Uploaded image for project: 'Confluence Data Center'
  1. Confluence Data Center
  2. CONFSERVER-9081

Improve search result ordering , giving prioity to pages that have the all the search works found closer together or as a phrase

XMLWordPrintable

    • We collect Confluence feedback from various sources, and we evaluate what we've collected when planning our product roadmap. To understand how this piece of feedback will be reviewed, see our Implementation of New Features Policy.

      NOTE: This suggestion is for Confluence Server. Using Confluence Cloud? See the corresponding suggestion.

      I sometimes find the ordering of search results a little difficult to use. In particular, results which have phrases or have matches closer together are not put high enough.

      An example:
      On our wiki I was searching for jira time tracker; there are 2 pages that have details about this little app that we've built. The one I was looking for was on page 3 of the search (see searchResults1.png). If I search with quotes "jira time tracker" then I only get the 2 pages back (see searchResults2.png).

      It would be better to prioritise this ranking of pages with phrases or matching words closer together.

        1. searchResults1.png
          searchResults1.png
          121 kB
        2. searchResults2.png
          searchResults2.png
          102 kB

            [CONFSERVER-9081] Improve search result ordering , giving prioity to pages that have the all the search works found closer together or as a phrase

            Paul Curren added a comment -

            Hi Alexey.
            This shouldn't be impossible to do. Possibilities include using the SpanTermQuery.

            Basically, tokens in Lucene have positional information and so there are a few different approaches we could take to implement this.

            In saying that, we unfortunately were not able to address this ranking improvement in the scope of Confluence 2.9, but will re-visit at some future time.

            Thanks.

            Paul Curren added a comment - Hi Alexey. This shouldn't be impossible to do. Possibilities include using the SpanTermQuery . Basically, tokens in Lucene have positional information and so there are a few different approaches we could take to implement this. In saying that, we unfortunately were not able to address this ranking improvement in the scope of Confluence 2.9, but will re-visit at some future time. Thanks.

            Alexey Alexey added a comment -

            I guess it is impossible to do in Lucene
            http://wiki.apache.org/lucene-java/LuceneFAQ#head-29e758aaaf252d77fe16e5e5f4b20fdb657f8059

            Alexey Alexey added a comment - I guess it is impossible to do in Lucene http://wiki.apache.org/lucene-java/LuceneFAQ#head-29e758aaaf252d77fe16e5e5f4b20fdb657f8059

            Paul Curren added a comment -

            We are going to make some minor changes to the ranking for Confluence 2.9. Hopefully we can address this problem as part of that.

            Paul Curren added a comment - We are going to make some minor changes to the ranking for Confluence 2.9. Hopefully we can address this problem as part of that.

              dave@atlassian.com dave (Inactive)
              a7240d48bceb David Ruddock
              Votes:
              10 Vote for this issue
              Watchers:
              6 Start watching this issue

                Created:
                Updated:
                Resolved: