Our newest Machine Learning Enhancement: Nutrient outlier detection
With Release 6.11 we will make use of our most recent Machine Learning Feature, a detection methodology for outliers within nutritional information of trade items classified within the GPC for Food. The Machine Learning makes use of the huge database of items and compares the current item opened in the editor or selected in the hit list to similar items in the database and so is able to detect patterns and provide hints regarding values related to nutrients that maybe unusual.
The scope of checks within WS Publishing covers the following anomalies regarding nutritional information:
- Check of items weather the basis quantity for the nutritional information is based on 100 ml or 100 gr depending on the usually applied basis for similar items
- Check if specific nutrients have been entered unexpectedly or if nutrients are missing which are usually given for similar items
- Check if a nutrient precision has been given that seems to be unusual compared to similar items
- Check if the values for certain nutrients are unexpectedly high or low or even untypical for that kind of item.
- Check if the the overall contained fats and carbohydrates matches to the sum of their sub-elements (e.g. carbohydrates are separated to sugars, starch, etc).
Clone current item and add it as part of the hierarchy
For sure you are also aware of it: in many cases a great amount of attributes and their values stay the same across all hierarchy levels, e.g. nutritional information or brand name, just to name two of them. But when adding a packaging item to a base level all of these information has to be re-entered.
With a new functionality in the hierarchy editor we offer you the possibility, to clone the current item with all of it’s attributes (of course the GTIN is excluded) and add it as under- or overlying element of the hierarchy. By doing this, the arduous manual input of the redundant data can be skipped and you have more time for the essential things.
New combination options for search criteria in Smart Search
We’ve implemented some adjustments to the smart search in WS Publishing.
In the previous option you were only able to select one of the many criteria of the smart search, but using the smart search with release 6.11 offers you the possibility to combine each search criteria with AND. Simply set your marks on the correct radio options and you are able to search e.g for all published items to a specific retailer which have CIC state Received.
In addition to that we’ve made some smaller adjustments to the Smart Search regarding layout and usability.
M2M Enhancements for Digital Asset Management
When providing trade item information as a hybrid supplier you are now able to enrich these M2M items within WS Publishing with digital asset information. Prior to this change, every update to your items via M2M would have wiped the relations to digital assets within WS Publishing.
Now the workflow is pretty simple:
- Send your items via M2M
- Log into WS Publishing, upload and link your digital assets to the item and sent it out.
- Linked digital assets will be kept with your trade items, independent of the digital asset information contained in the M2M CIN.
More power for the search due to most recent Elastic Search 6.4
With release 6.11 we’ve updated our underlying search index engine Elastic Search to the most recent version 6.4. What does this mean in detail?
- More performance for every search you trigger on the system
- Search for language dependent attributes now works perfectly smooth, that means your search will deliver results where you search term is either contained in the englisch attribute value or the German one. Or any other language you’ve filled in.
- Search for trade item dependent values is now also be possible. That means you will get all results where your search term is in any of the trading partner dependent values.
Elastic Search 6.4 … To boldly search what no one has searched before…
BMS 3.1.7 Support for our customers
Release 6.11 provides support for the Business Message Standard 3.1.7 to fulfill the requirements of our customers
- Data Model Changes
- New Validation Rules
- Modification to existing validation rules
- Support of Global Product Classification Update 11/2018
- Support of Target Market Codes D_A, 001 and 097