Information Systems and Networks

Transcription

1 Information Systems and Networks by Samuel Rota Bulò Department of Management Università Ca' Foscari Venezia

2 Lesson 5 Databased and Information Management

3 Case study: RR Donnelley giant commercial printing and service company providing printing services, forms and labels, direct mail, and other services The company s recent expansion has been fueled by a series of acquisitions commercial printer Moore Wallace in 2005 printing and supply chain management company Banta in January 2007 RR Donnelley s revenue has jumped from $2.4 billion in 2003 to over $9.8 billion today

4 Case study: RR Donnelley all that growth created information management challenges Each acquired company had its own systems and its own set of customer, vendor, and product data The data were often inconsistent, duplicated, or incomplete Donnelley had to use time-consuming manual processes to reconcile the data stored in multiple systems in order to get a clear enterprise-wide picture of each of its customers, since they might be doing business with several different units of the company. These conditions heightened inefficiencies and costs Donnelley still needed a clear single set of data that was accurate and consistent for the entire enterprise

5 Case study: RR Donnelley RR Donnelley turned to master data management (MDM) MDM seeks to ensure that an organization does not use multiple versions of the same piece of data in different parts of its operations by merging disparate records into a single authenticated master file Once the master file is in place, employees and applications access a single consolidated view of the company s data It is especially useful for companies such as Donnelley that have data integration problems as a result of mergers and acquisitions

6 Case study: RR Donnelley Implementing MDM is a multi-step process that includes: business process analysis data cleansing data consolidation and reconciliation data migration into a master file of all the company s data Companies must identify what group in the company owns each piece of data and is responsible for resolving inconsistent definitions of data and other discrepancies Donnelley launched its MDM program in late 2005 and began creating a single set of identifiers for its customer and vendor data The company opted for a registry model using Purisma s Data Hub customer data continue to reside in the system where they originate but are registered in a master hub and cross-referenced so applications can find the data

7 Case study: RR Donnelley Nearly a year later, Donnelley brought up its Customer Master Data Store, which integrates the data from numerous systems from Donnelley acquisitions Data that are outdated, incomplete, or incorrectly formatted are corrected or eliminated A registry points to where the source data are stored By having a single consistent enterprise-wide set of data with common definitions and standards, management is able to easily find out what kind of business and how much business it has with a particular customer to identify top customers and sales opportunities when Donelley acquires a company, it can quickly see a list of overlapping customers

8 Case study: RR Donnelley

9 Organizing data An effective IS provides users with accurate, timely, and relevant information timely: it is available to decision makers when it is needed relevant: it is useful and appropriate for the types of work and decisions that require it many businesses don't have timely, accurate or relevant information because the data in their IS have been poorly organized an mantained

10 File organization terms and concepts

11 Problems of the traditional file environment

12 Problems of the traditional file environment Data redundancy: presence of duplicate data in multiple data files occurs when different groups in an organization independently collect the same piece of data and store it independently of each other Data inconsistency: same logical piece of data that is stored in different locations with different values data updated in some systems but not in others use of different coding systems for attributes

13 Problems of the traditional file environment Program-data dependence: coupling of data stored in files and the specific programs required to update and maintain those files changes in programs require changes to the data such changes could cost millions of dollars to implement properly Lack of flexibility: a traditional filesystem can deliver routine scheduled reports after extensive programming efforts it cannot deliver ad hoc reports or respond to unanticipated information requirements in a timely fashion

14 Problems of the traditional file environment Poor security: access and dissemination of information may be out of control management has no way of knowing who is accessing or even making changes to the organization's data Lack of data sharing and availability: pieces of information in different files and different parts of the organization cannot be related to one another virtually impossible for information to be shared or accessed in a timely manner information cannot flow freely across different functional areas of different parts of the organization finding inconsistent pieces of information discourages users form trusting the accuracy of their data

15 Database: The database approach to data management collection of data organized to serve many applications efficiently by centralizing the data and controlling redundant data data appears as being stored in one location Database Management System (DBMS): software that permits an organization to centralize data, manage them efficiently, and provide access to stored data by application programs acts as an interface between application programs and the physical data files separates the logical and the physical views of the data the logical view presents the data as they would be perceived by end users or business specialists the physical view shows how data are actually organized and structured on physical storage media

16 The database approach to data management

17 The database approach to data management a DBMS reduces redundancy by minimizing isolated files in which the same data are repeated helps control redundancy by ensuring that every occurrence has the same value uncouples programs and data, enabling data to stand on their own access and availability of information will be increased program development and maintenance costs reduced programmers can perform ad hoc queries of data in the database enables the organization to centralize their data management, their use and security

18 Relational DBMS The most popular type of DBMS is the relational DBMS Relational databases represent data as 2-dimensional tables (relations) Tables contain data on an entity and their attributes

19 Main DBMS Commercial vendors DB2 (IBM) Caché (InterSystems) FileMaker Pro (FileMaker Inc) Microsoft SQL Server (Microsoft) Microsoft Access (Microsoft) Oracle (Oracle Corporation) SQL Anywhere Studio (SyBase) SyBase SQL Server (SyBase) Open source or free software Firebird SQL (The FirebirdSQL Foundation) HSQL Database Engine MySQL PicoSQL PostgresSQL SQLite ZODB

20 The Entity-Relationship (ER) model graphical model that describes the organization of data that can be then implemented with a DBMS it is composed by three elements: entities relationships attributes

21 Entities An entity may be defined as a thing which is recognized as being capable of an independent existence and which can be uniquely identified it is an abstraction from the complexities of a domain entities can be thought of as nouns e.g., suppliers, parts, customers,... Student

22 Relationship A relationship captures how entities are related to one another. relationships can be thought of as verbs, linking two or more nouns Person owns Car

23 Attributes Attributed provide entities with properties e.g., the attributes of an entity car can be: producer, model, power, price, n. of cylinders attributes have: type: string, number, date dimension: length of the representation Person owns Car - Surname - Name - Title - Tax code - Producer - Model - Power - n. of cylinders - price

24 Ont-to-one relationships every element of the first entity has one and only one corresponding element of the second entity Employee 1 1 uses Workstation

25 One-to-many relationship every element of the first entity can have many corresponding elements of the second entity, but an element of the second entity has only one corresponding element of the first one Person 1 * owns Car

26 Many-to-many relationship every element of the first entity can have many corresponding elements of the second entity and vice versa Student * * follows Course

27 Relational DB. Tables A database consists of a set of tables Tables commonly correspond to entities Each table consists of rows (records), each encoding an element of the entity it represents Each column encodes an attribute of the entity field attribute record

28 Relational DB. Tables fields of the same attribute (column) come from the same domain (have the same type) rows are regarded as different elements, even if they have fields with the same values attributes have different names order of rows and columns is irrelevant a table can have one or more attributes that form a key for the table the key must identify uniquely each record of the table keys are used to establish relations among tables

29 Relational DBMS

30 Operations of a Relational DBMS database tables can be combined easily to deliver data required by users basic operations are used to develop useful sets of data: select: filters rows according to some given criteria join: links multiple tables project: restricts the set of attributes grouping: groups rows according to some given criteria

31 Operations of a Relational DBMS

32 Object-oriented DBMS (OODBMS) another type of DBMS, different from the relational one an object-oriented DBMS stores data and procedures that can act on those data as objects that can be automatically stored and shared they a relatively slow compared to the relational counterparts for processing large numbers of transactions hybrid systems exist that provide both relational and object-oriented capabilities

33 Databases in the cloud cloud computing providers offer database management services, but typically they have less functionality primary customers are Web-focused start-ups or small to medium-sized businesses looking database capabilities at lower price than a standard relational DBMS

34 Capabilities of DBMS DBMS provide capabilities and tools for organizing, managing, and accessing and manipulating the data in the database the most important are: data definition language allows to specify the structure of the content of the DB data dictionary automated or manual file that stores definitions of data elements and their characteristics data manipulation language used to add, change, delete, retrieve the data in the DB

35 Database-oriented languages Languages designed to manage a database should provide the following functionalities: Data Control Language (DDL) allow to specify database security and access control Data Definition Language (DDL) allow to specify the logical schema of the DB (tables structure) Data Manipulation Language (DML) allow to read, insert, modify and delete the database data Query language allows to perform complex queries The most famous is SQL (Structured Query Language)

36 Query results Results of a query expressed by means of a query language is a table

37 Structured Query Language (SQL) Initially every DBMS adopted its own language The need of standards to help integration and interoperability of applications forced the main relational systems to adopt the SQL language SQL is not a mere query language but allows also to manipulate data, as well as define the structure of the tables in the database

38 Structured Query Language (SQL) SELECT FROM WHERE C1,...,Cn T1,...,Tm P Generates a new table showing fields C1,...,Cn extracted from the tables T1,...,Tm and only the rows that satisfy property P. SELECT FROM WHERE Name, Surname Students BirthPlace = 'Venice'

39 SQL: data insertion/deletion INSERT INTO Students VALUES (12345,'Verdi','Paolo',' ','Pisa'), (23456,'Rossi','Marco',' ','Roma') Inserts new records in the table Students DELETE FROM WHERE Students id_number='12345' Deletes a record from the table Students having id_number 12345

40 SQL: table creation/deletion CREATE TABLE Students( id_number integer, surname varchar(50), name varchar(50), birthday date, birthplace varchar(50) ) Creates a table Student in the DB DROP TABLE Students Deletes the table Student

41 Databases to improve business performance and decision making Businesses use their DB to keep track of basic transactions, such as paying suppliers, processing orders, keeping track of customers, and paying employees they need databases also to provide information that will help the company run the business more efficiently and help managers and employees make better decisions e.g., find the most profitable customer, or the most popular product

42 Data warehouse A data warehouse is a database that stores current and historical data of potential interest to decision makers throughout the company the data originate in many core operational transaction systems, such as systems for sales, customer accounts, and manufacturing, and may include data from Web sites transactions the data warehouse consolidates and standardizes information from different operational databases so that the information can be used across the enterprise for management analysis and decision making the data warehouse makes the data available for anyone to access as needed, but it cannot be altered

43 Data warehouse

44 Data Marts a data marts is a subset of a data warehouse in which a summarized or highly focused portion of the organization's data is placed in a separate database for a specific population of users e.g., marketing and sales data marts to deal with customer information data marts are constructed more rapidly and at lower cost than an enterprise-wide data warehouse

45 Tools for business intelligence once data is collected in data warehouses and in data marts, they are available for further analysis using tools for business intelligence business intelligence tools enable users to analyse data to see new patterns, relationships, and insights that are useful for guiding decision making principal tools: software for database querying and reporting tools for multidimensional data analysis tools for data mining

46 Tools for business intelligence: Online Analytical Processing (OLAP) suppose your company sells four products nuts, bolts, washers and screws in the East, West and Central regions of US you can easily answer questions like how many washers where sold during the past quarter by querying the sales database but when we want to know how many washers were sold in each of your sales region and compare actual results with projected sales you need online analytical processing tools

47 Tools for business intelligence: Online Analytical Processing (OLAP) OLAP supports multidimensional data analysis, enabling users to view the same data in different ways using multiple dimensions dimensions are represented by different aspects of information, e.g., product, pricing, cost, region, time of period OLAP enables users to obtain online answers to ad hoc questions in a fairly rapid amount of time, even when the data are stored in very large databases, such as sale figures for multiple years

48 Tools for business intelligence: Data mining with OLAP and query-oriented data analysis, users need to have a good idea about the information for which they are looking data mining is more discovery-driven it provides insights into corporate data that cannot be obtained with OLAP by finding hidden patterns and relationships in larger databases and inferring rules from them to predict future behaviour patterns and rules are used to guide decision making and forecast the effect of those decisions

49 Tools for business intelligence: Data mining types of information obtainable from data mining: associations: occurrences linked to a single event e.g., supermarket purchasing pattern: 65% of the time a cola drink is bought with pop-corn, but when there is promotion, 85% of the time sequences: events linked over time e.g., if a house is purchased, a new refrigerator is bought within two weeks 65% of the time, and an oven within one month, 45% of the time classification: recognizes patterns that describe the group to which an item belongs by examining existing items that have been classified and by inferring a set of rules e.g., discover the characteristics of customer that are likely to leave clustering: similar to classification but without prior knowledge of the groups. Clustering finds several groupings of within data e.g., divide customers into groups based on demographics and types of personal investments forecasting: uses a series of existing values to forecast what other value will be e.g., predict sale figures

50 Text mining business intelligence tools deal primarily with data that have been structured in databases and files unstructured data, most in the form of text files, is believed to account for over 80% of an organization's useful information e.g., s, memos, call center transcripts, survey responses, legal cases, patent descriptions, service reports text mining tools are available to help businesses analyse these data they extract key element from large unstructured data sets, discover patterns and relationships and summarize the information

51 Web mining the Web is another rich source of valuable information, some of which can now be mined for patterns, trends and insights into customer behaviour web mining refers to the discovery and analysis of useful patterns and information from the WWW Google Trends and Google Insights for Search services allow to track the popularity of words and phrases in Google search queries to learn what people are interested in and what they are interested in buying web mining looks for patterns in data through content mining, structure mining and usage mining web content mining is the process of extracting knowledge from the content of Web pages which may include text, image, audio and video data web structure mining extracts useful information from the links embedded in the Web documents web usage mining examines user interaction data recorded by a Web server whenever requests for a Web site's resources are received

52 Databases and the web many companies use the Web to make some of the information in their internal databases available to customers and business partners user access internet using a browser through which they request data to from the organization's database using HTML to communicate

53 Databases and the web the web server receiving the user's request translate them from HTML into SQL so that they can be processed by the DBMS working with the database in a client/server environment, the DBMS resides on a dedicated database server between the web server and the database server we find an application server hosting a software that handles all application operations including transaction processing and data access the application server takes requests from the Web server, runs the business logic to process transactions based on those requests and provides connectivity to the organization's back-end system or databases

54 Managing data resources: information policy an information policy specifies the organization's rules for sharing, disseminating, acquiring, standardizing, classifying and inventorying information it lays out specific procedures and accountabilities, identifying which users and organizational units can share information, where information can be distributed, and who is responsible for updating and maintaining information e.g., only selected members of the payroll and human resources department would have the right to change and view sensitive employee data

55 Managing data resources: information policy in small businesses, information policy is established by the owner or managers in large businesses, there is a data administration, who is responsible for the specific policies and procedures through which data can be managed as an organizational resource he develops the information policy, plans for data, oversees logical database design and data dictionary development, and monitoring how IS specialists and end-user groups use data data governance deals with the policies and processes for managing the availability, usability, integrity and security of the data employed in an enterprise, with special emphasis on promoting privacy, security, data quality, and compliance with government regulations a large corporate has also a database administration responsible for defining and organizing the structure and content of the database, and maintaining it.

56 Managing data resources: data quality having data that is inaccurate, untimely, or inconsistent with other sources of information leads to incorrect decisions, product recalls and financial losses 20% of US main and commercial package deliveries were returned because of incorrect names or addresses 25% of the critical data in large Fortune 1000 companies' databases is inaccurate or incomplete including bad product codes, descriptions, faulty inventory description, erroneous financial data, incorrect supplier information, and incorrect employee data if a database is properly designed and enterprise-wide data standards established, duplicated or inconsistent data elements should be minimal

57 Managing data resources: data quality most errors however stem from incorrect data input the incidence of such errors increases as companies move to the web suppliers and customers enter their data and directly update internal systems before new databases are in place, organizations need to identify and correct their faulty data and establish better routines for editing data once their database is in operation data quality audit is a structured survey of the data accuracy and level of completeness of the data in an IS data cleansing (data scrubbing) consists of activities for detecting and correcting data in a database that are incorrect, incomplete, improperly formatted, or redundant specialized data cleansing software exist to automatically survey data files, correct errors in the data, and integrate the data in a consistent company-wide format