redlink-gmbh / smarti

Smarti is a service component for conversational data.
http://dev.redlink.io/smarti
Apache License 2.0
6 stars 4 forks source link

How to deal with extracted URLs? #132

Open ruKurz opened 7 years ago

ruKurz commented 7 years ago

Currently extracted URLs are handled exactly like keywords (from UI perspective). Next to breaking the layout (https://github.com/redlink-gmbh/smarti/issues/128) questions about:

came up.

This issue has being opened in order to start a discussion about URLs from the end user perspective. How should URLs represented inside the SmartiWidget and can we use them to get a better search result ranking?

westei commented 7 years ago

As a first step I can map them into an own field in the LATCH template (used by all Information Retrieval QueryBuilder). This will allow to ignore them for now and implement a special handling later on

ruKurz commented 6 years ago

Is this issue also existing when using the conversation-search query builder?