1st Normal relation: if it does not have any multi-valued attribute, may not depend on the whole primary

key 2nd Normal relation: if each non-key

attribute depends on the whole Primary Key 3rd Normal relation: if no non-key attribute depends on another non-key attribute AI Grand Vision:
generalized human intelligence AI Narrow Definition: specialized computer programs, able to best humans in specific tasks Analytical CRM: Bring data
from multiple sources and consolidate them to identify profitable customers, find market segments, churn rates Analytical platforms: Combines relational
& non-relational tech optimized to process big data Associative Entity: created when we want to record data about a relationship, it must have a primary
key. represents a relationship between other entities and so relies on those foreign keys to maintain data integrity. All associative entity have foreign key
Attribute: property or characteristic of an entity Balanced scorecard: Measures four dimensions of firm performance(Financial, Business process,
Customer, learning and growth) Behavioral targeting: Tracking online behavior of individuals on thousands of Web sites Binary Relationship: 2 entities
Blockchain: Distributed database of transactions/on network/no central control/high security Bullwhip effect: As demand fluctuates inventory piles up
more and more as we move up the upstream supply chain due to uncertainties, inaccurate data and lack of trust CRM Failures: Reason include Lack of
senior management sponsorship, end-user incentives and understanding & preparation Candidate key:any column or set of columns that can uniquely
identify each row in a table, primary key is a chosen candidate key. Capability advantage: quality, new product, marketing(internal) Case-based
reasoning:从数据库找旧 case 来匹配并解决新 case Churn rate: Rate at which customers stop doing business with you Competitive advantage: any unique attribute or
capability that allows a company to outperform its rivals Concurrent Versions: use delay in some digital products to create multiple concurrent versions
Configuration tables: database tables that store settings, parameters, and rules used to customize the application's behavior Content provider: digital
content (WSJ.com) Contract purchasing: Involves long-term written agreements to purchase specified products, with agreed-upon terms and quality
Coordination: Unique business units, but need to share data Cross-functional modules: connect different departments and business processes and
enabling data sharing Cross-side network effects: Positive, as the number of buyers increases, value of the platform to sellers increases Customer touch
points: any interactions a customer has with a brand Dashboards:visual tools presenting user defined performance data Data Dictionary: Meta Data
containing data definitions Data Integration: combining data from multiple files Data Mart:Subset of data warehouses that is highly focused and isolated
for a specific population of users Data Warehouse: Benefit:Combines data from multiple sources for better decision making Data mining: Finds hidden
patterns and relationships in databases and infers rules from them to predict future behavior Data-Program Dependence: if data format changes,
programs need to be revised Database: Logically related set of files+definition of ALL Data DBMS: Manages Database structure/Controls access to data,
contains query language Decision Quality Dimensions: 1.Comprehensiveness: Decision reflects full consideration of facts 2.Fairness: Decision reflects the
interests of affected parties 3.Speed: Decision is not delayed 4. Coherence: Decision reflects a rational process 5.Due process: Decision based on known
process and can be appealed Deep learning: Multiple layers of transformations to produce a target output Deletion Anomaly: when delete a record that
may leads to the loss of other, it should be preserved Derived Attribute: value is derived from other attributes within the same Digital Goods: Goods that
can be delivered over a digital network. First one cost high, second cost low Direct goods: integrally involved in production process Disintermediation:
Removing intermediary like distributor or retailor Disruptive technology: 颠覆性技术 PC Distributed database: Benefits: Responsive to local needs and some
fault tolerance/Problems:Synchronization of replications, Integration of partitions Diversification: Independent business units with different customers
Downstream supply chain: Distributing the products to final customers thru distributors and retailors Drill-down: High level to more details E-
commerce: Using Internet & Web to transact business; digitally enabled transactions E-tailer: Internet retailer ERP system: Decision based on known
process and can be appealed Ad:enterprise database, standardize business process, integrate business process, lower IT cost. Dis:high acquisition and high
installation cost, considerable organizational disruption EDI: Computer-to-computer exchange of standard transactions such as invoices, purchase orders,
firm use EDI standards to define structure and information fields and for linking to other firms Entity: a collection of person, place, object, event, concept
Edge computing: Data are processed physically close to where they are collected thus improving speed Expert system: Model human
knowledge(knowledge base) / searches through the rules and “fires” those rules that are triggered by facts gathered(inference engine) Field: Meaningful
set of characters, number, etc. File: A set of records of the same type Foreign Key: 一对多里面选多 Links data in one table(PK) to the data in another table
Geoadvertising: Can tell you where is the nearest gas station Governance Costs: agency cost, bureaucratic delays Hadoop: software framework that
automatically handles hardware failures Hidden layer: Uses input data to help NN to arrive at results In memory computing: Relies on Ram than disk for
data storage Indirect goods: All goods not directly involved in production process and finished goods Information Asymmetry: One party has more info
than the other(eg. seller know more about a product's quality than buyer) Insertion Anomaly: occurs when adding a new row to a table leads to
inconsistencies IT Integration:跨系统数据分享 Ad:Coordination and agility Dis:bridging multiple systems Internal supply chain: Transforming raw materials and parts
into finished products in plants and assembly lines internally KPI: Measures selected by senior management Location Analytics: Helps decision makers
visualize problems with mapping Long tail marketing: Selling products with little demand Low-cost leadership: use IT to get low operational cost and low
price Mandatory many: Minimum One, Maximum Many Mandatory one: Minimum One, Maximum One MapReduce: framework used for processing
large datasets in a distributed manner, typically on clusters of computers Market creator: Bring buyers and suppliers together (eBay) Menu costs: Cost of
changing prices (merchant) Multivalued attribute: Net Marketplace: digital platforms where multiple sellers offer a wide range of products or services to
consumers Network effect: larger network means higher value to individual customer Neural network: Use hardware and software that parallel the
processing patterns of a biological brain. Uses a hidden layer NoSQL: allows scaling to clusters of machines OLAP: Enables users to obtain online answers
to ad hoc questions quickly Operational CRM: Customer-facing applications such as sales force automation, call center and customer service support, and
marketing automation Optional many: Minimum None, Maximum Many Optional one: Minimum None, Maximum One Partitioned database: a database
where large tables or collections of data are divided into smaller, more manageable segments called partitions Portal: point of entry and specialized
content (Yahoo, Bing, Google) Porter’s Five Forces: Competitive Rivalry, Supplier Power, Buyer Power, Threat of Substitution, and Threat of New Entry
Positional advantage: (external)geographical, distribution channel, brands Predictive analytics: Future trends and behavior based on analysis Primary
Key: unique identifiers for each row in table, should be permanent & unchangeable Primary activities: Directly adds value to company Private exchange:
Large firm using extranet to link to its suppliers, distributors, and other key business partners. Buyer is the owner, plays the central and controlling role
like set rules and control participation Project: Creates a subset consisting of selected columns in a table Push/Pull based model: push-based
systems(manufacture determine quantity) actively send data or updates to consumers, while pull-based systems require consumers to actively request
data Record: Related fields for a person, object or entity. Redundancy:Some data are contained in multiple files resulting inconsistency: e.g., employee
address may be in all three files. Relational DBMS Model: A collection of Tables: Each table has a primary key. Given the value of PK of any row, it
determines the value of each column for that row Relationship:Link between entities Replication:Independent but similar business units Richness: Text,
Audio, Video Sales force automation: Improves sales force productivity by automating routine tasks and help spot opportunities Select: Creates a subset
of all records meeting stated criteria Semi-structured decisions: Between unstructured and structured Sentiment Analysis:分析客户情绪了解市场 Silo-based
process: problem: inefficiencies, duplicated efforts, and a lack of cross-functional understanding within an organization Singled-valued attribute: can
hold only one value for a specific entity or record Spot purchasing: Involves purchase of goods based on immediateneeds in larger marketplaces that
involve many suppliers Standardization: 各部门统一数据格式 ad:efficiency and predictability Dis:requires tight management controlStrong Entity: independent and
has its own unique identifier Structured decisions: Repetitive and routine Super key: Any combination of attributes that uniquely identifies each entity
instance, can have more attributes than a Candidate key. Supervised learning: System trained with desired inputs and outputs Support activities: help
improve primary activities Switching costs:换品牌和服务产生的成本 Synergy: 协同效应 Ternary relationship: three entities relationship Text mining: allow business
extract key elements from, discover patterns in, and summarize large unstructured data sets Traditional File System: A collection of isolated files
(Employees file, Payroll file, Benefits file), each with its own programs Transaction broker: online transactions for a fee (Expedia) Transaction costs:
search, selection, bargain Transitive Dependencies:dependency of a dependency Ubiquity:无处不在 Unary relationship:只跟自己有关系 Unification: Single business
with global data access+same business processes Unstructured decisions: Novel, important, non-routine, requires judgment to solve problem
Unsupervised learning: Use brute force without verified examples Update Anomaly: redundant data in the database must be updated each place it exists
when it changes Upstream supply chain: Procuring raw materials and parts from suppliers and suppliers’ suppliers Value Chain model: Supportive &
primary activities. Independent firms coordinating activities – customer driven, less linear than value chain Value web: flexible and adapt to changes in
supply and demand Web mining: understand customer behavior, evaluate Website, success of marketing Join: Combines data from multiple relational
tables.