Use Cases
Dynamic search
Giving more specific, natural answers with the help of AI.
Conventional searches on websites, online shops or even desktop applications are mostly based on pure comparison of characters. The intention of the user is not considered, which can lead to poor search results.
By using machine learning-based NLUs (Natural Language Understanding) in the Ubitec Bot Framework, qualitatively better search results can be achieved without having to set up a completely new search system.
The analysis of the user input on the basis of aspects of natural language recognizes, for example. Filters, categories, brand names and other properties of your products or documents to enable a better and more natural search.
Search entries that do not actually achieve a hit in the search index can also be used to identify the intention of the user with the help of the Ubitec Bot Framework, and thus, for example, also make the delivery status or order costs in a web shop inquiries via the search input field .
Features
- Reduce results
-
Dynamic searches make it possible to display user-specific results depending on what the user entered.
- Step by step to the result
-
Depending on how many "variables" are entered in a search query, the chatbot can ask specific questions to improve the result.
- Improve third-party systems
-
As a rule, a search system does not have to be rebuilt. Normally it is sufficient to show the search algorithm the way using NLU.
- Dynamic conversation
-
During the process of making an appointment, the chatbot recognizes special requirements (e.g. documents) and notifies people accordingly.