2 Replies Latest reply on Sep 18, 2017 8:28 AM by Robert Kern

    Recommended Knowlege - Query? ( and influence of pinning ).

    Robert Kern
      Share This:

      Dear Community,

       

      we are trying to found out the exact query / logic of how knowledge articles are recommended for Incidents.

      So far we have not been able to find it in a documentation ... does anyone has a hint?

       

      Also bonus question :-) ... does pining ( save to ticket function ) a knowledge article influence this somehow?

       

       

      Thank you in advance,

      Robert

        • 1. Re: Recommended Knowlege - Query? ( and influence of pinning ).
          Karl-Anders Falk

          Hi Robert,

          I think this article is giving answers to many questions you might have ran into around how search works.

           

          Ash Hall is one of our most experienced consultants in this areas and has over the years built up a deep knowledge in how knowledge search works and how it can be tailored.

           

          The article also contains links to some documentation.

          Ash's Guide to Knowledge Management in Smart IT & RKM - Searching, Optimization, Process, Indexing & Images

           

           

          Most of the article refer to remedy Knowledge Management, but as MyIT/Digital WorkPlace and SmartIT is based on Remedy it is also valid for those products. I also suggest reading the MyIT and Digital workplace documentation about FTS (Full Text Search) and Search to understand how it is implemented and what documents and fields are searched depending on where you search.  Most details about how to configure FTS can be found in the Remedt AR system documentation, and if you want to go really deep, I am sure you can google Lucene Search Engine to get into the weeds of the algorithms.

           

          I hope this helps

           

          /KAF

           

          Karl-Anders Falk

          Ondemand Business Relationship Manager

          1 of 1 people found this helpful
          • 2. Re: Recommended Knowlege - Query? ( and influence of pinning ).
            Robert Kern

            Dear Karl-Anders ... thank you for the link to the KB Article.

             

            In the article it states about basic and advanced search ... I am not totally sure if "what fields are being searched" also relate to the "Recommended Knowledge" feature within SmartIT (to beginn with I do not know if ITSM as such has a similar feature of suggested Knowledge Articles - withouth directly searching for them).

             

            Also in the article it states that "Usage" influences the ranking ... however as it seems the incidents used itself are not included in the searched content (only the KB article itself - and then only "indirectly" the usage counter comes into play.

             

            In terms of SmartIT the "Usage Counter" possibly is "only" the "Save & Resolve" function ... also only this increases what is called in SmartIT "Linked Incidents" ... for example "Save to Ticket" function (also known as "pinning") does not increase the "Linked Tickets" counter within SmartIT ... but does it also increase the "Usage Counter"? (like are both the same or different).

             

            So quite something to chew on :-) ... anyway, I wish the documentation would be a bit more technically detailed.

             

            SmartIT

            [...] Relating the knowledge resource to a ticket will update the usage % of the resource increasing its credibility and use for resolving a request. It will also increase the resource's relevance to similar type of requests, placing it higher in the list when displaying Recommended Knowledge for similar requests [...]

             

            Article

            [...] Whenever a Knowledge Article is used (achieved by clicking the ‘Use’ button at the base of an article), a ‘Use’ counter is incremented. The same applies to whenever an article is opened and then closed using the ‘Close’ button, a ‘View’ counter is incremented [...]

             

             

            Best greetings,

            Robert