What needs improvement with KNIME?

One disadvantage is the data load. KNIME has limitations. It doesn't handle large datasets or a high number of records well. I haven't tried more than 10,000 to 20,000 records because the model prediction doesn't come out well with more data. The Enterprise Edition might work better, but I've only used the Community Edition. That's the only disadvantage I've encountered so far.

The current UI is primarily in English. Analyzing data in local languages might present challenges or issues.

The graphic features of KNIME need improvement, especially when working on dashboards.

The hardest part is keeping a tidy workspace because of the many nodes involved. When teaching, it would be helpful if there was more emphasis on how to group nodes effectively. For example, turning frequently used nodes into a single component can simplify things.

KNIME is not good at visualization. I would like to see NLQ (Natural language query) and automated visualizations added to KNIME.

In the last update, KNIME started hiding a lot of the nodes. It doesn't mean hiding, but you need to know what you're looking for. Before that, you had just a tree that you could click, and you could get an overview of what kind of nodes do I have. Right now, it's like you need to know which node you need, and then you can start typing, but it's actually more difficult to find them.

The main issue with KNIME is that it sometimes uses too much CPU and RAM when working with large amounts of data.

To enhance accessibility and user-friendliness, there is a need for improvements in the interface and usability of deep learning and large-scale learning languages.

The pricing needs improvement.

The most difficult part of the solution revolves around its areas concerning machine learning and deep learning. The aforementioned area can be considered for improvement.

One thing to consider is that the prebuilt nodes may not always be a perfect fit for your specific needs, although most of the time, they work quite well. However, if you encounter very complex requirements, you might need to add custom code to achieve your desired outcomes. This is an area that could use some improvement, but the advantage is that it encourages you to evaluate and minimize coding efforts. As a result, you can reduce the overall amount of coding required, which is a positive aspect of KNIME. Another area that could be improved is related to the libraries. While they are quite extensive, they might not always match your exact needs. In such cases, you might have to do some coding to tailor the solution accordingly. Therefore, one area for improvement is the flexibility of prebuilt nodes, as they may not always match complex needs perfectly. Also, enhancing clarity on what the nodes do would be beneficial. For additional features, there are a couple of things that come to mind. Firstly, it would be great to have more clarity on what each node does. Sometimes, it's not very apparent, and additional information would be helpful. Secondly, it would be beneficial to have better ways to interact with and manage nodes, enhancing the user experience. And finally, I think KNIME could improve on how easily it allows for extending functionalities with custom code. Although it's relatively straightforward now, making it even more accessible would be advantageous.

There are a few aspects that I am not entirely satisfied with. For instance, when integrating KNIME with our SAP system ERP and HANA, it's not as straightforward as expected. We need to find alternative connectors like the Teradata connector, which adds complexity. So far, I've had some problems integrating KNIME with other solutions. Thus, it could be an area of improvement.

KNIME is less secure than Alteryx. KNIME's documentation is not strong. I cannot make good documentation on a KNIME workflow like in Alteryx. Alteryx has more color options where I can put tools into different containers and write some annotations. I felt that was missing in KNIME.

The dynamic column name feature could be improved. When attempting to automate processes involving columns, such as with companies, it becomes difficult to achieve the same result when we make changes. So maybe it would be helpful to have a way to find similar column names automatically and execute a single approval, for instance. In future releases, I can suggest having an API along with tools or services. This would allow for customization since the API is already available in other versions. So, I suggest having an API for customization.

They should improve the solution's integration with other platforms. Also, they should add more functions to its server.

I downloaded KNIME myself, and it's for self-learning. It's difficult to provide input on the improvement area because it's more of self-learning. However, there are times when I am not able to do certain things. I don't know if it's because the solution doesn't allow me or if it's because of the lack of knowledge.

Though I can use KNIME in a 64-bit platform in the lab, it's missing some features. For example, from my laptop, I can use the image reader feature of KNIME. However, in the lab, the image reader node is missing. This is an area for improvement because not all nodes appear in the 64-bit system. In other systems, you get to use all nodes or features. It should be uniform in all systems, though I have no idea if it's a software problem or a corruption in the system that's in the college lab. At home, I can see and access the image reader node, but in school, that node is missing. What I want to be added to KNIME is the feature of extracting data from social media platforms, whether structured or unstructured and for that data to be automatically converted into a CSV file that I can use. I want a data cleaning and collection process from KNIME that works for social media platforms because datasets in social media can serve as business intelligence or can aid businesses. Social media is the trend nowadays, so I want a KNIME node to analyze data from social media platforms.

One area that could be improved is increasing awareness and adoption of KNIME among organizations. Despite its capabilities, it is not as well-known as other tools. The advertising and marketing efforts to reach out to companies and universities have not been very successful.

I want to access more of the popular deep learning libraries or frameworks as they have in other solutions, such as TensorFlow and PyTorch. There should be better integration, and a better way to use those libraries because they are very popular. Our company uses PyTorch and TensorFlow regularly to solve many computer vision problems. We expect better integration of those two tools. KNIME could improve when it comes to large data markets.

For now, the license is quite expensive for us and it would be helpful if that was reduced.

KNIME can improve by adding more automation tools in the query, similar to UiPath or Blue Prism. It would make the data collection and cleanup duties more versatile.

It's very general in terms of architecture, and as a result, it doesn't support efficient running. That is, the speed needs to be improved. The scalability needs to be improved as well.

There are some parameters that I would like to have at a bigger scale. The upper limit of one node that tries to find spots or areas in photos was too small for us. It would need to be bigger.

KNIME's licensing and data management aren't as straightforward relative to Alteryx. Alteryx's tools are more sophisticated, so you need fewer to use it compared to KNIME. I think tab implementation could be easier, too.

So far, I haven't had problems with it, so I haven't really thought about room for improvement. It's so much better than many other things. It's useful in that you can at least get people who are pretty averse to programming to start thinking about putting something into a program of any kind, because they can see what's happening. It's visual. It's codeless. For some purposes, I'd want to add Python or R, but I haven't had to do that so far, so I haven't seen the shortcomings of it. There must be some. All software has shortcomings, but I haven't recognized any myself. Not just for KNIME, but generally for software and analyzing data, I would welcome facilities for analyzing different sorts of scale data like Likert scales, Thurstone scales, magnitude ratio scales, and Guttman scales, which I don't use myself. I use both Thurstone scales and magnitude ratio scales quite a bit, and they're very powerful. But I've always had to do all the analysis myself in some simple code. I don't think that's provided. You could probably include it in KNIME, but I haven't tried to do it. If it just said, "Analyze scales," and you'd choose which sort of scale you want to analyze and it gave you the options of normalizing or reversing or whatever it happens to be, that would be helpful. There are lots of simple functions that you want to apply to scales, which would be useful in any software, including KNIME.

I would prefer to have more connectivity. The user documentation is insufficient. I would like to see more enterprise level application. There are high end features which should appear, the MLOps platform being one. This feature is key. There should be better connectivity to such platforms as AWS and SageMaker, as we rely heavily on AWS in RL. For certain South Asian markets, we plan to go with Azure, so it is important to have connectivity to both of the major clouds. There should be AI machine learning based algorithms. Such features should be available out of the box with good precision.

An improvement which can universally be made to products is to make them more simple. Code-less products are simplified. Both RapidMiner and KNIME should be made easier to use in the field of deep learning. While KNIME has all the requisite features, there is a shift from coding to programming with virtual language. It is only a process of making one's solution easier to use.

The user interface could be a bit better. It's currently very dated. They should look at other vendors like Alteryx that are more user friendly and modern. From a systems point of view, the tool is not completely user friendly. Users tell us they would like to do their own analytics and find it difficult to accomplish without the help of a technical service. You need to be a bit more knowledgeable in order to handle the solution. It's not difficult, it's just more technical than other options.

From the point of view of the interface, they can do a little bit better.

There are a lot of tools in the product and it would help if they were grouped into classes where you can select a function, rather than a specific tool. This would make workflow development faster because several tools could be used together, based on the function that is chosen. Each would complete one of the constituents of the task.

I think some of the online training content could be better, although I have been able to find all of the information. At times they're quite lengthy, and for me to go through everything and then get a resolution takes a good amount of time. They could have a more structured node-wise training model, where I can simply get into it. For example, if I need to understand a node to create pivot tables, I have to go through the training mode to understand what the functionality is. If they had a more structured training model it would be very helpful. It would be helpful to launch more certification programs online. There could be better marketing. The awareness of Knime is limited, especially for small organizations. When you compare with PowerBI, there is a lot of active marketing put into their product, also, having Microsoft associated with it is an advantage. They have to step up on the marketing aspect and the ability to digitize using Knime. Many are aware of other tools such as PowerBI. In the next release, I would like to see the certification available for active users. Also the costing aspect of the certification, there could be more local impact time zone programs with a bit of costing dissension to encourage more active users in Knime who can then move to the server version in their organization.

It is difficult to say at this time, as I am not using the latest version. I have noticed that I don't have the latest modules that were added, such as ML. In my opinion, there's one thing lacking. As far as algorithm notes go, it would be handy if category algorithms of C4 or C4.5 could be set with a checkbox or something like that. Once you go to the forum or the documentation, to see how to implement a C4.5, you mark the checkbox, and only then it would be content for C4 or C4.5. The documentation is lacking and it could be better. It's a community-driven product but there are a few crucial models missing such as ANOVA and MANOVA.

The user interface could be a little bit more comfortable. The usability, in general, could be much better. The predefined workflows could use a bit of improvement.

We are worried about the performance when it comes to using a lot of data that has many rows and columns. On the server-side, we are not sure whether KNIME can manage or handle large amounts of data without issue. It looks like it will easily work for small datasets but we are concerned about performance as the volume increases. KNIME needs to provide more documentation and training materials, including webinars or online seminars. At this time, it is not sufficient when compared to some other vendors. The user interface needs to be improved because it looks quite messy and I am not very comfortable using it.

I had some difficulty connecting to servers. It asked me to set something up on my server and it asked me for a code that I needed to generate on the server. There were several steps that I messed up. I followed all of the instructions but I couldn't manage it at all. I followed the directions in several forums to find out the problem. There should be better documentation and the steps should be easier.

The areas that I feel need improvement are: * It needs support for a joiner node to have three outputs (left unmatched, matched, right unmatched), as competitors do (have not checked 2019/20 releases). * I need the ability to add additional comparison conditions to a join. For example, in SQL you can specify only rows with a date fitting within a date range from the joined file. At the moment in KNIME, you should allow a join explosion to take place and filter what you need later, but sometimes the output becomes too big. * It would be helpful to have more examples of Java code for nodes, like Java Snippet. * I would like to have this solution show row counts on canvas, as it would improve the control and speed to build the workflow. * The pseudo-code types could be rationalised into one (e.g. only Java). * I would like to see better web scraping because every time I tried, it was not up to par, although you can use Python script.

One thing that I found was that in the open-source version of the KNIME analytics platform, we see difficulties in scheduling jobs. If the scheduler could be updated in the open-source version, the software will be easier to schedule properly and to use efficiently. The second time that I faced difficulty using KNIME was with data processing time. When we use large chunks of data for local processing, the processing is very slow. We do not want to move these big data often. For me, it seemed that moving one gigabyte of data went very slowly. So, the second thing that I would really like to see is a better ability to handle large amounts of data locally with KNIME in an efficient manner. The third area that might be improved is that when we have a large amount of data — let's say like five gigabytes — then there is one panel completely ignored. The impact of that on the results of our data processing is not good. So I would really like to see the load balancing and the overall processing time substantially reduced. So the things I would most like to see are the ability to handle large amounts of data and improved performance in processing.

It needs more examples, use cases, and MOOC to learn, especially with respect to the algorithms and how to practically create a flow from end-to-end. The learning curve is steep.

They could add more detailed examples of the functionality of every node, how it works and how we can use it, to make things easier at the beginning.