1. 1 Introduction to Databases Lecture One Relational and Object-Oriented Database Concepts Introduction to Database Design and Techniques

2. 2 Examples of Database Applications l Purchases from the supermarket l Purchases using your credit card l Booking a holiday at the travel agents l Using the local library l Taking out insurance l Using the Internet l Studying at university

3. 3 File-Based Systems l Collection of application programs that perform services for the end users (e.g. reports). l Each program defines and manages its own data.

4. 4 Limitations of File-Based Approach l Separation and isolation of data »Each program maintains its own set of data. »Users of one program may be unaware of potentially useful data held by other programs. l Duplication of data »Same data is held by different programs. »Wasted space and potentially different values and/or different formats for the same item.

5. 5 Database Approach l Arose because: »Definition of data was embedded in application programs, rather than being stored separately and independently. »No control over access and manipulation of data beyond that imposed by application programs. l Result: »the database and Database Management System (DBMS).

6. 6 Database l Shared collection of logically related data (and a description of this data), designed to meet the information needs of an organization. l System catalog (metadata) provides description of data to enable program–data independence. l Logically related data comprises entities, attributes, and relationships of an organization’s information.

7. 7 Database Approach l Controlled access to database may include: »A security system. »An integrity system. »A concurrency control system. »A recovery control system. »A user-accessible catalog. l A view mechanism. »Provides users with only the data they want or need to use.

8. 8 Views l Allows each user to have his or her own view of the database. l A view is essentially some subset of the database.

9. 9 Views l Benefits include: »Reduce complexity; »Provide a level of security; »Provide a mechanism to customize the appearance of the database; »Present a consistent, unchanging picture of the structure of the database, even if the underlying database is changed.

10. 10 History of Database Systems l First-generation »Hierarchical and Network l Second generation »Relational l Third generation »Object-Oriented

11. 11 The DBMS Marketplace l Relational DBMS companies – Oracle, Sybase – are among the largest software companies in the world. l IBM offers its relational DB2 system. With IMS, a non-relational system, IBM is by some accounts the largest DBMS vendor in the world. l Microsoft offers SQL-Server, plus Microsoft Access for the cheap DBMS on the desktop

12. 12 Terminology l Database: persistent collection of data l Database Management System (DBMS): software that controls access to the database l Database Administrator (DBA): person who controls database l Data Model: general structure of the data in the database l Data Language: commands used to define the data model and give users access to the database

13. 13 Utility of Databases l Data has value independent of use l Organized approach to data management l Eliminate redundancy in data l Share data l Archive data l Security of data l Integrity of data

14. 14 DB Terms and Techniques l Database access is a key feature of current enterprise computing l Relational DB: tables l To link/merge tables and extract/write information: Structured Query Language (SQL) – language of all modern databases (but many dialects) l SQL is transparent; operates with statements like SELECT, INSERT, DELETE, etc. l SQL provides its result sets in table format

15. 15 DB and the Internet l One vs. multiple user access l Internet browsers make it easy to access database programs (compared with traditional client/server programs)

16. 16 Relational Database Model l Database »Database is a collection of tables (relations) »Data are stored in tables l Tables »Each table has a name »Each table has a set of columns (fields) and rows of data (records) »Each table has a fixed number of columns »Each table has an arbitrary number of rows l Based on set theory l SQL (Structured Query Language) »DBMS independent language

17. 17 Database Columns (Fields) l Columns »Each column has a name »Columns are accessed by name »No standard column ordering »Data in a column belongs to a particular domain –Columns are the “attributes” of the dataset –Each value in a column is from the same domain –Each value in a column is of the same data type

18. 18 Database Rows (Records) l Rows »Each row entry is either a simple value or empty ("null") »Rows are sets of values for the columns (attribute values) »Primary key: a set of columns that uniquely identifies each row »Each row must be unique given the primary key (no duplicates) »Rows are referenced by the primary key »Row order cannot be determined by the user »Does not make sense to say “the fourth row” like it does in a “paper” table or spreadsheet

19. 19 Data Types l Each row value is an instance of a primitive data type »Integer »Real (e.g., number, currency »Character (e.g., text, hyperlink, yes/no) »Date/Time l No complex types in standard DBMS (matrix, drawing) »MS Access will allow drawings and some objects »Object oriented databases may allow objects and structures l Non existent value is “null”

20. 20 Database Design l Database design deals with how to design a database l Importance of Good Design »Poor design results in unwanted data redundancy »Poor design generates errors leading to bad decisions l Practical Approach »Focus on principles and concepts of database design »Importance of logical design

21. 21 Database Design Goals l Create a balanced design which is good for all users l Based on a set of assumptions about the world being modeled l Determine the data to be stored l Determine the relations among the data l Determine the operations to be performed l Specify the structure of the tables

22. 22 Database Design Process 1. Identify all the objects, entities, and attributes 2. Identify all the dependencies, draw a dependency diagram 3. Design tables to represent the data items and dependencies 4. Verify the design 5. Implement the database 6. Design the queries 7. Test and revise

23. 23 Identify All Objects and Entities l Determine the objects of your Database l For each object, describe each entity to be stored »example: better to store first name and last name separately l Determine the data type for each item »text, currency, date, etc. l Determine the range of allowable values for each item »non-negative? »greater than zero? »decimal points? »any of the 50 state abbreviations »zip code between 00000 and 99999 »phone number

24. 24 Turn Data Items into Attributes l Each attribute should have: »a meaningful name »a description of what the attribute means or what kind of data make up the attribute »a domain –the data type of the attribute –the range or a list of allowable values of the attribute

25. 25 Identify All the Dependencies l Assume a set of relationships between data items »a model of the world »may have to make assumptions »these assumptions should be listed clearly l Turn these relationships into dependencies »single-valued : there is one and only one value of ‘x’ for every value of ‘y’ –a person Y receives a grade X for a course in a semester –a person Y has a birth date X »multi-valued : there are zero (or one) or more values of ‘x’ for every value of ‘y’ –a student Y enrolls in one or more classes (X) each semester –a person Y has zero or more sisters l Draw a dependency diagram

26. 26 Single-Valued (One-to-One) Dependencies l Draw a single-headed arrow for single-valued dependencies PERSON BIRTHDATE STUDENT FINAL COURSE GRADE a person has one and only one birth date a student has one and only one final grade for a course

27. 27 Multi-Valued (One-to-Many) Dependencies l Draw a double-headed arrow between multi-valued dependencies STUDENT CLASSES PERSON SISTERS a student can enroll in one or more classes a person has zero or more sisters

28. 28 Independent vs. Dependent Attributes l Some attributes are independent »E.g., in a business – client relationship, your client’s phone number does not depend on when you are scheduled to meet him »your client still exists whether or not you have an appointment with him l Some attributes are dependent »the length of a side rails on a bridge is dependent on the structure of the bridge »the side rails of a bridge would not exist if the bridge itself was not there

29. 29 Dependent vs. Independent Attribute Representation l Start a new bubble around an independent attribute »properties of that attribute are attached to the new bubble »properties that are dependent on other attributes are attached to the old bubble »Each appointment is with one or more clients. Each appointment with one or more clients has a time. Each client has a single phone number. CLIENT PHONE NUMBER APPOINTMENT TIME

30. 30 Design the Tables l Draw a dependency diagram l Each dependency statement is a part of the diagram l Each statement is a single path through the diagram l Tables are formed by traversing the dependency diagram

31. 31 Traversing the Dependency Diagram l Choose an attribute at the end of a path l Follow the chain of arrows upwards »each multi-valued dependency on the path becomes a primary key for the table »combine all single-valued attributes at first level up into a single table »all attributes on the path should be included in the table »stop when you reach a bubble that has no arrows coming into it »each path becomes a separate table l Mark off your traversed path l Repeat until all paths have been traversed

32. 32 Verify the design l Inspect your tables »are all of the data included? l Do you have too many tables? too few? l If your design does not appear correct »go back to step 1 »you must repeat all steps of process in order »do not try to “rearrange” dependency diagram to give you the tables you think you should have

33. 33 Common Database Design Mistakes l Assuming the order of rows and columns is known »this is not a spreadsheet! »do not assume sorted order unless you explicitly sort l Guessing the design, not following the process l Storing what you can compute (when the value will change) – transitive dependency »e.g., do not store age if you are already storing birth date l Represent multi-valued dependencies in fixed size sets »if you know that there are exactly X number of something, create X single- valued dependencies, otherwise use multi-valued dependency l Adding a key when a unique value exists »adding an ID number for each person when you are already storing their social security number

34. 34 Results l If you follow the process correctly »you will not have redundant data »you will not lose unrelated data when you delete values l Databases with these characteristics are called 3NF (Third Normal Form) databases l Normalization to be discussed later in the course

35. 35 Goal: Build a Business Application Tools: Database Design SQL (queries) Programming Design SQL Program Design SQL Program Best: Spend your time on design and SQL. Worst: Compensate for poor design and limited SQL with programming.

36. 36 Application Development Feasibility Identify scope, costs, and schedule Implementation Transfer data, install, train, review Development Create forms, reports, and help; test Design Define tables, relationships, forms, reports Analysis Gather information from users tasks time

37. 37 DBMS Features/Components l Database engine »Storage »Retrieval »Update l Query Processor l Data dictionary l Utilities l Security l Report writer l Forms generator (input screens) l Application generator l Communications l Programming Interface

38. 38 DBMS Engine, Security, Utilities Data Tables Database Engine Product ItemIDDescription 887Dog food 946Cat food Order OrderIDODate 98743-3-97 98883-9-97 Customer CustomerID Name 1195Jones 2355Rojas Product ItemIDInteger, Unique DescriptionText, 100 char Customer CustomerIDInteger, Unique NameText, 50 char Security User Identification Access Rights Utilities Concurrency and Lock Manager Backup and Recovery Administration Data Dictionary

39. 39 Database Tables (MS Access)

40. 40 Database Tables (Oracle)

41. 41 DBMS Report Writer All Data Database Engine Data Dictionary Query Processor Report Writer Report Format and Query

42. 42 Report Writer (Oracle)

43. 43 DBMS Input Forms All Data Database Engine Data Dictionary Query Processor Form Builder Input Form Design

44. 44 DBMS Components All Data Database Engine Data Dictionary Security Query Processor Form Builder Report Writer Communication Network 3GL Connector Program Application Generator

45. 45 Relational Database Customer(CustomerID, Name, … Order(OrderID, CustomerID, OrderDate, … ItemsOrdered(OrderID, ItemID, Quantity, … Items(ItemID, Description, Price, …

46. 46 Object-Oriented DBMS Customer CustomerID Name … Add Customer Drop Customer Change Address Order OrderID CustomerID … NewOrder DeleteOrder … OrderItem OrderID ItemID … OrderItem DropOrderItem … Item ItemID Description … New Item Sell Item Buy Item … Government Customer ContactName ContactPhone Discount, … NewContact Commercial Customer ContactName ContactPhone … NewContact

47. 47 Objects l Object Definition »Object Name »Properties »Methods l Inheritance »Combine into one table. »Use multiple tables and link by primary key. –More efficient. –Need to add rows to many tables. Customer CustomerID Address Phone AddCustomer DropCustomer Class name Properties Methods Commercial Contact VolumeDiscount ComputeDiscount Government Contact BalanceDue BillLateFees AddCustomer Inheritance

48. 48 Objects in a Relational Database CustomerID Address Phone Customer CustomerID Contact VolumeDiscount CommercialCustomer CustomerID Contact BalanceDue GovernmentCustomer l Separate inherited classes. l Link by primary key. l Adding a new customer requires new rows in each table.

49. 49 OO Difficulties: Methods Database Object Customer Method: Add New Customer Application Customer Name Address Phone Personal Computer Unix Server IBM Server Program code Database Object How can a method run on different computers? Different processors use different code. Possibility: Java

50. 50 End of Lecture