Fundamentos y modelos de bases de datos by Mario G. Piattini así como otro al diseño lógico de bases de datos en el modelo relacional. De Miguel, A., Piattini, M., & Marcos, E. (). Diseno de bases de datos relacionales. Rama. Elmasri, R., & Navathe, S.B. (). Fundamentals of database. (), De Miguel, A., Piattini, M. and Marcos, E. Diseño de Bases de Datos Relacionales. Rama Ed, Dunn and Orlowska (), Dunn, L. and Orlowska.
|Published (Last):||17 June 2010|
|PDF File Size:||15.83 Mb|
|ePub File Size:||17.79 Mb|
|Price:||Free* [*Free Regsitration Required]|
Skip to main content. Log In Sign Up. Design method for a Historical Data Warehouse, explicit valid time in multidimensional models. Pons1 Recibido 19 de noviembre deaceptado 6 de enero de Received: El objetivo del modelo propuesto es resolver las limitaciones temporales de las estructuras multidimensionales tradicionales.
The objective of this model is to solve the temporal limitations of the traditional multidimensional structures. Although the Temporal DW considers, besides the temporal dimension, other aspects related to time, this model only takes into account the changes that occur in the DW schema, both in dimensions and in hierarchies.
Fundamentos y modelos de bases de datos
Therefore, considering the need for registering values that allow to evaluate trends, variations, maximum and minimum values, a problem to be solved is how to shape the values of entities, attributes or relationships that may vary in time, in the design of the structure.
The fact that the needed data was stored is already known, but the temporal search mechanisms would be complex. Design diweo, model driven software development, data warehouse, historical data base, valid time. A automatic execution of SQL statements . The relationships associated with these transactions research was conducted with students from the have varied through time.
As a first conclusion, from this model considers only the changes that occur in the assessment performance, we consider that the the Data Warehouse schema, both in dimensions and submitted proposal constitutes a valid alternative in hierarchies. Therefore, a problem to be solved in in the design of multidimensional structures, in the this type of multidimensional structure, considering proposal of the design baxes, and in the storage the need for registering values that allow to evaluate structure and in the graphical query interface.
For trends, variations, maximum and minimum values, more details see . A solution to this work; following this, we describe the conclusion problem is proposed by Model Driven Software and, doseo, the future works. This new paradigm aims to improve Data Warehouse, Temporal Data Base and Model productivity and software quality generated by Driven Software Development, which allow to reducing the semantic leap between the problem support the proposal.
Data Warehouse The proposed design method consists in successive The companies use the operational data accumulated automatic and semiautomatic transformations of over the years and stored in structures ad hoc called models, that begin with an ER model and that finally Data Warehouse to help understand and manage their allow to obtain a temporal multidimensional model activities.
While an Operational Data Base maintains expressed as a set of tables in the relational model. The multidimensional model cannot be changed . In this model the information is structured in facts and dimensions; The different types of Data Bases are connected, in a fact is a topic of interest for the company, it is its definition, disdo the concepts presented above.
A described by means of fact attributes, and these are Data Base that models only the valid time is called contained in cells or points in the data cube. A data Historical Data Base, a Data Base that models only cube is a multidimensional representation of data the transaction time, is called RollBack Data Base that can be seen from different points of view; a and the one that models both the valid time and data cube is formed by dimensions, which determine transaction time, is called BiTemporal Data Base .
This approach represents a new becomes effective when it is established and it is paradigm where models of the system, at different considered valid until a new update modifies it; levels of abstraction, are used to guide the entire therefore there is no distinction between the time development process. Models are implementation- of registration of that information in the Data Base independent and they are automatically transformed and the period during which the specific values of to executable code.
Thus, the Data Base represents This new paradigm aims to improve productivity only the current state and not the history of the and to generate software quality by reducing the reality facts that it was modeling. Furthermore, a semantic leap between the problem domain and Temporal Data Base supports some aspect of time, the solution .
The key underlying idea is without taking into account user-defined time . We can divide, briefly, the Model Driven time elapsed betw[een two instants; a timestamp Software Development process into three phases; is a temporal mark associated with an object or in the first one, a Platform Independent Model attribute of the Data Base. The valid time of a fact PIM is built, this is a high-level model of the is the time when the fact is true in the modeled system being developed independently of any reality.
A fact may have any number of instants and technology; then, the previous model is converted time intervals associated, these single instants and into one or more Platform Specific Models PSMintervals are important special cases.
Valid times these models are at a lower level than the PIM and are usually supplied by the user . In practice, describe the system in accordance with a particular the valid time is the most important concept since implementation technology; and finally, the last one it models the veracity of the facts recorded in the generates an Implementation Model IMthis is, a universe of discourse, which is the main objective source code from each PSM.
The division between of the information systems . Model Driven Software Development also until logically deleted. The transaction time of a presents a Computation Independent Model CIM Data Base fact is the time when the fact is present that describes the system within its environment and in the Data Base and may be retrieved.
Transaction shows what is expected from it without showing time values cannot be later modified. Also, as it is any details of how it will be built.
Moreover they may have descriptive model.
This model includes, besides the main attributes non-dimension attributes. The hierarchy analytical fact, temporal structures related to the levels e-tempi, a-tempi, r-tempi represent entities, levels of dimensional hierarchies that allow to record attributes, and temporal relationships respectively data and retrieve information that varies in time. The attributes IT temporal . Initial Time and FT Final Time will determine both the initial and final instances of the temporal To express the valid time, we will use the notation interval considered respectively.
For example, pitatini following are typical queries: At what intervals was a particular client active?
To use Loot.co.za, you must have cookies enabled in your browser.
On what dates and where did a particular client move? Which was the location of a particular client on a certain date? These are specific queries of a Historical Data Base and could not be performed in a common Data Warehouse. These queries and the ones of decision-making can be performed in our model using a Graphical Query interface . The transformation method will begin with an ER model and, by means of successive transformations, will allow to obtain a set of tables in a Relational Model expressed in SQL sentences.
Below, we will detail how the informal transformation process is; we will show, relacionalds by step, each of the transformations in detail. To explain the method, we will use and develop an example that shows the different transformations. The new entity will carry throughout the country. Design method for a Historical Data Warehouse, explicit valid time in pattini models temporal will remain as non-temporal to allow, in the subsequent transformation processes, to be source of a possible level of grouping in the dimension.
All the entities and non-temporal relationships, in the source model, will remain without modifications in the target model. All the non-temporal attributes both entities and relationshipsin the source model, will stay in the target model, unlike the attributes transformed in temporal attributes.
The fact if represented by a relationship in the source model, will become an entity related to the two entities that are part of the relationship in the target model. The fact will have the attributes of entity associated with the entity that owns the attribute.
The the relationship as a unique identifier. The As an intermediate step in the design of the Historical unique identifier compound of the temporal entity Data Warehouse, a modified Temporal Attribute will consist in the attribute called IT plus the identifier Graph  will be built, from the Temporal Data of the regular entity. The multiplicity of the role of the viseo Warehouse. Data Model, will be identified, first, as a concept of primary interest relacioanles the decision-making process A temporal relationship, in ve source model, will in the Historical Data Warehouse.
The fact will transform into a temporal entity weak linked to correspond to events that occur dynamically in the one of the entities that form the relationship in the reality and will be represented, in the adapted model, target model.
The name of the new entity will be by an entity. Pizttini new temporal entity will We will detail the characteristics of the graph: For each non-temporal have in the ER source model. The unique identifier vertex v, its corresponding associated attributes will compound of the temporal entity will consist in the determine functionally to all basez that apply attribute called IT plus the identifier attribute of the to the descendants of v.
The vertexes identifier attributes in the fact. All the vertexes linked to the root in the Temporal Attribute Graph, which are neither identifiers nor Construction of a Temporal Attribute Graph attributes that denote temporal aspects the latter will Given an identifier E that indicates a set of be transformed in leaves of the temporal hierarchy attributes that identify the entity E in the source level will correspond to measures. A fact may lack model, the Temporal Attribute Graph will be able measures.
The leaf levels in the hierarchy should be chosen in the Temporal Attribute Graph between vertexes Function translate E: Vertex children of the se. From Temporal Attribute Graph to The fact should have an associated vertex that denotes Multidimensional Model temporal aspects.
Unlike the choice of the fact, fact identifier. The multiplicity of the role fact will that will directly derive from the root of the graph. Datks concepts of root, leaf, child root, in the Temporal Attribute Graph, which are and parent were taken from .
The name of the parent level will be the transformation is as follows: Design method for a Historical Data Warehouse, explicit valid time in multidimensional models be the same as the identifier of the node, all the the primary key will be composed of a group of the vertexes linked to that level, which are identifiers, identifier attributes of each level; in addition, each will also be linked to parent levels.
Piathini vertexes table leaf will have a foreign key that will refer to that are not identifiers will be level attributes. All the each one of the tables that form the hierarchy levels.
In all The levels in the hierarchies in the Temporal the levels of the hierarchy, the multiplicity of role Multidimensional Model will become tables in the of the parent will be 1, 1the multiplicity of child Relational Model; the level attributes will be columns role will be 1, N.
If this becomes from a temporal will be the same as the identifier of the node. Besides, the table plus IT attribute.
If the temporal hierarchy is attribute that is not part of the temporal interval, derived from a temporal attribute, it will have the will make reference to the linked hierarchy level. All the vertexes linked the foreign key that will refer to this hierarchy table to the temporal nodes that are not identifiers, will plus the attribute IT.
If the temporal hierarchy level be temporal vertex attributes. Therefore, from the Temporal of the union of the primary key of the other related Multidimensional Model we will obtain a set of hierarchy table in addition, it will be the foreign relational data structures by applying the following key that will refer to this hierarchy table plus the transformation rules.
The transformation process attribute IT. Below, the Pseudotables the leaf levels; the attributes forming the primary from the Relational Model transformed in the key will also be foreign keys that will refer to each previous section are shown: In addition, we designed specific metamodels for the models used in the process: Therefore, a metamodel use the metamodel Temporal Attribute Graph.
We of a language graphical or textual is a precise do not use UML or profiles for the design of PIM, definition of its elements by using concepts and as we believe that the ER model is more expressive rules of a certain meta-language needed to describe for data modeling. All the classes, except those of the common and abstract language to define modeling Temporal Attribute Graph, inherit the name attribute languages and the form to access and exchange of a superclass Named, not shown in the graphs.
This process will allow the models. Therefore, in the Model Driven Software to transform the entities, the attributes and the Development, models are defined by metamodels.
Design method for a Historical Data Warehouse, explicit valid time in multidimensional models String Vertex Model to the Relational Model. Boloean in the Model Driven Software Development. String the mappings between the source and target model dataType: