CS P3 (General Vocab_

General Chatbot Concepts

• Chatbot: A software application that interacts with users using natural language.

• Latency: The delay between a user's query and the chatbot’s response.

• Natural Language Processing (NLP): A field of AI that enables machines to understand and respond to human language.

• Natural Language Understanding (NLU): A component of NLP focused on interpreting user inputs.

• Critical Path: The shortest sequence of models required to process a chatbot query.

• Dataset: A collection of data used to train and evaluate machine learning models.

Latency Reduction

• Complex NLP Models: Large AI models that can slow response time due to high computational complexity.

• High Query Volume: When a chatbot receives too many simultaneous requests, increasing latency.

• Dependencies Among ML Models: If one model in the pipeline is slow, it delays the entire chatbot response.

• GPUs (Graphical Processing Units): Specialized hardware that speeds up complex computations.

• TPUs (Tensor Processing Units): Hardware optimized for deep learning tasks, improving response speed.

Linguistic Nuances

• Linguistic Nuances: Subtle differences and complexities in language that affect chatbot responses.

• Emotion and Tone Detection: Identifying the emotional context of user messages.

• Contextual Understanding: Taking past interactions into account for coherent chatbot responses.

• Ambiguity Handling: The ability to manage statements with multiple possible meanings.

• Lexical Analysis: Breaking text into individual words and sentences.

• Syntactic Analysis (Parsing): Analyzing the grammatical structure of a sentence.

• Semantic Analysis: Understanding the meaning of words and sentences.

• Discourse Integration: Integrating a sentence into the broader conversation.

• Pragmatic Analysis: Considering social, legal, and cultural context for accurate responses.

Chatbot Architecture

• Chatbot Architecture: The structure and components that enable a chatbot to process and generate responses.

• Recurrent Neural Networks (RNNs): A type of neural network designed to handle sequential data.

• Long Short-Term Memory (LSTM) Networks: A type of RNN that overcomes the vanishing gradient problem.

• Transformer Neural Networks: A modern architecture that uses self-attention for parallel processing.

• Self-Attention Mechanism: A method in transformers that helps capture relationships between words.

Backpropagation & Learning

• Backpropagation: A training algorithm for neural networks where errors are propagated backward to update weights.

• Forward Pass: The phase where input data moves through the network to generate a prediction.

• Loss Function: A measure of the difference between the chatbot’s predicted response and the actual correct response.

• Gradient Descent: An optimization algorithm that adjusts neural network weights to minimize the loss function.

• Vanishing Gradient Problem: A challenge in deep learning where gradients become too small to update earlier layers in deep networks.

Datasets & Biases

• Dataset Diversity: Inclusion of various topics, languages, and user intents for better generalization.

• Bias in AI: Systematic errors in data or algorithms that lead to unfair outcomes.

• Synthetic Data: Artificially generated data to supplement real-world data.

• Data Augmentation: Methods like paraphrasing or synonym replacement to expand training datasets.

• Historical Bias: Bias introduced by outdated or historically skewed training data.

• Selection Bias: Errors caused by non-randomly chosen training data.

Processing Power & Optimization

• Processing Power: The computational capacity of a system to perform tasks efficiently.

• Central Processing Unit (CPU): The main processor responsible for executing chatbot-related tasks.

• Graphical Processing Unit (GPU): Hardware optimized for parallel computation, crucial for NLP tasks.

• Tensor Processing Unit (TPU): Specialized AI hardware designed to accelerate deep learning tasks.

• Cloud Computing: Using remote servers to store and process data.

• Distributed Computing: Parallel processing across multiple machines to handle high query loads.

• Parallel Processing: Splitting tasks into smaller parts to process them simultaneously.

Ethical Considerations

• Data Privacy: Ensuring user data remains confidential and secure.

• Fairness in AI: Ensuring chatbots provide equitable service to all users.

• Accountability in AI: Determining who is responsible for chatbot actions and decisions.

• Transparency in AI: Providing clarity on how chatbots make decisions.

• Misinformation in AI: Preventing the spread of false or misleading information.

• Explainable AI: Making AI decision-making processes understandable for users.