As you get started with Relative Insight, here are the key terms you should become familiar with.
AI Summary – Powered by artificial intelligence, AI Summary transforms your data into a concise, top-level overview, highlighting key trends and metrics. By selecting relevant verbatims and connections, the summaries provide an insight into what's happening and why.
Custom Themes – A feature that allows you to create your own topic categories, so you can better understand the specific needs of your customers. By selecting topics aligned with your industry's language, you eliminate complex workarounds and improve topic identification accuracy. In addition, Custom Themes enables you to measure relative differences between specific topic groups, bringing metrics to your analysis.
Data Cleaning – Removes individual verbatims from analysis that don’t add value, such as spam, duplicates, retweets, and brand content. Cleaning can be applied to specific language sets within the Library.
Data Discovery – A visualization tool that takes raw data and converts it into visually appealing representations. The visualizations (Relationship map, Metadata visualizations, Topic treemap, and Emotions overview) offer a high-level summary, indicating which areas might require further investigation in Explore.
Data set – An uploaded text data file from any source that relates to a question you’ve defined. Once uploaded, you can analyze the file in a comparison or perform additional manipulation on the file.
Data source – Where the analyzed data originates from.
Difference – Contrasting linguistic features when comparing two data sets that are statistically significant.
Explore – This is the main screen of the platform, which uses comparison to help you learn about target audiences, customer segments, and brands by looking at what makes them unique. Once you’ve uploaded your text data into the platform, Explore allows you to navigate through statistically significant differences and similarities between data sets with analysis split across five categories – topics, grammar, phrases, words, and emotions.
Emotions overview – A Data Discovery visualization, showing the emotions expressed within the data. It helps to give context by indicating how people feel about topics and themes.
Frequency – Frequency is a measure of how common a particular linguistic feature is within a data set. It is expressed as a percentage of the total word count.
Heartbeat – A visualization tool enabling organizations to get an objective view into how bespoke themes in customer and target audience conversations are changing over time. Changes in specific topics, phrases, words or grammar elements are visualized in Heartbeat charts,
Heatmaps – A multi-comparison visualization tool, that showcases which data sets in a group are most and least different. It is particularly useful when you are comparing more than two data sets.
Insight cards – Visual representations of the most interesting, relevant, and actionable discoveries found when exploring the comparison view. They can be used to bookmark and organize important findings and can easily be exported and customized for inclusion in reports and presentations.
Integration – Technical implementation to ingest data directly from the data source via an API or file sharing mechanism.
Library – Each project is saved in a specific folder from which you can upload, manage and manipulate your data.
Linguistic feature – Relative Insight performs analysis across five categories of linguistic features - topics, phrases, words, emotions, and grammar. The comparison view segments the output of the analysis according to these categories.
Merge/Combine – The opposite of Split/Segment, it combines multiple datasets into a single one to broaden the scope of analysis.
Message frequency – A measure of how many individual blocks of text (messages) contain a particular linguistic feature. If a linguistic feature appears multiple times within a single message, it will be treated the same as if it only appears once.
Metadata – Metadata is data that provides information about other data. In the context of Relative Insight, these are additional data points associated with the pieces of text that make up your data set.
Metadata visuals – A Data Discovery visualization, showing a series of charts and diagrams showcasing the metadata associated with the dataset. These visuals provide insights into the characteristics and attributes that shape the data, facilitating a better understanding of its composition.
Metric – Measures used to quantify linguistic features. Typical metrics used include: relative difference, word frequency, word percentage, message frequency and message percentage.
Overall impact metric – Shows the overall impact of aggregated items on an insight card or Custom Theme.
Project – All work in Relative Insight is organized into projects. A single project addresses an overarching business question, research area, or client brief.
Relative difference – A measure of how much more prevalent a particular linguistic feature is in one data set compared to others. It is presented in the right-hand column when viewing differences via the comparison view and on insight cards. Relative difference enables you to bring measurable metrics to your qualitative data.
Relationship map – A Data Discovery visualization, showing a Sankey diagram, that serves as a representation of how sentiment and topics are interconnected. The thickness of each line corresponds to the strength of the relationship, providing insights into the connections within your data.
Significance – The significance value compares items between the distributions of terms. If it passes a significance test with a confidence of 99% (>6.63) and is positive, it is deemed significant and shown as a difference. The ll value in the export shows the exact number and does not filter by significance. If they are fine with a lower confidence like 95%, they take the values higher than 3.84.
Similarity – Linguistic features that two data sets have in common.
Split/Segment – Divides a dataset into subsets for analyzing specific parts of your data. Segmentation can be based on metadata values, dates, or text content.
Standard English – Relative Insight’s Standard English model is a general representation of written English. It is comprised of 9,954,331 words representing 175,954 unique words from 100,760 different sources. It is comprised of a sample of Wikipedia articles and forum conversations on a wide variety of topics. This model has been built into the platform and can be used for many comparisons.
Standard Relative Flow – Standard flows enable a customer to automate the end-to-end process for adding data to the platform, preparing the data for analysis and tracking NPS to help you determine why your customers are promoters or detractors.
Statistical Significance – In Relative Insight, it measures the confidence that the differences identified are real and not due to chance.
Topics treemap – A Data Discovery visualization, showing an intuitive treemap that highlights the main topics within your dataset. These topics are arranged based on their frequency within the data, providing a clear indication of their significance.
Verbatim – An individual message or response in the source data. Typically, this relates to each open-ended text response from the rows in the data uploaded.
Verbatim Finder – Provides instant access to relevant quotes, trends, and patterns. It visually organizes words and verbatims by topic, with metrics for significance to enable a swift creation of insightful narratives.