Semantic Search for comments

All managers with access to comments can enter a search term to surface comments containing the searched keyword.

Semantic Search, on the other hand, understands the contextual meaning of a phrase, as well as relationships between words, without them explicitly being mentioned in the comment. Further advanced filtering helps speed up investigations and narrow results.

Semantic Search uses Natural Language Processing (NLP) to help you surface the most insightful and relevant comments related to a search query. It enables you to find out what's important to your employees in their own words.

Example: When searching for the word acquisition, Semantic Search would return comments also including the words buy-out, sale, purchase, and so on.

To enable Semantic Search for your organization, see Enable Semantic Search for comments.

Considerations

• The feature provides better results with longer search phrases and sentences. The semantic technology can return specific matches if provided with maximum context. Example: Searching for Efforts to support diversity and a diverse workforce yields better results than searching for just diversity.

• Semantic Search doesn't rank comments based on their sentiment, when matching the search query. Example: If you search I love my job, then I hate my job will display near the top of the results. Filter results based on their NPS, to account for sentiment and to find the positive or negative comments associated with a search.
• Semantic Search orders all available comments from most relevant to least relevant, based on the contextual meaning of your search query. Regardless if your search uses filtering, the displayed comments become less relevant the further you scroll through the results.
• The semantic search export enables you to download the first 10,000 results from the comment rankings, sorted by relevancy to the search query.
• The first iteration of Semantic Search only includes comments written in English.