Sunteți pe pagina 1din 6

Describing Data

http://www.nos.org/htm/sad3.htm

NOS : Certificate in Computer Applications | Home | Table of Contents | Up: Online Course Material Lesson 31 Describing Data 31.1 Introduction An important aspect of system analysis is analysis of system data. For describing the data in its meaningful form data dictionary is developed. Data dictionary also contains the different types of data as well as their meanings. It serves as a powerful tool of system analysis and designing. 31.2 Objectives After completing this lesson the you should be able to : describe data into its meaningful form understand different types of data prepare data dictionary for given system normalize the relations up to third normal form 31.3 Data definition Most people use the terms data & information interchangeably. But data & information are not the same. 31.4 Data and Information Data are raw facts and figures in isolation. These isolated facts and figures convey meanings but generally are not useful by them. For example, A, H, ,34, 26, 56.68,VISHAL, SCHOOL etc. Data are processed to form information. Data are processed in various meaningful ways to form information. For example, Prasad, whose enrolment no. is 270173002, has been certified in Sr.Secondary examination of National Open School is an information as it conveys some meaning. 31.5 Types of data Data are of various types, among them major ones are1. Character type 2. Numeric type 3. Alphanumeric type Character type are those data which is consisting of only the alphabets i.e. from A to Z For example Name, Age, Thing etc. Numeric data types are those data types which is consisting only of numbers or integers & the combination of digits from 0 to9. Example 254, 67, 2, 7, etc. Alphanumeric are those types of data which is constructed by the combinations of characters &Numeric data types. Example- A79, DD56, 4G6 etc.
1 of 6 17-01-2012 18:25

Describing Data

http://www.nos.org/htm/sad3.htm

31.6 Data Dictionary A data dictionary defines each term encountered during the analysis and design of a new system. Data dictionary is the place where we keep the details of the contents of data flows, data stores & processes. Data dictionary is an analysis tool, that primarily records the information content of data. Without a data dictionary the development of large systems becomes difficult. The data dictionary is an effective solution to the problem of complicated nature. The main purpose of a data dictionary is to provide a source of reference in which the analyst , the user, the designer can look up & find out its content and any other relevant information. Examples of Data dictionary Student Record = Enrolment Number + Name + Address + Sex + Date of Birth + Subject +

31.7 Levels of Data Dictionary We can define the data dictionary in three different levels. Data Elements Data Structure Data flows and Data Stores (a) Data Elements Data elements are pieces of data, which need not be broken further. Data elements can describe files, data flows, or processes. Often a data element is self defining such as Student name, enrolment no. (b) Data structure Data structure comprises of data elements. It is defined as collection of data elements. For example let us consider the following "Student Information Record". Student Information Enrolment Number Student name First name Second name Last name Sex

2 of 6

17-01-2012 18:25

Describing Data

http://www.nos.org/htm/sad3.htm

Student Address Address 1 Address 2 Address 3 Pin code Subject Details Subject1 Subject2 Subject3 Subject4 Subject5 Subject6 Background details Family background Fathers education Mothers education Family Income Urban/rural House hold items TV (Y?N) Radio (Y?N) Cycle(Y?N) Here "STUDENT INFORMATION" is a data structure, made up of data elements student name, enrolment number. "Subject details" is a data structure made up of six data elements subject1 to subject6.

(c) Data flows & Data stores Data flows are paths along which data travels & Data stores are places where data is stored until needed. So we can say that Data flows are Data structures in motion & Data stores are Data structures at rest. Data description Hierarchy is illustrated in figure 31.1

3 of 6

17-01-2012 18:25

Describing Data

http://www.nos.org/htm/sad3.htm

Figure 31.1 : Data description Hierarchy A data dictionary as we have seen, contains Data elements, Data structures and Data flows and data stores along with this processes, external entities & glossary of user terms. Processes are defined with number of the tools, such as Decision trees &Decision tables, which have already been discussed in previous lesson. In number of applications the users have their own vocabulary & this can be confusing to the analysts and the programmers. In that case data dictionary is a convenient place to keep these glossary items for reference purpose. 31.8 Normalization Normalization is a process by which the contents of the data store is simplified by removing redundant data elements. It is done by removing repeating groups and reorganizing the contents of data store. Let us consider the following data structure: Name Personnel-No. Address Salary-Slip-History* Date-of-change Job-title Salary Now, how can we get rid of the repeating group, Salary-Slip-History? The only way is, to split the main structure into two structures, both of which are simpler. After doing this normalization process, we end up with a structure containing name and address (occurring only once for each employee) and a structure containing change of title or salary (several for each employee). Each structure must contain Personnel-No., the only data element, which specifies each employee uniquely. In normalization the following vocabulary are used . In place of "data structure", relation is used. In place of "data element", the term domain is used. Each individual record is called a tuple. There are three types of normalized relations, called in increasing order of simplicity; first normal form, second normal form, and third normal form. (a) First Normal Form

4 of 6

17-01-2012 18:25

Describing Data

http://www.nos.org/htm/sad3.htm

Any normalized relation is automatically 1st normal form. Relations in first normal form may suffer from two kinds of complexity. i. If the primary key is concatenated (chained together), some of the non-key domains may depend on only part of the key, not the whole key. ii. Some of the non-key domains may be interrelated. Suppose we are trying to produce a normalized version of the data structure, "Orders", describing book purchases. We define an order than consists of the customer-name, order-date, ISBN for the book ordered, title, author, quantity of title that has been ordered and total cost of the order for a given book. We can create a normalized relation: Book-order (Customer-Name, Order-Date, ISBN, Author, Title, Quantity, Price, Order-Total) Here the underline shows that we have chosen the concatenated key: Customer Name/Order-Date/ISBN to uniquely identify each order. As we have noted, this relation is in 1st NF, by virtue of the fact that, it contains no repeating groups. The first complexity we would like to remove is the fact the several of the non-key, domains (Title, Author and Price) can be identified with only part of the key-the ISBN. In other words, if you are given the ISBN, you can know the Title, Author and Price not taking into account the "Customer-Name" or the "Order-Date". This is a different situation from that "Quantity". To know the quantity in any particular order, you have to know all of the three domains, which are concatenated, to make the key. In the vocabulary of normalization, "Quantity" is fully functionally dependent on the whole concatenated key. On the other hand, "Title" is not fully functionally dependent. Since, you only need to know part of the key (ISBN) to find the "Title", a domain is fully functionally dependent, but not fully so, if the value of the domains can be determined from only part of the key. This concept allows us to define second normal form. (b) Second Normal Form A normalized relation is in 2nd normal form if all the non-key domains are fully functionally dependent on the primary key "Book Order". To get "Book Order" into 2nd formal form, we must get rid of the partial functional dependence. This can be done by taking out the domains which describe the book and putting them in a separate relation:Order (Customer-Name, Order-Date, ISBN, Quantity Order-Total) Book (ISBN), Title, Author, Price) Order is now in 2nd normal form : each of the non-key domains (Quantity, Order-Total) can only be specified by knowing the fully concatenated key, i.e., all the non-key domains are fully functionally dependent on the primary key. We can simplify the situation even further, because "Order" has a complexity, still working in it. "Quantity" and "Order-total" are not mutually independent. For only given "Price" the "Quantity" determines the "Order-total". Therefore, "Order-Total" is functionally dependent on quantity. (c) Third Normal Form A normalized relation is in 3rd normal form if : a. All the non-key domains are "fully" functionally dependent on the primary key. b. No non-key domain is functionally dependent on any other non-key domains. So to transform a 2nd normal form relation into a 3rd normal form, we examine each of the non-key domains to see that they are independent of the other non-key domains and remove any such mutual independence. In the

5 of 6

17-01-2012 18:25

Describing Data

http://www.nos.org/htm/sad3.htm

case of "Order", we see that "Order-Total" is, in fact, a redundant data element, because it can be completed. So, we can remove it altogether and express each "Order" by means of two relations, both in 3rd normal form. Book-Order (Customer-Name, Order-Date, ISBN, Quantity) Book (ISBN, Title, Author, Price) Top 31.8 What You Have Learnt In this lesson data and types of data are explained. Data dictionary as analysis tool has been discussed in detail. The three levels of data dictionary are discussed. Normalization as techniques for simplifying the data structure is also explained. The three different normalized relations are explained. 31.9 Terminal Question 1. 2. 3. 4. 5. Define data and explain different types of data. What is data dictionary. Explain different levels of data dictionary. What is normalization. Explain different forms of normalization. Top

6 of 6

17-01-2012 18:25

S-ar putea să vă placă și