Steps of Designing Biodiversity Data Model Using ObjectRelational Data Model Techniques Mohd Taib Wahid1, A.Z.M Kamruzzaman2, Abdul Rashid Mohamed Shariff3 Harihodin Selamat4 1
[email protected],
[email protected] Department of Information Systems, Faculty of Computer Science & Information Systems University Technology of Malaysia, K.B. 791 81310 Skudai, Johor, Malaysia Tel: (607)-5532422, Fax: (607) 5565044
ABSTRACT The economic importance and uses of the large number of biodiversity data of Malaysia make it essential for their biodiversity to be conserved. The complexity of natural history collection information and similar information within the scope of biodiversity informatics poses significant challenges for effective management of biodiversity data. In order to undertake the steps for biodiversity conservation such as identification of species, monitoring climate conditions, it is essential to efficiently manage the vast amount of biodiversity related data. This paper discusses, first: discuss about biodiversity data model and secondly: object-relational conceptual design of biodiversity data model for long term stewardship of biodiversity information. A plenty of data models have been developed which only support spatial data. In addition to this, most of the data models are relational model. The purpose of this research is to design a biodiversity data model that better facilitates the exploration and analysis of spatio-temporal biodiversity data by using object relational data model techniques. The goal of this study is to design a biodiversity model that is efficient for data management and able to integrate diverse set of spatio-temporal data that maybe required for analysis, and monitoring biodiversity data. Keywords: Biodiversity Data, Data model, Object-relational Model, Spatio-temporal
1. Biodiversity Data Biological diversity plays a very important role in our lives. There are various definitions on biodiversity has been given by many researchers. Perhaps the best definition on biodiversity is the following (adapted from the Keystone Dialogue on Biodiversity in Federal Lands by Noss and Cooperrider, 1994): Simply put, biodiversity is the assortment of different types of organisms that co-occur in time and space. Biodiversity data is an assortment of different types of taxonomy or manifest itself on the genetic species, environment, landscape levels, manipulated to analyze the past, define the present and the consider possibilities of the future. Furthermore, all biodiversity data is an assortment of different types of organisms that co-occur in time and space [1].
2. Biodiversity Data Model During the past decade, research has been done in the area of biodiversity data. Malaysia is undoubtedly 12 mega diversity countries in the world. Several local organizations all over the country are working on the implementation of the biodiversity measures. One of the major hurdles faced by conservation scientists and researchers is the management of vast amount of biodiversity information. This information includes data on species characters (flora, fauna and genetic organism, peat swamp), geographic distribution of species, and geographic regions etc. It has been reported that over 15,000 of the described flowering plants (approximately 9% of the world’s total) and 185,000 described animal species (approximately 15.7% of the world’s total) exist in Malaysia (EPU, 1993) [2].
Broadly speaking, the above biodiversity data may be classified into three major groups: there taxonomy data, geo-spatial data and temporal data. These data types have complex and deeply-nested relationships within and between themselves. An important point to be noted is that data in these groups are both intrarelated and inter-related. For example, the geographical distribution of a species relates the taxonomic data and the geographical data [3]. Several projects have been undertaken by various organizations nationally and internationally, with the aim of addressing the biodiversity data management problem. Beside, data model for botanical collections for taxonomic databases have been developed by many researchers at various places since 1992, e.g. ASC (1992), Bolton et al. (1992), Sinnot (1993), Wilson (1993), NMNH (1994), and ITIS (1995) and APMIS (Alian Plant Management Information System) [7]. The researchers attempts to bring order into the complex data structure which are involve when plants are named, collected, classified and investigated as to their properties. Bodhi (Biodiversity Object Database architecture) is designed as a data model based on Indian plant biodiversity [3, 4]. Recently the academy of natural science of America developed a relational database which is implemented for biodiversity [5]. There are few others data model developed based on Malaysian plant such as Ethno-botany of Malaysia Plants Online [2, 6], and data model for botanical collection [2]. Based on our study on the above mentioned data models, biodiversity and GIS data model is relation. Relational data model cannot support huge (various level of data relation) data, complex data analysis and data manipulation, and time factor. After analyzing of all the models developed since 1993, most of the models are used to collect plant data, plant listing, and plant conservation. There is no data model exists which can support data
analysis, data retrieval, and temporal data (time). Early 2000, biodiversity object database arcHItecture (BODHI) was developed to handle plant taxonomies. To support spatial and temporal of plant biodiversity data, one robust data model is must be developed. Bio-diversity scientist faced many difficulties are the effective management and access of the large amounts and varied types of data that arise in their studies, ranging from micro-level biological information such as genetic makeup of organisms and plants. This problem is address in the Oshadhi project, which developed an object-based database system for efficiently handling plant taxonomies. Year after [8] developed an internet based next generation biodiversity information system, called BODHI (Biodiversity Object Database ArcHItecture). Joseph Maada Korsu Kendeh [7] had developed an Alian Plant Management Information System (APMIS) based on South Africa’s invasive alien plant species. In that system he has design the conceptual data model of the prototype geographic information system called APMIS. We also have been analyzed Pahang and different area from Johor state’s data and their databases system. In the Pahang forest database systems are currently storing biodiversity data such as flora and fauna, area, weather and water condition data. This system has been storing data with simple designing of tables. This system contains with 20 tables which are hard to manage and make relationship. Beside that, system is only storing general information about those mentioned data, does not support monitoring, and analysis data for prediction or making decision. From the Johor state, we have collected different year’s data including floara and fauna from Johor Park Corporation, Tanjung Piai, Pulao Kukup national park, Endau-Rompin National park. They have huge amount of biodiversity data which are collected traditional from different places
and different times. After the suggestion from those departments and analyzing biodiversity data, decision has been made which is consist designing of biodiversity data modeling to manage, to store, to analysis and monitoring biodiversity data including spatial and temporal factor. In studies of biodiversity data modeling, few problems have been categorized about biodiversity data modeling. First, large amount of data, biodiversity data is too large because of the variety of different types of species that growth in various landscape also co-occurs in time and space. Facilitate focus attention on specific area of plan biodiversity need to conserve. Second, biodiversity data are complex because biodiversity includes the variety of genes within one species through the complex interconnection of all life within an environment. There is no stable biodiversity database system and biodiversity analysis tool to maintain infrastructure and monitoring current and future data of biodiversity. Third, effective data management system means storing, retrieve and accessing data from the system. There is no such model exists which can support data management and time consuming accessing data from the system. Fifth, existed data model cannot support temporal (time) data such as Bodhi and Oshadhi data models. These models are object-relational model those do not support temporal data, only support spatial data. Sixth, most of the data models are relational data model. Rational data model cannot support temporal data, only support spatial data. Seventh, there is no proper retrieval system designed yet for the biodiversity data. Existed system only provided general information about biodiversity. Complexity of spatial data relationship which is adjacent, overlap, disjoint containments. These problems come when systems want to retrieve data about all areas biodiversity data. For example, in this case the entire polygon objects that overlap with another set of polygon objects that need to retrieve. Mostly all the queries in spatial domain like overlaps,
containments, include joints. This complexity and dominance of spatial queries in biodiversity application makes it essential to evaluate the spatial module in the system. Example of spatial relationship figure has shown in figure 1
Figure 1: spatial relationship
Another problem is temporal pattern of Biodiversity. There are various aspects of data values with respect to time. The time at which data values effective is called valid time. On the other hand, transaction time represents the time a data value is recorded in the database (R. Snodgrass, 2000). So the time varying data is commonly represented by time-stamping values. The time-stamps can be time points [9, 10], time interval [11, 12-14, 15], temporal sets [16] or temporal elements [17]. Further more, these time-stamps can be added to tuples or attributes, the two different approaches for handling temporal data in a relational data model. Time points, intervals, temporal elements, and temporal sets are used as attribute timestamps. Generally, the end points of an interval are added as two separate attributes to a relation. Time points, time intervals, temporal sets and temporal elements are all needed in temporal query languages. They all added to the expressive power of a temporal data model and its query languages [18] From the above discussion, it is clear that the biodiversity domain requires data management support. One possible approach is to use the popular objectrelational database management systems, which have been proved highly effective for commercial applications such as banking, inventory control and airlines
Taxonomy
reservation. However, they are not well suited for recent complex and advance database applications such as geographic information systems, computer aided design, and network management systemsbiodiversity data management also falls into this category of applications that are not amenable to object-relational database management system. This is because biodiversity (and the other) applications differ significantly from the commercial applications in both their modeling and processing requirements. 3.
Development methods
3.1 Pyramid framework and event based approach In the design of object-relational biodiversity data model to support spatiotemporal data, pyramid frame can make enhancement to support a biodiversity data model design. A conceptual framework (also known as pyramid framework) was designed which guide the implementation of the semantic geographical information system (GIS) data model [23]. Conceptual frame work (pyramid) is interrelated with two separate parts one is data component and another is knowledge component (Figure 2). Data component can be divided into three categories: location (position in the spatial three dimensions), time (position alone a time line) and theme (a set of observations, measurements, or attribute values associated with a particular location and time). Data components stores spatio-temporally referenced observational data such as spectral, climate, vegetation and other environmental attributes that maybe queried and visualized to reveal embedded spatio-temporal pattern and relationships. Data component can be computerized as a multi-dimensional, spatio-temporal referenced ‘hyper-cube’ of observational data that is similar to the ‘feature space’ concept commonly cited in analysis of remote sensing imagery.
What is it?
Knowledge Component
Object
What is it made of?
Theme
Where is it?
Location Y
When is it?
Data Component
Time T1
T2 T3
X
Figure 2 Pyramid Frame work: Data Component and Knowledge Component Knowledge component stores information about higher-level semantic ‘objects’ the geographic entities or process that are describe by the data. Information concerning an object’s location, time and composition. All the objects are also placed within two hierarchical relationship structures central to cognitive knowledge representational and object-oriented modeling, plant taxonomy (generalization) and partonomy (aggregation). The taxonomy structures groups similar objects within a category and stores a rule-base that describes how those objects may be identified within the data space. These rules may be derived from expert knowledge or from inductive analysis of the observation data. In an object-relational data model consists of a set of object classes (of different types or schemas). Each object class has an associated set of objects; each object has a number of attributes with values drawn from certain domains or atomic data types. Of course, there are additional features, such as object valued attributes (Oid), methods, object class hierarchies, etc. Besides objects, attributes describing geometries including time are of particular interest. Hence we would like to define collections of abstract data types, or in fact many-sorted algebras containing several related types and their operations, for spatial values changing over time. This section presents a simple and expressive system of abstract data types, comprising data types and encapsulating operations, which may be integrated into a query language, to yield a
biodiversity data). As an example in a spatial data model is implemented within a relational DBMS (i.e., Database Management System), which raises database requirements for the data model to consists of three data model components –spatial data types (e.g., points, lines, polygons), spatial operations (e.g., collection of biodiversity data from different places), and relational integrity constraints (including accessing methods). Specific user needs as functional requirements are proposed for different applications. For example, biodiversity data and forest area maps are specified as spatial data types (i.e., point, lines polygons) according to the need from exploring spatial and temporal data relevant to biodiversity data interaction. Operations and constraints corresponding to the interaction can also be developed for handling and facilitating biodiversity data management and manipulations.
powerful language for querying spatiotemporal data. To support temporal data, system needs another model such as event-based model. Pyramid system is a model to support multi-dimensional, spatio-temporal and geo-graphic objects (such as location and space). All the objects are also placed within two hierarchical relationship structures central to cognitive knowledge representational and object-oriented modeling, plant taxonomy (generalization). The plant taxonomy structures groups of similar objects within a category and stores a rule-base that describes how those objects maybe identified within the data space. Pyramid framework consists of main three components; objects, location and time. These features will be referred to accessed data from the database. This framework can be converted to table in stated figure 3. Object Object
Location
Time
Value n
Value n+1
Theme Location Y
Time T1
T2
T3
X
Event 1
Event 2
……
Event n
Figure 3: From the frame work data transform to the table and time modified as events 3.2 Time extension in biodiversity data model Figure 4 depicts a generic data model and its components. This figure indicated that the data model captures essential information from the real world based on function and database requirements [19]. The functional requirements (indicated as hatched arrows) refer to the user-defined data types, operations, and integrity constrains that will be customized and applied to the data model. The database requirements (indicated as gray arrows) directly lead to the specification of data types, operations, and integrity constraints for real world object (in this case
Alternatively, a data model M can be described, modified, and accessed in a uniform way, is expressed as M= (DT, OP, and C). The data model M includes three components data types DT, query enhancing operation OP and integrity constraints C [20]. A corresponding spatial data model enhancing the three components of the data model can be expressed as MS= (DTS, OPS, CS). Meanwhile, as a spatial-temporal data model extended from MS is denoted as MST = (DTST, OPST, CST) where time is stamped o every component of the spatial data model. Its data types accommodate time-varying spatial data. Its query
operations are redefined with spatial, temporal and spatio-temporal functions [21]. Additionally, each integrity constraint in the spatial data model CS has a temporal version CST. The data types DT are the first, compared with the operations OP and the constraints C, to reflect changes enforced for applications. The literature reveals that spatio-temporal data models developed form spatial data models usually start with the design and implementation of new data types and data structures. Query languages and access methods, although they are components equally important to the data types in the data model, are the next for modification.
3.3 Objectrelation data model techniques The object relational model (ORM) is a way of describing or representing objects, classes of objects, relationships between objects and classes, and real world memberships. The ORM can be considered the static part of object-oriented systems modeling. That is, the ORM describes the “database” of a model: what objects may exists, what objects classes they belong to and what relationships exists between objects. In the object-relational model also consists of object, object classes, relationships relationship sets and constraints.
Real world
Database Requirements Function Requirements
Data Types
Operations
Constraints
Figure 4: Data Model Components (source: Steiner, 1998)
Usually, extending the data structures with time attributes does not cause any severe problems. When time-stamping data; two different time dimension can be distinguished. Valid time records time when data was true in reality. Transaction time records when data has stored in the system. To store valid time data, two additional attributes of type Date, VTS (Valid Time Start) and VTE (Valid Time End), can be added to (maybe already existed) non-temporal data structures denoting the start and the end point of valid time. The same can be done for transaction time. However, the idea presented could easily be generalized to deal also with transaction time.
3.3.1
Objects Types:
An object includes a static part, a set of data, and a dynamic part, a set of procedures manipulating this data. an object is defined by its behavior, and represented by the set of its procedures, rather than by its structure. An object has a unique and immutable identify that differentiates it from other objects. An object identity (OID) comes into being when the object is created and can never be confused with that on an other object even if the original object has been deleted. The OID never changes even if all the properties of the object change. It is depend on the object’s state. An object identifier further has the following characteristics: 1) it identifies an object not
only in an extension of a class, but also in database, 2) it is constant throughout the lifetime of the object, 3) it is not visible to the user, and 4) it is created and maintained by the system. An object has a set of attributes (Instance variables), and relationship with other objects. An attribute models the state of an object. Spatial and/or temporal information may be associated to objects, independently from the characteristics of their attributes. Consequently, an object type can be plain (neither spatial nor temporal), spatial, temporal, or spatiotemporal. The term object, however, is fundamentally different from the term entity. Entity is concerned merely with data. We typically store a record from each entity. Object concerned with data and the methods with which it is manipulated. For example, in biodiversity data model different object represents the different real world entities as shown in Figure 5 Biodiversity Data
Taxonomy Data Geospatial Data Temporal Data
Geo-region
SpatioObject
Point
Line
SpatialCollection
Polygon
Figure 5: Object type and spatial data as object There are two types of object diagrams: class diagrams and instance diagrams. A class diagram is a schema, pattern, or template for describing many possible instances of data. A class diagram describes object classes. An instance diagram describes how a particular set of objects relates to each other. Example of a class diagram has shown in figure 7.
3.3.2
Abstraction mechanism
The object-relational data model is built on few basic concept of abstraction: relationship, generalization, association and aggregation. These abstraction concepts may be described by the means of relationship sets. A relationship set is a set of relationships, all of which have the same role to between objects. In an objectrelational diagram, a relationship set is represented by lines connecting associated objects. As for objects, relationship may be located in space and time, via two attributes. In this case they are referred to as spatial and/or temporal relationships. CollectionObjects
Identifications
Objects_ID
Identification_ID
IdentifiedBy Date_of_ident
Taxon_name Taxon_id IdentfiedBy
Figure 6: Identifies relationship
3.3.3 Generalization and specialization Generalization and specialization represents the “Is-a” relationship set, an essential element of object-relational paradigm. The main idea in generalization and specialization is that one object class (specialization) is a subset of another (the generalization). The process of generalization consists of putting classes together in what is called a superclass. Generalization enables us to perceive that all instances of more specific concept are also instance of general concept. Specialization, on the other hand, breaks apart a class by differentiation of properties or forms a set of lower class called subclass. Specialization allows us to describe each subclass of a more general class by specifying only the additional details necessary for us define. Subclass describes a specialization of the superclass. The terms superclass and subclass characterize generalization and specialization. It is important to note that superclass an subclass are abstraction of the same object, and do not describe two different objects. The direction of the “Is-
certainly not a new concept. Associations are widely used throughout the database modeling community for years.
a” relationship goes from the specialization to the generalization, that is, it may be stated as, “Specialization Class is-a Generalization Class”. Figure 8 shows Generalization and Specialization for the data model design. Flora_details Flora_ID Author Identifier Date_of_Idnt Area_code Area_name Family_name Species_name Genus Class
Flora
Fauna
Flora_ID Flora_code Com_Name Scientific_Name Local_name notes
Fauna_ID Com_name Scientific_Name Local_name notes
Fauna_Details Fauna_ID Author Area_Code Author Obs_Date Date_of_idnt Area_name Identifier Family_name Species_name Genus Class
Area Area_Code Area_name Vertical_position Horizontal_possition Area_Type Details
Organization Org_Code Org_name Area_Code Address Data_type Telephone web
Survey Survey_key Survey_code Data_type Date_start Date_finished Author notes
Figure 7: class diagram for biodiversity data model
Biodiversity Data Generalization
Is-a
Specialization
Taxonomy data
Geospatial data
Temporal data
Figure 8: Generalization and Specialization Association An association describes a group of links with common structure and common semantics. A link is a physical or conceptual connection between object classes. All the links in an association connect objects from the same classes. Association and links often appear as verbs in a problem. An association describes a set of potential links in the same way that a class depicts a set of potential objects. In fact, the notion of an association is 3.3.4
In an object-relational model diagram, association is also known as the “member of” relationship set. For example, clubs and teams are usually modeled using association. Accordingly, an object in the Taxonomy data object class is a set of objects from the PlantSpecies, IdentCharacteristic, IdentLevel. The relationship set is read “PlantSpecies is a member of Taxonomy Data”.
it is not obvious if an association should be modeled as an aggregation [22]
Taxonomy Data Member - of Species
IdentCharacteristic
IdentLevel
Figure 9: Association
3.3.5 Aggregation An aggregation represent the “a –part-of” or “is subpart of” relationship set, in which objects representing the components of something are associated with an object representing the entire assembly. Aggregation is a strong from of association in which an aggregate object is made of components. Components are part of the aggregate. The aggregate is semantically an extended object that is treated as a unit in many operations, although physically it is made of several lesser objects. A single aggregate object may have several parts: each part-whole relationship is treated as a separate aggregation in order to emphasize its similarity to association. For example, Geospatial object in as aggregation of the SpatialColelction, point, line and polygon that are the part of it. GeoSpatial Data Part - of
SpatialCollection
Point
Line
Polygon
Figure 10: Aggregation
However, there are may be some discussions of the differences between association and aggregation. Aggregation is a spatial form of association, not an independent concept. Aggregation adds semantic connotations in certain cases. If two objects are tightly bound by a partwhole relationship, it is an aggregation. If the two objects are usually considered as independent, even though they are often linked, it is an association. But, the decision to use aggregation is often arbitrary a being matter of judgment. Often
3.3.6 Constraint Most of the system can be model with objects, objects classes, and relationship sets. Constraints are functional relationships between entities of an object relationship model. Constraints provide one certain for measuring the quality of an object relational model. In object relational model, a constraint restricts the membership of one or more object classes and relationship sets. The object relational model allows analysis to express several different types of constraints. However, in the object behavior model and interaction model, it may be useful to use this constraint in order to describe some restricted condition.
4. Conceptual Biodiversity Data Model Structure After explanation of several components of object relational model, this section shows some relationships among the components of object relational model for biodiversity. The proposed spatio-temporal data model is described by means of collection of one or more class diagrams that form the object model and connected with its related relational tables to form an objectrelational data model. The class diagram which describes the structural characteristics of the proposed spatio-temporal data model is presented in Figure 11. It shows that five major classes were identified and incorporated into the data model: BiodiversityFeature (BF), BiodiversityData (BD), Spatial Data (SD), Taxonomy Data, and Temporal Data (TD). Object-relation is the most powerful techniques to design object-relational data model. The design of an object-Relational Data Model generally involves two basic steps. First, structural modeling, which describes the structure of similar objects in terms of classes, their similarities and differences (generalization), the
associations or connections among these classes, and the structural constrains. The second step corresponds to behavioral modeling, which describes the behavior of different classes in terms of operations and relationships
This paper has reviewed biodiversity data models and spatio-temporal data modeling approaches and tries to overcome problems with temporal factors in biodiversity data. We have also presented two general approaches to design spatio-temporal data model in object relational techniques.
BiodiversityFeature BioFeatureID: IdentifierType BioFeatureName: NameType
BiodiversityData
Biodiversity Data
Biodiversity Feature
BiodiversityID1[]: identifierType BiodiversityName[1..*]:NameType
TaxonomyData SpatialDataID[1]: IdentifierType SpatialDataName[1]:NameType Color[1]: ColorType
TemporalD[1]:IdentificationType TemporalReferencing:Reference TemporalResulution:
DateTime
Texonomy Data
SpatialData
TemporalData
TaxonName[1]:NameType TaxonID[1]: identifierType TaxonValue[1]: AttributeType taxonDataType[0..1]:NameType TaxonIdntDate[1..*]:DateType
Date
Time
Day Month year
Hour Minute Second
Spatial Data
Interval
YearMonthInterval Month Year
Time-stamp
DayTimeInterval Days Hours Minutes Seconds Milliseconds
Figure 11: Detailed Structure of Spatio-temporal Object-Relational Model for Biodiversity Object-relation is the most powerful techniques to design object-relational data model. The design of an object-Relational Data Model generally involves two basic steps. First, structural modeling, which describes the structure of similar objects in terms of classes, their similarities and differences (generalization), the associations or connections among these classes, and the structural constrains. The second step corresponds to behavioral modeling, which describes the behavior of different classes in terms of operations and relationships
5. Conclusion
The first approach is to design data model, thereby supporting efficient storage and processing data as well as an ability to handle temporal and spatial data. The second approach is to design objectrelational data model for the management of biodiversity data. This model is built up from the basic attributes and behaviors of spatial temporal objects. Based on the object-relational strategy, it’s extensible to any applications by generalization, specialization, association and aggregation. Thus, this model can completely support the multiple semantics of space and time.
Although the research has made a progress, much work still needs to be done. There are few more steps should be taken immediately to protect and manage vast amount of biodiversity data those steps are: • Development of query languages for spatio-temporal biodiversity data. • Graphical user interfaces for spatiotemporal biodiversity information. Efficient implementations will be carried out in future research. Work on specification of integrating space-time query language is also under way.
References [1] Noss, R. F. and A. Y. Cooperrider (1994), “Saving Nature's Legacy”, Island Press, Wash. D.C. 416 pp. [2] Mat Saleh, K. & A Latif, (1997), “Towards a Comprised Database of Malaysian Medicinal and Aromatic Plants” 1997 [3] Srikant B.J, Jayant R. Haritsa (2000), “Design and implementation of Biodiversity Information Management System”, 2000 [4] Srikant B.J, Jayant R. Haritsa, Uday S. Sen (2000) “The Building of BODHI, a Biodiversity Database System”, 2000 [5] Paul J. Morris (2005), “Relational Database Design and Implementation for Biodiversity Informatics”, The Academy of Natural Sciences1900 Ben Franklin Parkway, Philadelphia, 2005, PA 19103 USA
Geographical Information Systems for Rural Applications, March 2002, international institute for geoinformation science and earth observation, enschede, the Netherlands [8] Srikant B.J, Jayant R. Haritsa, Uday S. Sen, 2000 “The Building of BODHI, a Biodiversity Database System” [9] Adnan Younas, Mohd Noor, Mohd Taib, Mohd Shafry (2004). An Effective Spatial and Temporal Data Management Using: Spatio-temporal Object-Relational Model (STORM), ISG’04, Kuala Lumpur 21-23 September. [10] Dangermond, J. (1990). A classification of software components commonly used in geographical information systems, Pequet, D., & Marble,D.(Eds.),Introductory readings in Geographical Information Systems. (pp. 31-51). London; New York: Taylor & Francis. [11] Bellamy, S.P. (1996). MHIS Developtment, Database/ArcInfo LinkHeigh Level Design, Technical Report, Issue V1.R1.M0 ( 24/7) [12] Benoit, D., Voisard A., (1993). A Unified Approach to Geographic Data Modeling, Technical Report 9316, University of Munich (LMU). [13] Blaha, M., & Premerlani, W. (1998). Object-Oriented Modeling and Design for Database Applications. Upper Saddle River, N.J.:Prentice Hall.
[6] Suhami Napis, Kamarudin, Khairudin & A.Latif (2001), “Biodiversity Database for Malaysian Flora and Fauna: Update” University Putra Malaysia, 2001
[14] Bonan Li, Guoray Cai (2002). A General Object Oriented Spatial Temporal Data Model, Symposium on Geospatial Theory, Processing and application, Ottawa 2002.
[7] Joseph Maada Korsu Kandeh (2002), “geographic information system on invasive alien plant species”,
[15] Castagneri J. (1998). Temporal GIS Explores New Dimension in Time. GIS World, 11:9, 48-51.
[16] Claramunt and Thériault, M. (1995). Managing Time in GIS: An EventOriented Approach. Clifford, J., & Tuzhilin, A. (Eds), Recent Advances in Temporal Databases: Proceedings of the International Workshop on TemporalDatabases. (pp. 23-42). Berlin; New York: Springer. [17] Armstrong, M. P., (1988). Temporality in spatial databases. Proceedings: GIS/LIS'88, 2:880-889. [18] Tansel A. U., "Temporal Relational Data Model", IEEE Transactions on Knowledge and Data Engineering, Vol. 9, No. 3, May 1997. [19] Elmasri and Navathe (2004), “Fundamental of Database Systems”, fourth edition, 2004, Pearson Edition Inc. [20] Steiner, A. (1998). A generalization approach to temporal data models and their implementations, Swiss Federal Institute of Technology, Zurich, Zurich. [21] Faria, G., Mederios, C. B. and Nascimento, M. A. (1998). "An extensible framework for spatiotemporal database applications", TimeCenter. [22] James E. Rumbaugh, Michael R. Blaha, William J. Premerlani, Frederick Eddy, William E.Lorensen: Object-Oriented Modeling and Design Prentice-Hall 1991 [23] Jeremy L. Mennis, Donna J. Peuquet Diansheng Gua, (2000), “A semantic GIS Database Model for the Exploration of Spatio-temporal Environmental Data”, 4th international conf. on Integrating GIS and Environmental Modeling (GIS/EMA4)
