Emerging Trends

Identify key trends in the most recent data

Identification of the emerging trends is one of the key factors when defining the main key action points for individuals, teams and companies.
The context is very broad and applicable to indefinite scenarios:

  • In the news field: what have been the most viewed articles about?
  • In tweets: what have been people talking about the most?
  • In a review platform: what have been the most recent complaints about?
  • In tech and programming: what have been the most common tools?
    And the list goes on.

On this page, we explain how Relevance AI can help you identify the emerging trends on a no-code platform.


A dataset containing the data that you work with in day-to-day work-life. Text data such as news, tweets, customer reviews, and even images can be processed on Relevance AI's platform. Your dataset must

  • be in valid CSV or JSON format
  • contain the main data field as well as the corresponding dates in "yyyy-mm-dd" format
    For instance, a CSV file of all tweets that were posted last month and the date when each tweet was published.

Sample dataset - with fields such as tweet, Date, hashtags, etc.


  1. Upload your data
    Follow the simple and completely automated upload process to upload your CSV file to Relevance AI's platform.
  2. Follow the no-code clustering process on the field you want to analyse (e.g. News, tweets, reviews or image URL) to categorize your data. Clustering means grouping items similar to each other together. This helps identify general topics/categories existing in your dataset. For instance, in a News dataset one cluster can be "sport", and the other can be "politics"; or in a review dataset, one cluster can be about "response time" and the other can be the "quality".
  3. Work on insights on the Explorer app. The Explorer app is pre-filled. However, we have put together a thorough guide on how to configure the Explorer app so that everyone can configure the page the best way that suits them.

Identify the emerging trend

Our sample data

We are working on a tweet dataset as shown in the image above. These tweets are all made by Disney from 2010 to January 2021. Topics vary from movies or animations related announcements, and celebrations to new releases.

Most frequent topics in the last week

We follow the four above-mentioned steps (i.e. upload, AI Clustering and set up the Explorer page). Tweets are clustered into 10 categories and we label the categories based on their content on the Explorer app and under Cluster explorer view. To identify the most frequent trends, we will need to:

  1. filter the data by date and only visualize the last week; read more at Search and filters.
  2. sort data descendingly based on the frequency
    As can be seen in the image below the top three emerging trends were "New on Disney Plus", "Anniversaries" and "All magic".

Relevance AI - Last week's most frequent topics

Most liked topics in the last 30 days

Another way of identifying the emerging trends in a tweeter dataset is to study the average number of likes in the recent period. We will need to

  1. filter the data by date and only visualize the last month; read more about date filtering at Search and filters.
  2. define metrics on the average number of likes; read more about defining metrics using Metrics and aggregation.
  3. sort data descending based on the metric
    As can be seen in the image below the top three emerging trends from Dec 23rd 2021 to Jan 23rd 2022 were "Happy new year", "Anniversaries" and "Marvel Studio".

Relevance AI - Last month's most-liked tweets

Note: We are using the Cluster explorer view on the Explorer app.