You are on page 1of 8

V. Rajeswari et al /International Journal on Computer Science and Engineering Vol.

1(3), 2009, 227-234

V. Rajeswari , Dr. Dharmishtan K. Varughese ,
Assistant Professor, Department of Information Technology, Professor, Department of Electronics and Communication
Karpagam College of Engineering, Coimbatore –32, Engg, Karpagam College of Engineering, Coimbatore – 32,

Abstract The various semantic conflicts occurring among
heterogeneous data cubes, the system architecture and related
In the contemporary business and industrial environment, the resolution procedures for all kinds of semantic conflicts are
variety of data used by organizations are increasing rapidly. Also, discussed[2]. For local data cubes with different schemas, a
there is an increasing demand for accessing this data. The size, global XML Schema is defined to integrate the local cube
complexity and variety of databases used for data handling cause structures [2] and each local cube is transformed into an XML
serious problems in manipulating this distributed information. document conforming to the global XML Schema. These
Integrating all the information from different databases into one
transformed XML documents obtained from local cubes will be
database is a challenging problem.
manipulated by pre-defined XQuery commands to form a
XML has been in use in recent times to handle data in web unified XML document, which can be regarded as the global
appliccations. XML (eXtensible Markup Language) is a very open cube. The integrated global cube can be easily stored and
way of data communication. XML has become the undisputable manipulated in native XML databases.
standard both for data exchange and content management. XML
is supported by the giants of the software industry like IBM, II. FUNDAMENTALS OF DATABASE MANAGEMENT SYSTEMS
Oracle and Microsoft. Relation schema is defined as a set of elements (attributes)
connected by a structure (table) to describe something (an entity)
The XML markup language should be the lingua franca of data in the relational database [3]. This entity should have a certain
interchange; but it’s rate of acceptance has been limited by a meaning associated with these elements, and the semantics
mismatch between XML and legacy databases. This in turn, has specify how to interpret data values stored in the schema.
created a need for a mapping tool to integrate the XML and
databases. A classification scheme is proposed for various kinds of
semantic conflicts, and it is applied to develop a query execution
This paper highlights the merging of heterogeneous database plan which optimizes multi-database queries based on different
resource. This can be achieved by means of conversion of types of schematic conflicts with the least execution cost. The
relational mode to XML schema and vice versa and by adding the conflicts are classified as value-to-value conflicts, value-to-
semantic constraints to the XML Schema. The developments that attribute conflicts, value-to-table conflicts, attribute-to-attribute
the industry has seen in recent times in this field is referred to as conflicts, attribute-to-table conflicts, table-to-table conflicts[2].
the basis. These types of conflict classifications were further partitioned
into two categories such as conflicts of similar schema structures
Keywords : XQuery, XML, Schema, XSLT, Heterogeneous
and conflicts of different schema structures.
The semantic conflicts in a heterogeneous database system
I. INTRODUCTION are further classified as schema conflicts and data conflicts[2].
The most important cause of heterogeneous database is There are two causes for schema conflicts. One is the use of
transformation and sharing of data. By using the metadata of different structures for the same concept, and the other is the use
participate sites [1], a framework is proposed for integrating of different specifications for the same structure.
heterogeneous database through XML technologies. XML is
fully compatible with applications like JAVA and it can be Semantic Conflicts
combined with any application which is capable of processing 1. Value-to-value conflicts: These conflicts occur when
XML irrespective of the platform it is being used on. XML is an databases use different representations for the same data. This
extremely portable language to the extent that it can be used on type of conflicts can be further distinguished into the following
large networks with multiple platforms like the internet and it types:
can be used on handhelds or palm-tops or PDAs. XML is an
extendable language meaning that new tags can be created or (a) Data representation conflicts: These conflicts occur when
the tags which have already been created can be used. semantically related data items are represented in different data

227 ISSN : 0975-3397

and XHTML.0" encoding='UTF-8'?> semantically related data items are named differently or <painting> semantically unrelated data items are named equivalently[2]. <caption>This is painted in 5. Rajeswari et al /International Journal on Computer Science and Engineering Vol. which is well-formed. and • Elements may nest but may not overlap. This category • Every start – tag must have a matching end tag. and date. which is a child of painting. XML AND XQUERY • None of the special syntax characters such as "<" and "&" A ) XML Introduction appear except when performing their markup-delineation roles. 2009. Atom. V. img. provided in the specification. <img src="India.Table-to-Table conflicts: These conflicts occur when </painting> information of a set of semantically equivalent tables[1] are Figure 1 XML Program represented in a different number of tables in another databases. has become the default file format for most office-productivity Existing APIs for XML processing tend to fall into these categories: tools.Attribute-to-Attribute conflicts: This occurs when <?xml version="1.0" encoding='UTF-8'?> database and as an attribute name in another database. attribute values in one database [5] are expressed as table names in another database. an • Attribute value must be quoted. generality. attribute-to-attribute conflicts. attribute-to-table conflicts[2]. for example SAX and StAX. into a unified one. it satisfies a list of syntax rules and table-to-table conflicts. • Comments and processing instructions may not appear inside the tags. C) Well-formedness and error-handling a. caption. value-to-table conflicts.e. includes value-to-attribute conflicts. 227-234 (b)Data scaling conflicts: These conflicts occur when There are a variety of programming interfaces[6] which semantically related data items are represented in different software developers may use to access XML data and several databases using different units of measure[2]. • There must be exactly on root element. Conflicts of similar schema structures. Value-to-Attribute conflicts: These conflicts occur when XML documents[5]. hundreds of XML-based languages have been A variety of APIs[5] for accessing XML have been developed[6]. Such phenomenon is They are painting. This is a string that is often found at the very beginning of 2.. date is a child of also called missing data. XML is a set of rules for encoding documents electronically. for example in web services. including Microsoft Office. for the representation of arbitrary data structures. OpenOffice. • XML data binding which provides an automated translation between an XML document and programming-language objects. schema systems designed to aid in the definition of XML-based languages (c)Inconsistent data: These conflicts occur when semantically related attributes for the same entity have different B) XML Declaration definite data values in different databases. D) Programming interfaces As of 2009. Stream-oriented facilities require less memory for certain tasks which are based on a linear traversal of an XML document. It is a textual data format. 228 ISSN : 0975-3397 . null values are usually generated. including RSS.jpg" alt=' Unity in The former case is also called synonyms and the latter case Diversity'/> homonyms. i. outerjoin operation is usually employed to integrate the tables • An element may not have two attribute with the same </caption> 6. two categories. SOAP. • Stream-oriented APIs accessible from a programming language. and some have been standardized.1(3). When integrating relations with such conflicts into a global There are four elements in this example document. 4. src The types of conflict are further classified into the following and alt. b. the same information is expressed as attribute values in one <?xml version="1. img has two attributes. For tables with conflicts of similar schema structures.Attribute-to-Table conflicts: These conflicts occur if an <date>1511</date> attribute name of a table in a database is represented as a table <date>1512</date> name in another database. Value-to-Table conflicts: These conflicts occur when the concepts. XML developed and used. XML’s design goals emphasize simplicity. III. and Apple's iWork. with example DOM and XOM. AbiWord. relation. it is widely used • Declarative transformation languages such as XSLT and XQuery. The following XML document uses all the constructs and 3. Conflicts of different schema structures. Although XML’s design focuses on documents. This category The XML specification defines an XML document as a text includes value-to-value conflicts. caption. strong support via Unicode for the languages of the world. and • Tree-traversal APIs accessible from a programming language for usability over the Internet [5].

H) Examples in XML-QL Modular XHTML lets you embed RDF cataloging information. XML The language is based on a tree-structured model of the pages are less likely to have annoying cross-browser information content of an XML document. 6. The simplest XML-QL queries extract data from an MathML equations. WHERE. E) XML on the Web The language also provides syntax allowing new XML XML began as an effort to bring the full power and structure documents to be constructed. sequences. an XMLized variant of HTML 4. Splitting up an XML document that represents multiple browsers such as Netscape 4. All these constructs are defined as envisioned. browsers. XQuery A publisher contains name and address elements and an author 229 ISSN : 0975-3397 .This flavor requires an 5. an XML-like syntax can be used nonexperts to use. Since XML documents must be well- formed and parsers must reject malformed documents.a. An article also contains a type attribute. Tree.c/bib. and it contains one shortversion from XML documents or any data source that can be viewed as or longversion element. XHTML.0 became a W3C <!ATTLIST article type CDATA> Recommendation on January 23. Selecting and transforming XML data to XHTML to be either by a CSS stylesheet or by an XSLT stylesheet that published on the Web. 2009. The second flavor of XML on the Web is direct display of 3. such as MathML and SVG. lastname)> flexible query facilities to extract data from real and virtual documents on the World Wide Web. Rajeswari et al /International Journal on Computer Science and Engineering Vol. <!ELEMENT publisher (name. A FLWOR APIs include declarative retrieval of document components via expression is constructed from the five clauses after which it is the use of XPath expressions. but its year element is XQuery provides the means to extract and manipulate data optional. named: FOR. ORDER BY. with a few notable exceptions. entries that conform to the following DTD: F) XQUERY XQuery is a query and functional programming language <!ELEMENT book (author+. Namespaces sort out which elements www. The work <!ATTLIST book year CDATA> was closely coordinated with the development of XSLT 2. They are XML document. strings.1(3). XHTML also adds a few bits of syntax to HTML. 2. address)> The mission of the XML Query[7] project is to provide <!ELEMENT author (firstname?. and more inside your XML document. A. LET. search-engine robots can more easily determine the true meaning The type system of the language models all values as of a page. transforms the document into HTML. A third option is to mix raw XML vocabularies. uses XPath expression syntax to address specific parts of an traversal and data-binding APIs require more memory. booleans. 1. and can be arbitrarily nested. 227-234 They are faster and simpler than other alternatives. XML is still a very attractive language in which to expressions within the language. The items in a sequence can either be nodes or XML on the Web comes in three flavors[7]. SVG pictures.0 by the <!ELEMENT article (author+. title.0. Therefore it finally provides the interaction Figure 2 DTD between the Web world and the database world. processing instructions and attribute names reflect the nature of the content they hold. namespaces. Like most great inventions. For instance. it omits the publisher. such as the XML declaration and empty. expressions referred to as dynamic node out to have uses far beyond what its creators originally constructors are available. Searching textual documents on the Web for relevant XML documents that use arbitrary vocabularies in web information and compiling the results. comments. write and serve web pages. Pulling data from databases to be used for the application XML-aware browser and is not supported by older web integration. The first is atomic values. V. XQuery 1. with XHTML using Modular XHTML. the two groups shared responsibility for XPath (shortversion|longversion))> 2. such as relational databases or office documents. containing seven incompatibilities. Most of XHTML can be displayed service. XSL Working Group. Generally. transactions into multiple XML documents. that all start-tags correspond to a matching end-tag and that all Uses of Xquery are as follows : attribute values be quoted.0 [7] was publisher)> developed by the XML Query working group of the W3C. one title. attribute they're much easier for robots to parse.b. author elements. 2007. Where the element and attribute of SGML to the Web in a form that was simple enough for names are known in advance. year?. database. An article is similar. Because "FLWOR[7] expression" for performing joins. and assume that it contains bibliography belong to which applications. kinds of node such as document nodes.0 that tightens up and so on. Since XML element and node. XML. RETURN. XHTML requires defined in XML Schema.0. which is a subset of XQuery 1.0. and one publisher element and has a G ) Features year attribute. Atomic values may be integers. XML [4] turned in other cases. The example XML input is in the document XHTML documents. Since XML documents are highly structured. Extracting information from a database [1] for a use in web element tags that end with />. Ultimately collections This DTD [6] specifies that a book element contains one or more of XML files will be accessed like databases. It supplements this with a SQL-like often found more convenient for use by programmers. the formatting of the document is supplied 4. text node. Generating summary reports on [3] data stored in an XML quite well in legacy browsers.xml. element node. title. the full list of types is based on the primitive types HTML to match XML's syntax. that is designed to query collections of XML data.XQuery 1.

the example abbreviation is a relaxation of the XML syntax.1(3). The variable names are preceded by $ to distinguish them Figure 4 Bibliographic Data from string literals in the XML document (like Mc-Graw-Hill).a.b. When a user poses a <author> $a </> global query on the integrated system.c/bib. For convenience. The following query returns both <title>Foundation for Object/Relational <author> and <title> and groups them in a new <result> element: Databases</title> </result> WHERE <book> Figure 5 Result of Bibliographic Data <publisher> <name>Addison-Wesley </> </> <title> $t </> <author> $a </> IV.c/bib. Figure 3. 227-234 contains an optional firstname and one required lastname. This example produces all authors of books <title> Database System Concepts </title> whose publisher is Mc Graw-Hill in the XML document <author> www.a. Often it is useful to construct </lastname></author> new XML data in the result. 2009.III string values.xml.a. The query above produces an XML document that contains <result> a list of (<firstname>.b. </lastname> </author> <publisher> WHERE <book> <name>Mc-Graw-Hill</name> <publisher><name>Mc-Graw-Hill </publisher> </name></publisher> </book> <title> $t </title> <book year="2003"> <author> $a </author> <title>Operating Systems </title> </book> IN "www.2 Element Patterns .b. this query matches every <book> element in the <lastname>Williams</lastname></author> XML document www.xml" <author> CONSTRUCT $a <lastname>WilliamStallings </lastname> Figure 3 Element Patterns . The </> name.xml" <result> CONSTRUCT $a <author><lastname>WilliamStallings </lastname></author> Figure 3. I) Matching Data Using Patterns.e. Any URI that represents an XML-data source <lastname>Silberschatz Korth Sudharsan may appear on the right-hand side of IN. address. i. For each such </publisher> match. SYSTEM STRUCTURE </> IN "www.c/bib.a.xml that has at least one <title> <publisher> element one <author> element.b.<lastname> ) [5] pairs or (<lastname> ) <author><lastname>Williams elements from the input document.A good introductory text --> XML document.I </author> <author> Informally. </element> can be abbreviate by </>.1 Element Patterns .a. <bib><book year="2006"> XML-QL uses element patterns[7] to match data in an <!-. This When applied to the example data in Figure 4. V.. it binds the variables $t and $a to every title and author </book></bib> pair.c/bib. firstname and lastname are all CDATA.c/bib.xml" The data warehouse creation [1] module can be regarded as CONSTRUCT <result> user of the heterogeneous database system. Rajeswari et al /International Journal on Computer Science and Engineering Vol. and one <publisher> element <name>Mc-Graw-Hill </name > whose <name> element is equal to Mc Graw-Hill.b. the global site <title> $t </> decomposes the global query into sub-queries to request each 230 ISSN : 0975-3397 . Thus the query query would produce the following result: above can be rewritten as: <result> <author><lastname>SilberschatzKorth WHERE <book> sudharsan</lastname> <publisher><name> Mc-Graw-Hill </></> </author> <title> $t </> <title>Database System Concepts </title> <author> $a </> </result> </> IN "www.II <title>Operating Systems</title> </result> J)Constructing XML Data.

together with some necessary template files.location.. and ProductsData. Sites X and Y XSLT and template files are shown in Figure 10 and Figure 11 contain tables named products and productslist respectively.component is /></TD> missing in the relation in Site X are respectively xmlns:xsl=" named productlist. This can be easily achieved in contemporary commercial database products (e. V.pno and Products. ProductData as Figure 12 illustrates. There are three value-to-value conflicts: The products. in advance to transform local data into the global schema format.0"> 2. For Site Y. manufacture_year).pid and productlist. cost. The two tables products and prductlists can be integrated as ProductsData(pno. Besides. There is a table-to-table conflict: The products. When the user wishes to show ProductsData. components. location. both tables can be outer-joined into There are some semantic discrepancies between these sites. but the productlist.manufacturing_year is also missing in the relation /></B></TD> products. <TR><TD><xsl:value-of select = '@pno' 3.w3.cost stores <xsl:template match = '*'> the list price of every product. The </xsl:template> productlist. Figure 8 and Figure 9 There are two databases containing semantically related show the XSLT and template files for Site X. Rajeswari et al /International Journal on Computer Science and Engineering Vol. Finally.location stores more detailed data than <xsl:template match = 'products'> products. 2009. <TD><B><xsl:value-of select = '@company' /></B></TD> <TD><B><xsl:value-of select = '@cost' /></B></TD> <TD><B><xsl:value-of select = '@dealer' /></B></TD> <TD><B><xsl:value-of select = '@location' /></B></TD> 231 ISSN : 0975-3397 . ProductsData. the DBA in each site should prepare some XSLT[2] format files. dealer. Figure 7 Two Sites with Conflicts of Similar Schema Structures. Finally. company. respectively. MS SQL Server 2000.1(3). Besides. sform" version="1. name.productname in Site Y. The integration process utilizes the semantic knowledge of all participate local Figure 6 Heterogenous Database Architectures by the full names.g. the transformed data are integrated into a consolidated view with the global query applied on it to return global data to the user.location by the concise names two sets of XSLT and template files are used to transform A) Two Databases With Different Semantic Conflicts both tables into ProductData. They are listed as follows: 1.cost stores the <xsl:apply-templates /> discounted price (with 40% off). the information about books but in different formats. IBM DB2 Extender and Oracle).cost by the original list price. 227-234 participant to return the data in XML format[3]. This should be prepared or discovered before integration. the <TD><B><xsl:value-of select = '@name' productlist. respectively. There are two attribute-to-attribute conflicts: The attributes <xsl:stylesheet products. Besides.

productlist. INTEGRATING DATA FROM MULTIPLE XML <TD><B><xsl:value-of select = '@dealer' SOURCES /></B></TD> In XML-QL.xml Fig 11 The Template file for Site Y. /></B></TD> <TD><B><xsl:value-of select = '@cost' /></B></TD> V. company. right(rtrim(location). productname.Products. name.xsl com:xml-sql" sql:xsl="productlist. 2009. dealer.productlist. Rajeswari et al /International Journal on Computer Science and Engineering Vol.xsl"> <sql:query> <ROOT xmlns:sql="urn:schemas-microsoft. dealer.xsl </xsl:template> </xsl:stylesheet> <ROOT xmlns:sql="urn:schemas-microsoft- Figure 8 The XSLT for Site X . cost.4 com:xml-sql" sql:xsl="Products.0"> i ame y atio onent ctureyea <xsl:template match = '*'> d n s r <xsl:apply-templates /> </xsl:template> 1 MotherB Heiss 2000 IBM Indi 100 2000 <xsl:template match = 'productlist'> oard Compan a <TR> y <TD><xsl:value-of select = '@bid' /></TD> 2 Micro Joe 2500 Microns US 65 2009 <TD><B><xsl:value-of select = Controll Compan A '@productname' /></B></TD> ers y <TD><B><xsl:value-of select = '@company' Fig 12 Integrated Relation ProductData in the Global site. components FROM products manuafactureyear FOR XML AUTO FROM productlist </sql:query> FOR XML AUTO </ROOT> </sql:query> </ROOT> Figure 9 The Template file for Site X . In this example.'> <TR><TH colspan='7'> ProductsData <TR><TH colspan='7'>Productdata</TH></TR> </TH></TR> <TR><TH>pno</TH><TH>name</TH><TH>company</ <TR><TH>pno</TH><TH>Name</TH><TH>compnay</ TH><TH>cost</TH> TH><TH>cost</TH> <TH>dealer</TH><TH>location</TH><TH>manufa <TH>dealer</TH><TH>location</TH><TH>compon ctureyear</TH></TR> ents</TH></TR> <xsl:apply-templates select = 'ROOT' /> <xsl:apply-templates select = 'ROOT' /> </TABLE></BODY></HTML> </TABLE> </xsl:template> </BODY> </xsl:stylesheet> </HTML> Figure 10 The XSLT for Site Y .xml <xsl:stylesheet xmlns:xsl="http://www. SELECT pid.xsl"> as cost. 227-234 <TD><B><xsl:value-of select = <TD><B><xsl:value-of select = '@components' /></B></TD> '@manufactureyear' /></B></TD> </TR> </TR> </xsl:template> </xsl:template> <xsl:template match = '/'> <xsl:template match = '/'> <HTML> <HTML> <HEAD> <HEAD> <STYLE>th { background-color: #CCCCCC <STYLE>th { background-color: #CCCCCC }</STYLE> }</STYLE> </HEAD> </HEAD> <BODY> <BODY> <TABLE border='1' style='width:600. several sources may be queried simultaneously <TD><B><xsl:value-of select = '@location' and produce an integrated view of their data. 2) as location.w3. V. <sql:query> rtrim(publisher) + ' Company' as dealer. location.'> <TABLE border='1' style='width:600. we /></B></TD> 232 ISSN : 0975-3397 . cost/0.1(3). SELECT p productn compan cost dealer loc comp manufa sform" version="1.Products.

CONSTRUCT <result ID=ResultID($p)> <month> $m </> </> WHERE <person> } <name></> ELEMENT_AS $n { <ssn> $ssn </> WHERE $e = "book".gov/taxpayers. into blocks and sub-blocks: the latter are enclosed in braces ({..xml" XML – based heterogonous data integration and query has CONSTRUCT <result ID=SSNID($ssn)> <ssn> been a hot technology nowadays.xml" mapping from heterogenous database integration schemas to CONSTRUCT <result ID=SSNID($ssn)> ssn> XML documents can be prepared according the proposed procedures. The second sub block checks if e is a book and has a publisher.b. transforming the information system of enterprise { WHERE <taxpayer> inside. persons that are not taxpayers.c/data. in addition to the previous query.xml. the root block is empty and [5] www.a. the <income> will be appended to the object rather than included in a new <person>. and their semantics is related to that of outer joins in Heterogeneous Databases Integration” relational databases.xml". pp. The first sub block but not equivalent. we can realize integrating. $ssn </> $i </>} Figure 14 Fragmented Queries REFERENCES [1] Frank S.c/data. In the second sub block.xml and </> CONTENT_A $p IN "www.. The following query retrieves all titles published in 2007 from the bibliography database. XML-QL queries are structured Journal of CIIE. In the first sub block. “An Interactive Tool Based on XML join of the entities "www.irs..c/data. sharing and utilizing of the <ssn> $ssn </> hetrerogenous data source differently and effectively. which has <income> </> ELEMENT_AS $i improved the whole efficiency of the Information System. This query contains.b. it adds the month element to the same publication. If so. Through the Intranet and XML $ssn </> $n </>} Technology.273-278(2005). Heterogeneous database Integration Using XML. Kuiheyang “Research and Realization of XML Based }). produces a result element with their title. person's incomes are produced.a. </> IN "www.b. Query blocks are powerful.a.C. Volume "www. and adds that publisher { WHERE <person> element <name> </> ELEMENT_AS $n VI. The two sources are joined on the CONSTRUCT <result ID=ResultID($p)> <title> social-security number. and vice versa.c/bib.xml[7].a.c/data.C. Wherever there is a match with a previously generated OID.4.w3schools. No.xml" and Technology for Data Exchange between Heterogeneous ERP systems”. Rajeswari et al /International Journal on Computer Science and Engineering Vol. <publisher>$q </> IN $p <taxpayer> CONSTRUCT <ssn> $ssn </> <result ID=ResultID($p)> <publisher>$q <income></> ELEMENT_AS $i </> </> </>IN "www. [4] Yanxinwang. CONCLUSION <ssn> $ssn </> </> IN "www.The </> IN "www.irs. The result contains only those elements that have {WHERE $e = "journal-paper". 233 ISSN : 0975-3397 . all [2] Frank S. XML based heterogeneous Database Integration for Data warehouse Creation. Each result has a unique OID given by that Technologies” Dec.irs. all persons' [6] Giovanni Guardalven “ Integrating XML and Relational Database names are produced. <month> $m both a name element in the first source and an income element </> IN $p in the second source.b. This is the outer [3] Wei-Jung Shiang.xml". In this" } CONSTRUCT <result> <ssn> $ssn </>$n $i </> Figure 15 Output from Bibliography database Figure 13 Integrating data from multiple XML sources The outer block runs over all publications in 2007.xml" www. and Alternatively. 227-234 produce all pairs of names and social-security numbers by WHERE <$e> <title> $t </><year> 2007 </> querying the sources www. and it also retrieves the publication month for journal articles and the publisher for books. Minng-Ying Ho.1(3). persons and their [7] www. V. which is bound to SSN in both $t </> </> there are two sub blocks[7]. query is written in two fragments: checks if e is journal-paper and has a month tag. Skolem functions can be used and a related. Tseng.irs. 2004.

His areas of interest are Database Management Systems & Antenna and wave propagation. 234 ISSN : 0975-3397 . Rajeswari et al /International Journal on Computer Science and Engineering Vol. Karpagam College of Engineering. 2009. 227-234 Author’s Biography Ms. Rajeswari is Assistant Professor in the department of Information Technology. V. Karpagam College of Engineering. Her areas of interest are J2EE and Database Management Systems. Varughese is Professor in the department of Electronics and Communication Engg.1(3). Coimbatore. Dharmishtan K. V. Dr. Coimbatore.