All Collections
Uploading Data
How to upload your own data
How to upload your own data
Alex avatar
Written by Alex
Updated over a week ago

Hi!

We are very excited that you are visiting our text analytics platform.

To analyze your own data, just follow these simple steps to get you started:

  1. Go to the Analysis Page

  2. Click on "+ New Analysis"

  3. Select the Data Type that you want to analyze and click on "Next"

  4. Enter a Name for the Analysis and a short description, then click "Submit"

  5. To upload your own data, click on "File Upload"

  6. Click on "Select File", and chose the .csv file that you want to analyze.

  7. You arrive to the most difficult part, the "Column Mapping".

    We need you to let us know a bit about each column of the file. We need to know:

    Which of the columns of your data file contains...

    1. ...the text that you want to analyze (required)

      • The texts usually are answers to open-ended questions

      • The texts will be used to predict the sentiment, extract keywords, detect the categories etc...

      • There can be only one text column. At this moment we do not support files with more than one text column. (We are working on supporting this)

    2. ...the date when that text was created (optional)

      • This date will be used to show time evolutions on your dashboard

      • A date filter will be created on your dashboard

    3. ...the NPS evaluation (optional)

      • This column will be used to calculated the NPS metric

      • A range filter will be created on your dashboard

      • NPS values must be between 0 and 10

    4. ...the CSAT evaluation (optional)

      • This column will be used to calculate the Customer Satisfaction metric

      • A range filter will be created on your dashboard

      • CSAT values can have different ranges, but usually lie between 0 and 5, 0 and 7, or, 0 and 9.

    5. ...categorical metadata (optional)

      • These columns will be used to create multi-select filters on the dashboard

      • Questions where the user must select one of multiple options usually are categorical metadata (single-select questions)

      • The values in these columns cannot be answers to open-ended questions.

    6. ...numerical metadata (optional)

      • These columns will be used to create range filters on the dashboard

      • The values in these columns must be numbers

  8. Let's take the airline demo file as an example

    1. Column "airline" contains a predefined set of different airlines. We want this to be a multi-select filter on the dashboard. We have to select "Multiselect Filter / Categories"

    2. The "review_date" is our date column

    3. The "overall" column contains a rating from 0 to 10 which we take as the NPS score

    4. The "traveller_type" column again contains predefined values, such as "Business", "VIP", or "Family Leisure". Again we select "Multiselect Filter/Categories"

    5. The "food_bev" column again contains a numerical rating of the Food and Beverages offered. This column contains numbers and we want a range filter on our dashboard. So, we select "Range Filter/Numbers"

  9. Click on Submit. After validation passes, you are greeted with the screen to connect our predefined Deepers to the dashboard. Choose the Deepers that you want to use, then click "Continue"

    That's all. On the next page, you simply need to confirm to start running the analysis. You will receive an email notification once everything is finished.

Thanks for reading. Do not hesitate to contact us if you have any unresolved questions!

Did this answer your question?