Smart tag suggestions

Introduction

The smart tag suggestion feature enhances the tagging process with intelligent recommendations based on the content being created or edited.

By leveraging machine learning algorithms, the smart suggestion tool analyzes content patterns to suggest the most appropriate tags, ensuring that content is accurately categorized and easily discoverable.

This feature allows editors to quickly assign relevant tags without extensive manual searching.

Using the smart tag suggestion

To use the smart suggestion feature, editors log into the FORGE Content Manager. There’s no need for dedicated permissions.

On the entity detail page, the tags tab on the right side displays the Suggested tags button.

Selecting Suggested tags activates the smart suggestion feature, which displays a list of recommended tags based on the current entity's content.

By default, all suggested tags are automatically selected, but editors can unselect the tags they consider not relevant. For example, editors can unselect sports, like in the screenshot below.

Editors can ask for more suggestions by selecting Suggest more. Then, they can apply the suggestion by selecting Add tags. The chosen tags appear in the tag list like ordinary tags.

How smart suggestions work

Data processing

The smart tag suggestions feature processes and stores the text from which it infers tag within the FORGE ecosystem. Also, no sensitive data (e.g., the name of the user asking for tag suggestions) are involved in the process.

The smart tag suggestion workflow runs in three steps:

1. FORGE submits the entity text to Azure OpenAI to generate tags.

2. FORGE matches the generated tags with the existing published tags. It’s a semantic match. For example, if Azure OpenAI suggests sportsperson and FORGE already has the published tag athlete, then the suggested tag is athlete.

   When there’s no suggestion

   In the case the semantic match is empty, there’s no tag suggestion available.

3. FORGE lists the suggested tags on the entity edit page. The suggested tags are ordered by content relevance, which is how important a tag is compared to the analyzed text.

Azure OpenAI is just an option

Based on client requirements, FORGE is open to plug any other Large Language Models (LLMs).

Benefits of using smart suggestions

Implementing the smart suggestion feature offers several advantages:

• Efficiency: Editors can save time by quickly applying suggested tags rather than manually searching for them.

• Consistency: The feature promotes uniformity in tagging, which enhances the overall organization of content.

• Improved Discoverability: By using relevant tags, content becomes more accessible to users searching for specific topics.

Conclusion

The smart suggestion feature in FORGE makes it easier for editors to categorize content.

By utilizing intelligent recommendations, editors can improve the visibility and organization of their digital products, ultimately leading to a better user experience.