The nodes are scattered around the graph, we don’t know which topics belong together, the colors seem to be there just for decoration.Īlternatively, a word cloud made using InfraNodus generator also offers a set of powerful analytical tools and insights that can be used to understand the underlying text better and to tell a more effective story about the content: In InfraNodus, we can also see analytics - which topics are the most relevant in the text - and turn on the network view, which enables us to see which concepts are connected and how. What matters is the context and the traditional word cloud tools just cannot analyze that.Įven if we look at a more advanced MonkeyLearn word cloud generator, which uses machine learning under the hood, we will see that the results we obtain still lack the contextual information, even if they can show the most frequently occurring phrases within the text: MonkeyLearn’s word cloud has a bit more information, but still lacks the context.
We cannot remove the word “applause” and even if we did, we just see general terms like “people” or “freedom” or “American”, which we can see in any presidential address.
If we were to put this same text into a popular word cloud generator, TagCrowd, we’ll see something like that: A word cloud made using TagCrowdĪs you can see, this visualization lacks context. We understand that he’s talking about the demands of the time, the journey that America has to make, equal opportunities, and belief. “American” is the most influential term in the middle of the graph, connecting all the topics together. (We explain the science in the next section of this article below)īased on this approach, we can quickly see that Obama is talking about “time requires” a lot, as well as “make people believe”, “America journey complete” and “create equal man”. If the words have the same color, they belong to the same cluster, meaning that they tend to occur next to each other in this text. On this word cloud the closer the words are to each other, the more related they are in the context of the text. Take a look at this mind map of Obama’s 2013 address made using InfraNodus: Word cloud made using InfraNodus text network visualization tool - the more influential words are bigger, the words that appear more often together have the same color and are closer on the graph. You will be able to see what are the main topics in any text and, more importantly, how they relate to each other. Using InfraNodus word cloud generator, you can generate word clouds where the relevant terms will be aligned next to each other if they tend to appear in the same context. This is a problem because a text is not just a combination of words, it’s all about the relations - and this information is lost. The words will just be randomly aligned on the screen and the most frequently mentioned ones will be bigger and towards the center, while the less frequent words will be smaller on the periphery of the graph. When you generate a standard word cloud, it will normally not have any information about the context.