Artigo versão em Português

The topics tab is the main area where you can manage the unsupervised topics our AI is continuously discovering and organizing them into a supervised topic structure. To access this view, click on the "Topics" tab icon located in the side menu. Here is all you need to know about it.

The topics tab works with two different areas. The unsupervised topics area is located in the central area and allows you to delete topics created by the AI, check the keywords associated with any topic, analyze the numbers (volume, net sentiment), and, ultimately, view findings generated by the AI related to each topic.

The supervised topics area is located in the left-hand side of the screen and allows you to easily create topics and subtopics in order to organize the unsupervised topics. For this organization, you can align the vision inspired by structures you are already used to see in your company.

💡Pro tip #1: But what exactly do we mean by unsupervised and supervised topics? The unsupervised topics are the ones automatically discovered by the AI by combining clustering techniques and advanced analytics tools. This approach involves using machine learning algorithms to analyze the text data and identify patterns, similarities, and relationships among the words or phrases in the data. It is automatically done by the AI as soon as the dataset is posted.

The supervised topics refer to a method where the topics or subtopics in the text data are predefined or known in advance. In this approach, a human expert or analyst provides a set of predefined topics, or subtopics to the system.

.

💡Pro tip #2: You don't need to rely exclusively on the data categorization and the taxonomy structure to have your analysis. However, by using them, you can extract the maximum potential of the Birdie platform. In other words, data categorization and the taxonomy structure will enhance your experience.

The unsupervised topics area is composed by the following parts:

Topics: List of topics found by the IA.

Subtopic: If an AI Topic is classified inside a supervised topic structure, it will show here.

Net Sentiment: Of the topic found considering all extractions.

Volume: Of the number of feedback that mentions the topic.

View Findings: Read an AI summary overview of the key findings related to each topic.

💡Pro tip #1: You can check the keywords associated with each topic by clicking on the arrow before each Topic. This action will open a list of keywords associated to each topic.

💡Pro tip #2: By clicking over each topic, a screen will pop-up bringing (1) the summary of each thread in the case of conversation-type sources (Support tickets, meeting transcriptions, ...) or (2) the feedback in the case of single-feedback-type sources (NPS, Reviews, ...) associated with that specific topic.

The supervised topics area is composed by the following parts:

Product Areas: Supervised Topic Structure created according to client’s needs and directly connected to the product (These topics are tracked via Dashboards).

Other Topics: Free space to allow users to create and measure their own topics.