenac-Gilles, N., Biébow, B., & Szulman, S. (2000).

Revisiting ontology design: A method based on corpus analysis. In R. Dieng & O. Corby (Eds.), Proceedings of the 12th European Knowledge Acquisition Workshop (EKAW’00) (pp. 172-188).

Aussenac-Gilles, N., Biébow, B., & Szulman, S. (2002).

Revisiting ontology design: A methodology based on corpus analysis.InProceedingsofthe12thInternationalConference in Knowledge Engineering and Knowledge Management (EKAW), Juan-Les-Pins, France.

AxIS. (2005). 2005 AxIS research project activity report.

Section ‘Overall Objectives.’ http://www.inria.fr/rapportsactivite/RA2005/axis/axis_tf.html

Backhouse,J.,&Cheng,E.K.(2000).Signallingintentions and obliging behavior online: An application of semiotic and legal modeling in e-commerce. Journal of End User Computing, 12(2), 33-42.

Barzilay,R.,&Elhadad,M.(1997).Lexical chains for text summarization. Master’s thesis, Ben-Gurion University.

Baxevanis,A.(2002).Themolecularbiologydatacollection: 2002 update. Nucleic Acids Research, 30, 1-12.

BEA, IBM, Microsoft, SAP, & Siebel. (2003). Business process execution language for Web services. Version 1.1. Specification. Retrieved May 15, 2006, from ftp://www6. software.ibm.com/software/developer/library/ws-bpel. pdf

Bechhofer, S., Moller, R., & Crowther, P. (2003). The DIG description logic interface. International Workshop on Description Logics, Rome, Italy.

Beckett, D. (2004). RDF/XML Syntax Specification (Revised),W3CRecommendation. http://www.w3.org/TR/rdf- syntax-grammar/

Beco, S., Cantalupo, B., Matskanis, N., & Surridge M.

(2006). Putting semantics in Grid workflow management: TheOWL-WSapproach.GGF16 Semantic Grid Workshop, Athens, Greece.

Ben-natan, R. (1995). CORBA. New York: McGraw Hill.

Benson, D., Karsch-mizrachi, I., Lipman, D., Ostell, J., Rapp, B., & Wheeler, D. (2000). GenBank. Nucleic Acids Research, 28, 8-15.

Berendt, B., Hotho, A., & Stumme. G. (2002). Towards

Semantic Web mining. In Proceedings of the First International Semantic Web Conference on the Semantic Web

(pp. 264-278). Springer.

Berendt,B.,Hotho,A.,&Stumme.G.(2005,September15-

16). Semantic Web mining and the representation, analysis, and evolution of Web space. In ProceedingsofRAWS’2005

— Workshop on the Representation and Analysis of Web Space, Prague-Tocna.

Bernstein, A., Hill, S., & Provost, F. (2001). An intelligent assistant for the knowledge discovery process.InProceed- ingsoftheIJCAI-01WorkshoponWrappersforPerformance Enhancement in KDD. Seattle, WA: Morgan Kaufmann.

Bernstein, F.C., Koetzle, T.F., Williams, G.J., Meyer, E.F.,

Brice, M.D., Rodgers, J.R., et al. (1977). The protein data bank: A computer-based archival file for macromolecular structures. Journal of Molecular Biology, 112, 535-542.

Berzal, F., Blanco, I., Sanchez, D., & Vila, M.-A. (2001).

Measuring the accuracy and interest of association rules: A new framework. Intelligent Data Analysis, 6(3), 221-235.