Quiz title: C6 Practical Implementation of ML and NLP

Title: Q1 Multiple Choice
Points: 1
1. What is the main challenge when transitioning from individual ML models to business systems?
a) Models become less accurate in business environments
*b) Integration complexity and real-world data differences create new challenges
c) Business systems require different algorithms
d) Models need to be retrained completely

Title: Q2 Multiple Choice
Points: 1
2. What is data drift in the context of deployed ML systems?
a) Data becoming corrupted over time
*b) Real-world data differing from the training data used to build models
c) Data storage systems failing
d) Data being accidentally deleted

Title: Q3 Multiple Choice
Points: 1
3. Why might a customer churn prediction model with 85% accuracy still fail in business deployment?
a) 85% accuracy is too low for business use
*b) It may not integrate well with business workflows or account for changing customer behavior
c) Business systems require 100% accuracy
d) Churn prediction is not useful for businesses

Title: Q4 Multiple Choice
Points: 1
4. What is a key difference between technical performance metrics and business success metrics?
a) Technical metrics are more important than business metrics
*b) Business metrics measure user experience and business impact, not just model accuracy
c) Business metrics are easier to calculate
d) Technical metrics are only used during development

Title: Q5 Multiple Choice
Points: 1
5. What is bias in ML systems?
a) A preference for certain data formats
*b) When models systematically produce worse or unfair outcomes for certain groups
c) When models are too complex
d) When models run too slowly

Title: Q6 Multiple Choice
Points: 1
6. How does bias typically arise in ML systems?
a) Through intentional discrimination by developers
*b) From incomplete, unbalanced, or historically biased training data
c) From using too much data
d) From models being too accurate

Title: Q7 Multiple Choice
Points: 1
7. Why can bias remain hidden during model evaluation?
a) Bias detection tools don't exist
*b) Models may appear accurate overall while performing poorly for specific demographic groups
c) Bias only appears in production systems
d) Evaluation metrics cannot measure bias

Title: Q8 Multiple Choice
Points: 1
8. What is the first step in a conversational AI system when processing user input?
a) Sentiment analysis
*b) Intent classification to determine what the user wants to accomplish
c) Response generation
d) Entity extraction

Title: Q9 Multiple Choice
Points: 1
9. Intent classification in chatbots is fundamentally what type of machine learning problem?
a) Unsupervised learning
*b) Supervised learning classification
c) Reinforcement learning
d) Clustering

Title: Q10 Multiple Choice
Points: 1
10. What is Named Entity Recognition (NER) in conversational AI?
a) Recognizing famous people in conversations
*b) Extracting specific information like products, dates, or locations from user messages
c) Identifying the user's name
d) Recognizing the AI system's name

Title: Q11 Multiple Choice
Points: 1
11. Why is context management important in conversational AI?
a) To make conversations longer
*b) To understand references across multiple messages in a conversation
c) To store user personal information
d) To improve response speed

Title: Q12 Multiple Choice
Points: 1
12. How does sentiment analysis help in conversational AI systems?
a) It makes responses more accurate
*b) It helps adapt responses to the customer's emotional state
c) It identifies the customer's location
d) It translates messages to different languages

Title: Q13 Multiple Choice
Points: 1
13. What are the two main approaches to response generation in conversational AI?
a) Fast and slow responses
*b) Template-based and neural response generation
c) Automated and manual responses
d) Short and long responses

Title: Q14 Multiple Choice
Points: 1
14. What is extractive summarization?
a) Creating completely new summaries from scratch
*b) Pulling key sentences directly from documents
c) Translating documents to other languages
d) Converting documents to different formats

Title: Q15 Multiple Choice
Points: 1
15. What is abstractive summarization?
a) Removing abstract concepts from documents
*b) Rewriting information in a concise form using new language
c) Creating abstracts for research papers only
d) Summarizing only numerical data

Title: Q16 Multiple Choice
Points: 1
16. How are text summarization systems typically used in business?
a) To replace human decision-making entirely
*b) To help people process large amounts of information more quickly
c) Only for academic research
d) To translate documents

Title: Q17 Multiple Choice
Points: 1
17. What is the "humans in the loop" approach for AI systems?
a) Having humans write all the code
*b) AI handles initial processing while humans review, correct, and approve outputs
c) Humans and AI working on completely separate tasks
d) AI systems that only work when humans are present

Title: Q18 Multiple Choice
Points: 1
18. Why is human oversight essential in AI-powered business systems?
a) AI systems are always inaccurate
*b) Humans provide context, empathy, and ethical reasoning that AI systems lack
c) AI systems are too expensive to run alone
d) Legal requirements mandate human involvement

Title: Q19 Multiple Choice
Points: 1
19. What is the primary goal of responsible AI implementation?
a) To build the most sophisticated AI system possible
*b) To create solutions that improve business outcomes while maintaining human judgment
c) To replace all human workers with AI
d) To maximize AI system accuracy at any cost

Title: Q20 Multiple Choice
Points: 1
20. Why is regular performance monitoring important for deployed AI systems?
a) To increase system costs
*b) To ensure systems continue working effectively as conditions change
c) To make systems more complex
d) To satisfy regulatory requirements only

Title: Q21 Multiple Choice
Points: 1
21. What does fairness evaluation in AI systems assess?
a) Whether the system runs fast enough
*b) Whether the system treats all users appropriately without discrimination
c) Whether the system is profitable
d) Whether the system uses the latest technology

Title: Q22 Multiple Choice
Points: 1
22. In the context of AI systems, what role does domain expertise play?
a) It's not necessary if the AI is sophisticated enough
*b) It helps ensure AI systems are applied appropriately and outputs are interpreted correctly
c) It only matters during system development
d) It slows down AI implementation

Title: Q23 Multiple Choice
Points: 1
23. What is a key principle for successful AI deployment in business?
a) Fully automate all processes immediately
*b) Thoughtfully integrate automated capabilities with human oversight
c) Use the most complex AI models available
d) Replace human judgment with AI decisions

Title: Q24 Multiple Choice
Points: 1
24. How should businesses measure the success of AI systems?
a) Only through technical accuracy metrics
*b) Through a combination of technical performance, user experience, and business impact
c) Only through cost reduction
d) Only through speed improvements

Title: Q25 Multiple Choice
Points: 1
25. What is the most important lesson for responsible AI implementation?
a) Always use the newest AI technology
*b) Choose appropriate models, evaluate them honestly, and use them to support human judgment
c) Maximize automation in all business processes
d) Focus only on technical performance metrics