MIS 3

What is data

Raw facts and figures

What is information

Data presented in context so it can answer a question or support decision-making

What is knowledge

Insight derived from experience and expertise based on data and information

What is a database

A single table or a collection of related tables

What is a DBMS

Software for creating, maintaining, and manipulating data

What is SQL

A language used to create and manipulate databases

What is a DBA

A professional responsible for database design, implementation, maintenance, backup, recovery, policy enforcement, and security

What is a record in a database

A row representing one instance of whatever the table tracks

What is a field in a database

A column representing a category of data in a record

What is a key in a database

A field or combination of fields used to uniquely identify a record and relate tables

What is a relational database

A database where tables are related based on common keys

What is a data aggregator

A firm that collects and resells data

What is big data

Extremely large, complex, and often unstructured datasets used to generate insights that would otherwise be difficult or impossible to obtain

What are the three Vs of big data

Volume, velocity, and variety

What is business intelligence

Reporting, data exploration, ad hoc queries, and sophisticated data modelling and analysis for decision-making

What is analytics

The extensive use of data, statistical and quantitative analysis, explanatory and predictive models, and fact-based management to drive decisions

What is a data lake

A storage and access system for big data that can hold structured and raw unfiltered data

What is a data warehouse

A set of databases used to support decision-making in an organisation

What is a data mart

A database focused on a specific problem or business unit

What is ETL

Extract, Transform, Load, copying data from multiple sources, cleaning it into a common format, and loading it into a usable form

What is a canned report

A regular summary of information in a predetermined format

What is an ad hoc report

A custom report created by users as needed by selecting their own parameters

What is a dashboard

A visual heads-up display of critical indicators and key performance metrics

What is OLAP

Online analytical processing that summarises data from relational databases and stores it in a data cube

What is a data cube

A special database used to store data for OLAP reporting

What is data mining

The use of computers to identify hidden patterns in and build models from large datasets

What are prerequisites for data mining to work

Clean consistent data and data events that reflect real trends

What is one risk of bad data

It can produce wrong estimates and expose the firm to risk

What is over-engineering in data analysis

Building a model with so many variables that it only works on the data used to create it

What is generative AI

A type of AI that uses neural networks and deep learning to learn patterns from existing data and generate original content

How is generative AI different from traditional AI

Traditional AI mainly predicts or classifies, while generative AI creates new outputs

What kinds of output can generative AI create

Text, images, audio, video, and other original content

What is machine learning

Training algorithms to learn patterns and make predictions from data

What is a neural network

A system inspired by the brain that uses interconnected nodes to process information

What is deep learning

A type of neural network with multiple layers that enables more sophisticated learning

What is an LLM

A large language model designed to understand human language and generate intelligent responses

What are LLMs trained on

Enormous datasets, often sourced from books, articles, websites, and other text-based sources

What architecture enabled modern LLM breakthroughs

The transformer architecture

Why are transformers important

They are especially effective at understanding context and generating text, images, audio, and code

What does ChatGPT stand for

Chatbot Generative Pretrained Transformer

Why are GPUs important for LLMs

They handle the parallel processing and matrix calculations needed for faster training and inference

Why are CPUs less suited than GPUs for LLM workloads

CPUs are general-purpose processors, while GPUs are better for highly parallel machine learning tasks

What share of company data is often unstructured according to the PDF

About 80 to 90 percent

What is first-party data

Internal data produced through a company’s everyday interactions with customers and prospects

What is second-party data

Data produced by or in collaboration with trusted partners

What is third-party data

External data acquired to enrich internal datasets

How can LLMs help with content generation

They can generate drafts of blogs, posts, press releases, product descriptions, and images

How can LLMs help enterprises with chatbots

They can answer business questions using private company data while supporting personalised responses

How can LLMs help with document analysis

They can summarise unstructured text and convert it into structured table formats

How can LLMs act as reasoning engines

They can extract meaning from text for tasks like sentiment analysis and insight generation

How can LLMs act as translation engines

They can translate text across languages

How can LLMs support retrieval and summarisation

They can search large bodies of text and generate concise summaries

What is a task-specific LLM

An LLM specialised for a narrow task such as generating software code

What is a domain-specific LLM

An LLM trained for a specific field such as biomedicine, law, or cybersecurity

What does task alignment mean in bringing LLMs into production

Matching the model to the actual business task it must perform

What is semantic caching

Temporarily storing semantic representations or embeddings for more efficient and meaningful responses

What is feature engineering and injection in LLM production

Adding extra information or features to improve personalisation and task performance

What does reducing latency mean in LLM systems

Reducing the time it takes the model to produce a response after receiving input

What is a data centre

A facility that hosts computer systems and associated components to store, process, manage, and distribute large amounts of data

What does a standard data centre include

Multiple networked computers, storage, computing infrastructure, and cooling systems

What is one security benefit of data centres

They typically provide enterprise-grade security for critical business and customer information

What is resource optimisation in data centres

Delegating much of the IT workload such as computation and storage to the hosting provider

How do data centres support cloud offerings

They enable businesses to provide cloud-based solutions such as SaaS web and mobile apps

Why is physical location important for a data centre

It affects available space, security, legal compliance, and network performance

Why do data protection laws matter in data centre location

Personal data may be subject to rules like GDPR that restrict where data can be stored or transferred

Why does proximity to users matter for a data centre

Being closer to the target audience can improve speed and reduce latency

What do data centre staff do

They monitor, maintain, and manage infrastructure around the clock

What is core hardware in a data centre

The servers, processors, storage devices, networking equipment, and security features needed for IT operations

What is support infrastructure in a data centre

Systems that keep the data centre operational and maximise uptime

What is a UPS system

An uninterrupted power supply that helps keep systems running during power issues

What is HVAC in a data centre

Heating, ventilation, and air conditioning systems used to control environmental conditions

What is an enterprise data centre

A company-owned data centre used for its own storage and computing needs

What is a hyperscale data centre

A very large data centre built to support large-scale IT projects such as search engines, social media, and cloud computing

What is a managed data centre

A data centre owned by a third party that leases services to other businesses

What is a cloud-based data centre

A distributed data centre run by a third-party provider where space and infrastructure can be rented as needed

What is colocation

A multi-tenant data centre where businesses host their own servers off-site while the provider supplies power, cooling, security, and networking

What is an edge data centre

A smaller data centre located near end users to support low-latency real-time processing for uses like IoT and automation