Generate a Word Cloud and Sentiment Analysis for Your Steemit Account

in #steemit6 years ago

Yesterday, I came across two Steemit tools by @egmracer01.

The first tool generates a word cloud for a given Steemit account. The word cloud that I generated for my Steemit account yesterday is below.

wordcloud.png

The second tool generates a sentiment analysis for a given Steemit account. The sentiment analysis of my Steemit account is below.

sentiment.png

Potential Use in Steemit

They are both nice ideas and actually both of them can be useful if they are embedded in Steemit profiles.

When you check someone's profile for the first time, their word cloud and sentiment analysis can give an idea about the posts of that person, much faster than reading their posts.

I can imagine that some people would like to follow accounts that have more positive sentiment. The word cloud also says a lot whether the account is relevant or not. Both of the tools would add value to the platform if embedded in Profile.

Feedback for Writers

Moreover, serious writers could analyze their word cloud to improve their use of language. If you're using the same words too often, they would stand out in the cloud and you might want to avoid them and you might want to use their synonyms instead.

Potential Use in Investing

The sentiment analysis have greater applications though. In investing, contrarian investors look for extreme positive or extreme negative sentiments in the markets. Using this tool can help automate measuring the sentiment of the market.

In order for the sentiment tool to be useful as a technical indicator, it has to be tested on real data. So, without any statistical analysis, we can't say that it's a reliable investing method.

And who knows maybe it has some room for improvement, because when I look at my sentiment analysis, it's slightly positive, but I wouldn't say that my last posts were very positive.

Maybe this tool looks at the individual words and doesn't recognize the meaning when they are used together. I don't know.

Shout out to @egmracer01

In either case, both tools have some potential to be used on the Steemit platform, so make sure you check the results for your own account via this link and check other posts of @egmracer01 in his profile.

Sort:  

Nice tools. Thanks to both of you, @egmracer01 and @bbilgin.
I have tried it and this is the result I got:

word cloud.png

sent.png

Your balance is definitely more tilted towards the positive side, which is good.

The word cloud gives a good idea about your posts as expected. It would be difficult for me to keep in my mind what each person that I follow posts about, but with this tool I can see the "big picture" of an account. I really like that.

Thanks for the shoutout. If anyone has recommendations or ideas for things to build in the future, I'm all ears.

I have noticed that the word cloud includes common non-significant words like: "Thank", "post" and "think". Maybe they should be excluded.

Actually, it's not that bad that it shows thank as your biggest word. It shows that you're a grateful person.

Moreover words like "think" and "believe" could be interesting for writers. If they pop up too much on the word cloud, the writer needs to pay attention to use them less, because they make one's writing weaker.

So, this can be a parameter for the user. Just an empty text box, with the option to populate it with common words like you mention. And the user could have the option to add extra words to be excluded. Eventually, this text box could be prepopulated with a standard list.

Yes, I agree.
Including those words was useful for me. However, if someone else wanted to know about my interests through this word cloud, they should be excluded. So:

this can be a parameter for the user. Just an empty text box, with the option to populate it with common words like you mention.

That's a good suggestion. Here's the list of words I currently have ignored: https://github.com/evanmarshall/steemit-cloud/blob/master/src/ignored-words.js

I think that you should exclude the following words too:
thank
think
please
moreover
maybe
soon
like
liked
used
already
keep
instead
example
really
give
btw
post
lot
away
around
issue

Yesterday, we had a talk about the friction to go to another website. Today, I realized that I check my steemd feed way more often than steemit feed, because there are no notifications feed on steemit and steemd acts as a notification feed.

If a decent discovery algorithm for Steemit was implemented on another domain, even on a webpage on another domain like your tool above, and it opened the Steemit posts in new tabs, I'd definitely use it as my start page for my steemit browsing.