Happy Wednesday, folks! The last “Just KNIME It!” challenge of the season is out now!
Thank you so much to all who participated! And if you still haven’t experienced our weekly series of data puzzles, no problem: consider this email as a last call for the season.
After cleaning and preparing your office equipment data, it’s finally time to put it to use. The management team now wants a dashboard to explore product options, compare prices, and uncover insights from customer reviews — all in one place. You should build a product intelligence and review analytics data app in KNIME. Starting from data on products and reviews, create an interactive experience where users can browse categories, find similar products by price and rating, visualize review trends over time, extract key phrases from customer feedback, and even map review activity by country.
Here is the challenge. Let’s use this thread to post our solutions to it, which should be uploaded to your public KNIME Hub spaces with tag JKISeason4-30 .
Need help with tags? To add tag JKISeason4-30 to your workflow, go to the description panel in KNIME Analytics Platform, click the pencil to edit it, and you will see the option for adding tags right there. Let us know if you have any problems!
As I do not have much time I prepared what I could (I assumed what the challange will be, sadly there were really lot of things I have prepared but I had to delete, but that’s life )
So for the visualization. It has several parts, I try to logically group that:
In that view I have prepared a heatmap, where we can see the brands and the count of the review of differenc scores (5,4,3,2,1). So we can see what were the distribution of the scores in one brand
In the Geo analysis I put the countries to the map (colored by review count) and it has a table with the data on the map
There is option to filter the data put onto the map, you can select your filters on the left side, then click refresh and the whole map refreshes (it is really good in theory but it’s really slow, rendering the map)
In this visual and text I tried with some ML (linear regression so not a model you should throw you hat away ) saying if the brand is going up, or down, based on month coefficient
So that was the visualization part. Then the reporting. I sadly noticed that the filter visualizations couldn’t be put into a pdf. So I changed them to texts. I just prepared the PDF to trend analysis. As with all the filtering it would create a pdf with 867 432 pages long (all the combinations of the filters) and I didn’t feel if that is the scope for now
Altogether I really really enjoyed this season. I think I could really sharpen my visualization skills in KNIME (with nearly all challanges), I get a grasp of network mining, deeped dive into ML models I didn’t know before. So all in all it was a really great and thank you for everyone. I really enjoy Just KNIME It!
Find herewith my Submission: so the challenges re exhaustive ones at end… Ending up the season which is the 4th one consecutive with participation in more then 125 + challenges .. Every challenge and season creates more curiosity to keep pushing and need to learn more to brush up in the vast area.. Kudos to Knime team who really tries their best to bring the new flavour to the challenge and season. JKISeason4-30 – KNIME Community Hub
Did not have the time to run all the activities. Ran a similarity of the products so that the final recommendation presents the pareto ranked best choice and the most similar product based on the product description.
Here is my solution to this week’s and final challenge for Just KNIMT It Season 4. JKISeason4-30
This challenge is bringing all topics that we have learned from each of the previous challenges in Season 4. I tried to address the objectives in this challenge and if not, it was educational and fun to at least try.
I leveraged the data preparation and transformation we had done in challenge 28 and 29 so that part of the work was straightforward to leverage in this challenge.
Building the interactive dashboard as pre-scribed in this challenge was a challenge for me and luckily I had the L4-DA Data Analytics and Visualization 03 workflow to leverage.
I was able to build the category selection where the end user can also see the product ID (or ASIN).
With the ASIN, the end user can now view similar products and can also view a line chart for product reviews over time and a tag cloud based on the reviews for the similar products.
Adding the geospatial context for the Review Country aspect of the similar products required me to think about how to integrate into the dashboard.
I initially tried downloading the spatial data but that was a massive dataset so abandoned that idea.
Luckily, KNIME provides the OSM Boundary Map node that will give you the geospatial data for a given OSM location (even country). By integrating the Table Row to Variable node I was able to link to the OSM Boundary Map node based on the selection for each product so that the geospatial data changes.
Here is also the PDF report of the dashboard based on the selections the end-users makes.
the only drawback of the PDF report is that the geospatial view is not supported.
What a great way to end the season and I learned so much from all the participants and how they solved these challenges.
Cheers, Happy Holidays and Happy New Year.
Difficult one for me…I did not succeed with the pdf, both the tag cloud and geospatial view do not show up. You can find my workflow here. I will certainly participate again in the next season. Kudos to everyone .
Happy Wednesday, folks! The last “Just KNIME It!” challenge of the season is out now! 
After cleaning and preparing your office equipment data, it’s finally time to put it to use. The management team now wants a dashboard to explore product options, compare prices, and uncover insights from customer reviews — all in one place. 
You should build a product intelligence and review analytics data app in KNIME. Starting from data on products and reviews, create an interactive experience where users can browse categories, find similar products by price and rating, visualize review trends over time, extract key phrases from customer feedback, and even map review activity by country. 
Need help with tags? To add tag JKISeason4-30 to your workflow, go to the description panel in KNIME Analytics Platform, click the pencil to edit it, and you will see the option for adding tags right there. 
Let us know if you have any problems!
)
