A Clarity Score is a measure that indicates how easily a piece of text can be understood by its intended audience. It typically uses factors such as sentence length, word complexity, and structure to evaluate readability. The score helps writers and editors make sure their content is clear and accessible.
Annotator scores are numerical ratings or evaluations given by people who label or review data, such as texts, images or videos. These scores reflect the quality, relevance or accuracy of the information being labelled. Collecting annotator scores helps measure agreement between different annotators and improves the reliability of data used in research or machine learning.
Response labelling is the process of assigning descriptive tags or categories to answers or outputs in a dataset. This helps to organise and identify different types of responses, making it easier to analyse and understand the data. It is commonly used in machine learning, surveys, or customer service systems to classify and manage information efficiently.
Result feedback is information given to someone about the outcome of an action or task they have completed. It helps people understand how well they performed and what they might improve next time. This process is important in learning, work, and technology, as it guides future behaviour and decision-making.
Concept recall is the ability to remember and retrieve key ideas, facts or principles that you have previously learned. It is an important part of learning because it helps you use information when you need it rather than just recognising it when you see it. Strong concept recall means you can explain or use a…
Output labels are the names or categories that a system or model assigns to its results. In machine learning or data processing, these labels represent the possible answers or outcomes that a model can predict. They help users understand what each result means and make sense of the data produced.
Retry reasoning is a process where a system or program decides whether to try an action again after it fails. Instead of simply repeating the same step blindly, the system analyses why the failure happened and chooses the best way to proceed. This approach helps to avoid repeating mistakes and increases the chances of eventual…
Output guards are mechanisms or rules that check and control what information or data is allowed to be sent out from a system. They work by reviewing the output before it leaves, ensuring it meets certain safety, privacy, or correctness standards. These are important for preventing mistakes, leaks, or harmful content from reaching users or…
Label calibration is the process of adjusting the confidence scores produced by a machine learning model so they better reflect the true likelihood of an outcome. This helps ensure that, for example, if a model predicts something with 80 percent confidence, it will be correct about 80 percent of the time. Calibrating labels can improve…
A stability index is a measure used to assess how stable or consistent something is over time. It is often used in fields like engineering, finance, and data analysis to identify changes or shifts in a system, data set, or process. By calculating the stability index, organisations can monitor trends, detect problems early, and make…