Quiz 2 Concepts PDF

Python for Non-Programmers - Basics 1. Variables - Variables are like containers that store data, such as numbers or text. For example, name = "John" saves the name "John" so you can use it later in the program. 2. Data Types - These are categories for data. Common types include...

Python for Non-Programmers - Basics 1. Variables - Variables are like containers that store data, such as numbers or text. For example, name = "John" saves the name "John" so you can use it later in the program. 2. Data Types - These are categories for data. Common types include numbers (like 5), text (like "hello"), and lists (a collection like [1, 2, 3]). Knowing types helps you decide how to use each variable. 3. Operators - Symbols like + (add) and == (equal) perform actions in code. For example, 3 + 4 gives 7, and 5 == 5 checks if 5 equals 5 (True). 4. Control Structures - These are commands that help programs make decisions (like if statements) or repeat actions (for loops). They let you control what happens next. 5. Functions - Functions are shortcuts for repeating code. You define them once, give them inputs, and they return results. For example, a function to add two numbers lets you reuse the adding code whenever needed. 6. Lists - Lists hold multiple items in a specific order, like a list of favorite colors: ["red", "blue", "green"]. They’re useful when you need to group items together. 7. Strings - Text is stored as strings in Python, which lets you work with letters, words, and sentences. You can also change or combine strings, like adding “Hello, " and "world" to make "Hello, world." 8. Input and Output - input() gets data from users, while print() shows data on the screen. They let your program interact with people using it. 9. Errors - Mistakes (like typos) can stop a program. Error handling helps manage these issues so your program doesn’t crash when something goes wrong. 10. Libraries - Python has built-in and extra libraries (collections of tools) for doing di_erent tasks, like math or working with dates. Libraries make coding faster and easier. Social Media Data Collection and Annotation Pipeline 1. Data Collection - Collecting posts, comments, or messages from social media platforms to analyze what people are sharing and saying.  2. APIs - APIs (Application Programming Interfaces) allow programs to get data directly from a website. For example, the Twitter API lets you access tweets without visiting the site. 3. Web Scraping - Collecting information from a website’s pages. For example, you might “scrape” a news site for the latest headlines. 4. Data Privacy - When collecting data, it’s important to respect people’s privacy, especially with personal information. 5. Annotations - Adding tags to data. For example, tagging tweets as “positive” or “negative” helps when analyzing sentiment. 6. Metadata - This is extra data like date, time, and location of a post. It provides more context about when, where, and how data was created. 7. Data Formats - Formats (like JSON or CSV) organize collected data so it’s easier to store, read, and analyze. 8. Data Cleaning - Fixing or removing errors in the data, like deleting repeated posts or fixing typos. 9. Sampling - Choosing a smaller part of the data to study, especially if the full dataset is too large. 10. Data Bias - When data is unbalanced or unfairly represents one view or group. Detecting bias is key to making sure your analysis is accurate. Social Media Data Collection and Annotation Pipeline (2) 1. Manual Annotation - Humans add tags to data by reading and understanding it. This often produces the highest-quality labels. 2. Automated Annotation - Using tools or software to tag data automatically. It’s faster than manual but may miss details. 3. Consistency - Ensuring that data is labeled the same way by all annotators. This keeps the data accurate. 4. Annotation Rules - Clear rules that help annotators label data correctly. Rules avoid confusion and increase consistency. 5. Training Data - Labeled data used to teach AI how to recognize patterns, like teaching it what “positive” and “negative” look like in tweets.  6. Annotation Tools - Software that helps annotators label data faster, like Doccano, where you can quickly tag words or sentences. 7. Data Storage - Keeping data safe and well-organized so it’s easy to access for analysis. 8. Versioning - Tracking changes in data as it’s annotated or cleaned, so you always know the latest version. 9. Dataset Checks - Checking data to make sure all entries are correct and complete before analysis. 10. Social Media Data Privacy - Ensuring that you have permission to use social media data according to rules and privacy laws. NLP for Social Media Listening and Analysis 1. Sentiment - Analyzing if social media posts are positive, negative, or neutral. This shows public opinion or mood. 2. Topics - The main ideas discussed across posts, like common themes in tweets during an event. 3. Recognizing Names - Finding mentions of people, places, or brands. It helps in tracking specific entities in conversations. 4. Cleaning Text - Removing extra symbols (like hashtags) to make the text easier to analyze. 5. Breaking Text - Dividing sentences into words to process them separately. 6. Slang - Informal words, acronyms, or emojis that are popular on social media. Recognizing these helps make sense of casual language. 7. Language Models - Programs that help understand the meaning behind social media text. 8. Important Words - Key words or phrases that summarize the main message of a post. 9. Finding Trends - Tracking how popular topics change over time. 10. Engagement - Measuring likes, comments, and shares to understand how people respond to content. Case Studies: NLP and Media Analytics - Fake News Detection 1. Fake News - Information that isn’t true but looks real, often spread to mislead or influence people. 2. Fact-Checking - Methods to verify if information is accurate, such as checking against reliable sources. 3. Features - Characteristics that help spot fake news, like suspicious headlines or exaggerations. 4. Text Sorting - Grouping articles or posts as either real or fake. 5. Misleading Language - Words that are intended to confuse or mislead readers. 6. Source Trust - Checking if the source of information is known to be reliable. 7. Simple Models - Basic AI programs that can help identify fake news by analyzing certain features. 8. Emotion - Looking at the emotions in the text (like anger or fear) since fake news often uses strong emotions to influence people. 9. Spread of News - Seeing how fake news moves across social networks and who shares it. 10. Evaluating Success - Measuring how well fake news detection tools are working. Conversational AI in Practice: A Deep Dive into Chatbots 1. Chatbot Basics - Programs that can talk to people. They follow rules or use AI to answer questions. 2. Understanding User Goals - Identifying what the user needs based on what they ask or say. 3. Dialogue Flow - The order of responses that keep the conversation moving smoothly. 4. Understanding Language - The chatbot “understands” messages so it can reply appropriately.  5. Response Creation - Making a reply that fits the question, so users get a useful answer. 6. Backup Responses - Generic responses for when the chatbot isn’t sure how to respond. 7. Learning from Data - Training the chatbot with examples to improve its responses. 8. Personalized Replies - Customizing responses based on what the bot knows about the user. 9. Feedback - User ratings or comments that help improve the chatbot’s answers. 10. Privacy - Ensuring user data is protected and not misused. Exploring ChatGPT - Introduction to ChatGPT and its Capabilities 1. ChatGPT - An AI that can generate text based on the input it receives, creating conversations. 2. Language Model - A tool trained on a lot of text so it can answer questions or hold discussions. 3. Making Text - ChatGPT creates responses similar to human conversation. 4. Customizing - Adapting ChatGPT for tasks like answering questions about specific topics. 5. Keeping Context - Remembering what was discussed earlier in the conversation to stay relevant. 6. Good Prompts - Writing questions or instructions that guide ChatGPT’s responses. 7. Practical Uses - Examples include helping with customer service or writing assistance. 8. Limitations - Knowing where ChatGPT might make mistakes or give incomplete answers. 9. Safety - Ensuring responses don’t harm users or give risky advice. 10. Interactivity - Using ChatGPT in a back-and-forth conversation style. Prompt Engineering Basics - Introduction  1. Prompts - Questions or statements given to the AI to guide its answer. 2. Prompt Format - Writing prompts clearly and simply to get good answers. 3. Giving Instructions - Detailed directions help AI focus on the right answer. 4. Randomness - Adjusting how creative or predictable the AI’s responses are. 5. Examples - Adding sample text to show AI what kind of answers you need. 6. Zero-Shot vs. Few-Shot - No examples vs. a few examples to improve responses. 7. Unbiased Prompts - Writing prompts that avoid hinting at a particular answer. 8. Testing Prompts - Trying multiple prompts to see which one works best. 9. Templates - Pre-made prompts for common questions or tasks. 10. Limits of Prompts - Knowing that not all prompts will give perfect answers. Creative Writing and Media Content Generation with AI 1. AI Writing Tools - Tools that help you create content, like ChatGPT. 2. Adjusting Style - Changing tone (e.g., formal, casual) to fit the content purpose. 3. Story Writing - Using AI to help create story ideas or complete story sections. 4. Summarizing - Making long content shorter by focusing on the main ideas. 5. Rewriting - Restating information in di_erent words for variety or clarity. 6. Idea Brainstorming - Using AI to suggest new ideas for content. 7. Grammar Help - AI can correct spelling or grammar errors. 8. Creative Boundaries - Setting guidelines to steer AI’s creativity in a specific direction. 9. Content Types - AI can create di_erent content, like blog posts, emails, or social media posts. 10. Originality - Avoiding plagiarism by ensuring AI creates unique content.