Terminology

Question 1

Define ‘Chatbot / Language model chatbot’.

Answer 1

A chatbot is an AI system that uses natural language processing—often powered by a trained language model—to converse with users and generate human-like responses.

Question 2

Define ‘Latency’.

Answer 2

The delay between a user’s input and the chatbot’s response, which affects how quickly the system interacts.

Question 3

Define ‘Response time’.

Answer 3

The amount of time the chatbot takes to produce an answer after receiving a query.

Question 4

Define ‘Critical path’.

Answer 4

The minimal sequence of dependent processing steps or models required to transform user input into a response; delays in any step slow the overall output.

Question 5

Define ‘Natural language understanding (NLU)’.

Answer 5

The component of NLP that interprets and converts unstructured user input into structured, machine-actionable data.

Question 6

Define ‘Machine learning models’.

Answer 6

Algorithms trained on data to recognize patterns and make predictions; in chatbots, they learn language patterns to understand and generate responses.

Question 7

Define ‘Machine‐actionable information’.

Answer 7

Data that has been structured or processed into a format that a computer system can directly use to perform tasks or decisions.

Question 8

Define ‘Training dataset’.

Answer 8

A collection of examples used to train a model so that it can learn to understand language and generate appropriate responses.

Question 9

Define ‘Domain‐specific dataset’.

Answer 9

A dataset focused on a particular subject area, providing the specialized vocabulary and context needed for a chatbot in that field.

Question 10

Define ‘Linguistic nuances’.

Answer 10

Subtle variations in language such as tone, emotion, or ambiguity that impact meaning and must be understood for accurate responses.

Question 11

Define ‘Context’.

Answer 11

The surrounding conversation and background information that gives meaning to the current user input.

Question 12

Define ‘Natural language processing (NLP)’.

Answer 12

The field of AI that enables computers to understand, interpret, and generate human language.

Question 13

Define ‘Lexical analysis’.

Answer 13

The process of breaking down text into words or tokens as the first step in understanding language.

Question 14

Define ‘Tokenization’.

Answer 14

Splitting text into smaller units (tokens) like words or subwords, making it easier for the model to process.

Question 15

Define ‘Stemming’.

Answer 15

A technique that reduces words to their root form by cutting off endings, aiding in recognizing word variants.

Question 16

Define ‘Lemmatization’.

Answer 16

The process of converting a word to its base or dictionary form using language rules, improving consistency in text analysis.

Question 17

Define ‘Syntactic analysis (Parsing)’.

Answer 17

Analyzing a sentence’s grammatical structure—often via a parse tree—to understand how words relate to each other.

Question 18

Define ‘Parse tree’.

Answer 18

A tree diagram that represents the grammatical structure of a sentence, showing the relationships among its parts.

Question 19

Define ‘Semantic analysis’.

Answer 19

The process of determining the meaning of a sentence by interpreting the definitions and relationships of its words.

Question 20

Define ‘Discourse integration’.

Answer 20

Linking the meaning of a sentence with previous parts of the conversation to maintain context over multiple exchanges.

Question 21

Define ‘Pragmatic analysis’.

Answer 21

Understanding the intended meaning behind words by considering context, real-world knowledge, and implied nuances.

Question 22

Define ‘Sentiment analysis’.

Answer 22

Evaluating the emotional tone of text to determine whether it is positive, negative, or neutral.

Question 23

Define ‘Dialogue state tracking’.

Answer 23

Maintaining and updating the conversation context over multiple turns to ensure coherent and relevant responses.

Question 24

Define ‘Multi‐turn dialog management’.

Answer 24

Handling conversations that involve several back-and-forth exchanges, ensuring continuity and context retention.

Question 25

Define 'Entity recognition'.

Answer 25

Identifying and extracting key pieces of information (like names or dates) from user input.

Question 26

Define 'Intent recognition'.

Answer 26

Determining the user's purpose or goal behind a message, guiding the chatbot’s response.

Question 27

Define 'Dialogue management'.

Answer 27

The system that decides the next action or response based on user input and conversation context.

Question 28

Define 'Clarification handling'.

Answer 28

Strategies the chatbot uses to ask follow-up questions or rephrase ambiguous queries to understand user intent.

Question 29

Define 'Natural language generation (NLG)'.

Answer 29

The process of converting structured data or intents into fluent, human-like language for responses.

Question 30

Define 'Architecture (of the chatbot)'.

Answer 30

The overall design and structure of the chatbot system, including its components and their interactions.

Question 31

Define 'Machine learning algorithms'.

Answer 31

The methods used to train models from data, enabling the chatbot to learn language patterns and decision-making rules.

Question 32

Define 'Recurrent neural networks (RNNs)'.

Answer 32

Neural networks designed for sequential data that process one input at a time while maintaining a memory of previous inputs.

Question 33

Define 'Transformer neural networks (Transformer NNs)'.

Answer 33

Neural networks that use self-attention to process entire sequences in parallel, capturing long-range dependencies more effectively.

Question 34

Define 'Neural network'.

Answer 34

A computational model composed of interconnected layers of nodes that learns patterns from data through adjustable weights.

Question 35

Define 'Sequential data'.

Answer 35

Data where the order of elements matters, such as words in a sentence or turns in a conversation.

Question 36

Define 'Input layer'.

Answer 36

The first layer of a neural network that receives and processes the initial input data.

Question 37

Define 'Hidden layers'.

Answer 37

Intermediate layers in a neural network that extract and transform features from the input data.

Question 38

Define 'Output layer'.

Answer 38

The final layer of a neural network that produces the model’s response or prediction.

Question 39

Define 'Backpropagation through time (BPTT)'.

Answer 39

A training algorithm for RNNs that unfolds the network over time and propagates error gradients back through each time step.

Question 40

Define 'Vanishing gradient problem'.

Answer 40

A challenge in training deep or sequential models where gradients diminish over time, hindering learning of long-term dependencies.

Question 41

Define 'Long Short-Term Memory (LSTM)'.

Answer 41

A type of RNN that uses gating mechanisms to maintain and control long-term memory, mitigating the vanishing gradient problem.

Question 42

Define 'Input gate'.

Answer 42

In an LSTM cell, the component that controls how much new information is added to the cell state.

Question 43

Define 'Forget gate'.

Answer 43

In an LSTM cell, the mechanism that decides which information to discard from the cell state.

Question 44

Define 'Output gate'.

Answer 44

In an LSTM cell, the mechanism that determines what part of the cell state is output at each time step.

Question 45

Define 'Memory cell state'.

Answer 45

The internal storage within an LSTM cell that retains information over long sequences.

Question 46

Define 'Self-attention mechanism'.

Answer 46

A method used in Transformers where each token in a sequence assesses its relevance to every other token, allowing the model to capture global dependencies.

Question 47

Define 'Generative Pre-trained Transformer 3 (GPT-3)'.

Answer 47

A large-scale language model based on the Transformer architecture, known for generating highly fluent and contextually relevant text.

Question 48

Define 'Unbiased dataset'.

Answer 48

A dataset curated to be as balanced and representative as possible, minimizing skew that could cause the chatbot to favor certain patterns or groups.

Question 49

Define 'High-quality dataset'.

Answer 49

A large, well-organized, and accurate collection of data that is relevant to the chatbot's domain, ensuring reliable learning and performance.

Question 50

Define 'Real data'.

Answer 50

Data collected from actual user interactions or real-world events, as opposed to artificially generated data.

Question 51

Define 'Synthetic data'.

Answer 51

Artificially created data used to augment real data, often generated to simulate specific scenarios or increase dataset diversity.

Question 52

Define 'Confirmation bias'.

Answer 52

This form of bias occurs when the dataset is biased towards a particular viewpoint, such as training data that only include customer queries related to certain types of policies.

Question 53

Define 'Historical bias'.

Answer 53

This form of bias occurs when the training data do not reflect changes over time. For example, if the natural language processing model is trained on data from several years ago, it may not be able to accurately predict recent customer queries.

Question 54

Define 'Labelling bias'.

Answer 54

This form of bias occurs when the labels applied to the data are subjective, inaccurate or incomplete. For example, if the labels assigned to customer queries are too generic, the model may not be able to accurately predict the customer’s intent.

Question 55

Define 'Linguistic bias'.

Answer 55

This form of bias occurs when the dataset is biased towards certain linguistic features, such as dialect or vocabulary. For example, if a dataset is built on formal written language, the model may not be able to accurately interpret informal language.

Question 56

Define 'Sampling bias'.

Answer 56

This form of bias occurs when the training dataset is not representative of the entire population, such as training data that only include customer queries from one demographic.

Question 57

Define 'Selection bias'.

Answer 57

This form of bias occurs when the training data are not randomly selected but are instead chosen based on some criteria. A language model trained on data that suggest certain demographics may be more likely to file insurance claims that are biased towards people who fall under that category.

Question 58

Define 'Pre-processing'.

Answer 58

The process of cleaning and transforming raw text data into a usable format for model training, such as tokenization and normalization.

Question 59

Define 'Bag-of-words algorithm'.

Answer 59

A simple method for text representation that counts word frequencies, disregarding grammar and order, to provide a basic input for models.

Question 60

Define 'Vector representation'.

Answer 60

Encoding text as numerical vectors (embeddings) that capture semantic meaning, allowing the chatbot to process and compare textual information.

Question 61

Define 'Training the model'.

Answer 61

The process of adjusting a machine learning model’s weights by exposing it to data so that it learns to perform its task accurately.

Question 62

Define 'Deployment of the model'.

Answer 62

Integrating and running the trained model in a live environment where it can interact with users in real time.

Question 63

Define 'Computational power'.

Answer 63

The processing capability of hardware (like CPUs, GPUs, or TPUs) that is required to train and run models efficiently.

Question 64

Define 'Central Processing Unit (CPU)'.

Answer 64

A general-purpose processor that executes instructions and performs basic computations; adequate for simple tasks but slower for large-scale neural network processing.

Question 65

Define 'Graphics Processing Unit (GPU)'.

Answer 65

A specialized processor that excels at parallel computations, significantly speeding up the training and inference of neural networks.

Question 66

Define 'Tensor Processing Unit (TPU)'.

Answer 66

A custom hardware accelerator designed specifically for machine learning tasks, offering high efficiency for large-scale model training and inference.

Question 67

Define 'Memory (RAM/VRAM)'.

Answer 67

Fast-access storage used during computation; RAM is used by CPUs and VRAM by GPUs, both essential for holding model data and intermediate results during processing.

Question 68

Define 'Storage'.

Answer 68

Long-term data retention hardware (such as SSDs or HDDs) used to store datasets and model files, which are loaded into memory for processing.

Question 69

Define 'Bandwidth'.

Answer 69

The data transfer rate available within a system or network, important for quickly moving data between storage, memory, and processors.

Question 70

Define 'Distributed computing'.

Answer 70

Using multiple computers or servers simultaneously to share the computational load, enabling faster training or processing of large models.

Question 71

Define 'Parallel processing'.

Answer 71

Executing many calculations simultaneously, which is critical for efficiently training or running neural networks in chatbots.

Question 72

Define 'Data parallelism'.

Answer 72

A method where multiple processors handle different portions of a dataset concurrently during model training, speeding up the process.

Question 73

Define 'Model parallelism'.

Answer 73

A technique that splits different parts of a large model across multiple processors, allowing training of models too big for a single device.

Question 74

Define 'Batching'.

Answer 74

Grouping multiple inputs together to process them simultaneously, improving efficiency by leveraging parallel computations.

Question 75

Define 'Vectorization'.

Answer 75

Optimizing computations by processing data in batches as vectors or matrices, rather than one element at a time, to fully utilize hardware capabilities.

Question 76

Define 'Profiling'.

Answer 76

The practice of analyzing a program's performance to identify bottlenecks or inefficient code segments for optimization.

Question 77

Define 'Fine-tuning code'.

Answer 77

Optimizing and refining the model’s implementation or hyperparameters after initial training to improve performance without altering core functionality.

Question 78

Define 'Load balancing'.

Answer 78

Distributing workload evenly across multiple servers or processors to ensure no single component is overwhelmed, which is essential for handling many chatbot requests concurrently.

Question 79

Define 'Cloud services'.

Answer 79

Remote computing resources provided over the internet that enable scalable deployment and training of chatbots without maintaining physical hardware.

Question 80

Define 'Two-tier model'.

Answer 80

An architecture separating the system into two layers—typically a front-end interface and a back-end processing server—used to efficiently manage chatbot interactions.

Question 81

Define 'Encryption (in transit and at rest)'.

Answer 81

The process of encoding data to prevent unauthorized access, both during transmission (in transit) and when stored (at rest).

Question 82

Define 'Anonymization'.

Answer 82

Removing or masking personal identifiers in data so that individual users cannot be recognized, protecting privacy in training and logs.

Question 83

Define 'Secure APIs'.

Answer 83

Application interfaces that are protected with authentication and encryption, ensuring safe data exchange between the chatbot and other systems.

Question 84

Define 'Bias'.

Answer 84

Systematic errors or unfair tendencies in a model’s outputs, often resulting from skewed training data.

Question 85

Define 'Fairness'.

Answer 85

The principle that a chatbot should treat all users equitably without discrimination or undue preference.

Question 86

Define 'Transparency'.

Answer 86

Clarity regarding how a chatbot functions, including its decision-making process and limitations, to build user trust.

Question 87

Define 'Misinformation'.

Answer 87

False or inaccurate information generated by a chatbot, which can mislead users if not properly controlled.

Question 88

Define 'Conversation logging'.

Answer 88

Recording chatbot interactions for monitoring, troubleshooting, and future model improvements, while ensuring user privacy.

Question 89

Define 'Sensitive content filtering'.

Answer 89

The process of detecting and blocking inappropriate or harmful content in a chatbot's responses or user inputs.

Question 90

Define 'Reinforcement Learning from Human Feedback (RLHF)'.

Answer 90

A training method where human evaluations of responses guide the model to align more closely with desired behaviors and ethical standards.

Question 91

Define 'Loss function'.

Answer 91

A mathematical function that measures the error between a model's predictions and the actual target values; minimizing this function improves model accuracy.

Question 92

Define 'Deep learning'.

Answer 92

A subset of machine learning that uses deep (multi-layered) neural networks to learn complex representations from large amounts of data.

Question 93

Define 'Hyperparameter tuning'.

Answer 93

The process of adjusting the external parameters of a model (such as learning rate or number of layers) to optimize performance.

Question 94

Define 'Large language model (LLM)'.

Answer 94

A neural network with a vast number of parameters trained on extensive text data, capable of understanding and generating human-like language.

Question 95

Define 'Weights'.

Answer 95

The adjustable parameters within a neural network that are optimized during training to capture patterns and make accurate predictions.