Analyze My First Dataset
Steps to analyse a text field with two herotText workflows
On this page, we explain two text analysis scenarios, both of which rely on state-of-the-art AI techniques.
- Identifying Themes: one-to-one grouping
Here, all entries in a dataset are processed. Themes are identified based on conceptual similarities. Each entry is assigned to only one of the themes. This scenario is called clustering.
Example:
Sydney's weather and landscape is amazing=>theme = Sydney
Update: We have introduced a new version of AI-Clustering where "One-To-Many" clustering is available. It works by breaking the text into its composing sentences behind the scene to decide on all possible clusters.
- Identifying Themes: one-to-many grouping
Here, all entries in a dataset are processed. Tags/code frames are extracted. Each entry is tagged with relevant tags based on conceptual similarity. This scenario is called tagging.
Example:
Sydney's weather and landscape is amazing=>tags = Sydney, weather, landscape
Note: Following steps require zero coding/programming.
Clustering
Clustering
Identifying themes: Applicable for bothone-to-one and one-to-manygrouping
The 2 main components for text clustering are shown in the image below. Each component is a no-code workflow. Step-by-step guide is available via the provided links.
Relevance AI - AI Clustering flow
1. Upload Data
Prepare your dataset as a CSV file and upload your data to Relevance AI's platform.
Non-English Data? Not a problem :)
If your free-text data is not in English or is composed of multiple languages, make sure to use our no-code translation workflow after uploading your data.
2. AI Clustering
AI Clustering workflow groups your data entries based on their conceptual similarities and identifies themes. You will be presented with a pre-filled data Explorer dashboard. Explorer is your fully configurable dashboard to understand your data and extract insights. Have a glance at the Explorer many functionalities and follow our guides to create your Explorer dashboard.
Tagging
Tagging
Identifying themes: one-to-many grouping
The 2 main components for Tagging are shown in the image below. Each component is a no-code workflow. Step-by-step guide is available via the provided links.
Relevance AI - AI Tagging flow
1. Upload Data
Prepare your dataset as a CSV file and upload your data to Relevance AI's platform.
Non-English Data? Not a problem :)
If your free-text data is not in English or is composed of multiple languages, make sure to use our no-code translation workflow after uploading your data.
2. AI Tagging
Under Relevance AI's AI Tagging workflow, you will
- use AI to automatically extract tags/code frames from your dataset
- apply your knowledge of the field to modify (i.e. add to/remove from) the extracted tags
- use AI to label your dataset
- be presented with a pre-filled data Explorer dashboard. Explorer is your fully configurable dashboard to understand your data and extract insights. Have a glance at the Explorer many functionalities and follow our guides to create your Explorer dashboard.
Common Questions?
- What if I have my own tags and don't need AI to generate the tags/code frames?
You can skip step 2 and 3 in the Tagging scenario and follow the instruction at Tag With Your Code Frames instead. - How can I perform sentiment analysis?
You can run Sentiment Analysis at any stage after your data is uploaded. - How can I perform emotion analysis?
You can run Emotion Analysis at any stage after your data is uploaded.
Useful links
Find insights through clustering
Updated 30 days ago