It was [September 2018](https://hive.blog/contest/@abh12345/positive-phases-commentators-contest-win-a-month-long-1000-steem-power-delegation) when I last did something like this and it's time to dust of some code and find out who's been positive in their commentary on Hive.

<center>
![image.png](https://images.hive.blog/DQmeBmG5Wiy55AWMPG9ijBP5quUTptn5kV7fCCFdW49CTQa/image.png)
<sub><b>[source](https://hipwallpaper.com/view/YnfoeZ)</b></b>
</center>

The data I'm looking at are comments made from the 20th March 2020, 3 pm UTC right through until today.  You will need to ask for a chart to be in the running for the prize, which is 2.5k HP delegated to an account of your choice (yourself is fine!) for a month.

---------------

#### Text Mining / Text Analytics 
####

For a while now, I've been trying to surface various data and information using text held in posts/comments on Hive.  The approach has been quite straight-forward, it works (sort of), but is only really scratching the surface of what could be done with a bit more time and programming competency. 

Recently, I was passed [this link](https://www.datacamp.com/community/tutorials/text-analytics-beginners-nltk) by someone who I think can help me do more in depth and cool stuff in the future using the methods discussed in the article.

One of the text mining elements discussed is Sentiment Analysis.

> Quantifying users content, idea, belief, and opinion is known as sentiment analysis. User's online post, blogs, tweets, feedback of product helps business people to the target audience and innovate in products and services. Sentiment analysis helps in understanding people in a better and more accurate way. It is not only limited to marketing, but it can also be utilized in politics, research, and security. [source](https://www.datacamp.com/community/tutorials/text-analytics-beginners-nltk)

An off the top of my head example would be to use a sentiment analysis method on with all posts with say, #leofinance and #hive, or #leofinance and #bitcoin.

Would it be possible to collect this information from post/comments and find out how, in a given week, those discussing Hive or Bitcoin on leofinance feel about these coins?

No clue, but at some point i'd like to give it a go and find out.  And I wonder how it would compare to coingecko's feature that records user sentiment towards a token:

<center>
![image.png](https://images.hive.blog/DQmbR8RaoHkYToDQ6mq4S2NCMcwfHuf5qbzf6Mm1gryATK2/image.png)
<sub><b>[source](https://www.coingecko.com/en/coins/hive)</b></b>
</center>

-----------------------

Anyway, I'm not there with that yet, and so it's back to some dirty SQL script to find out who, based on my selection of words and phrases, has been positive in their comments on Hive since it arrived late March.  The words/phrases checked far are:

- I love this
- very kind
- much appreciated
- thank you so much
- fantastic
- well done
- thank you very much
- you are very kind
- hive on!
- I applaud you
- you rock
- excellent!
- i am very happy
- this is awesome
- my pleasure
- you are welcome
- congratulations
- a lovely
- great!


My results:


![image.png](https://images.hive.blog/DQmSepdnatU2ot4vNTsbLeCqsyXPyUdrmoLe9QXfUb1f2Zf/image.png)


And a grand total of 155, which I'm sure some of you will be able top.  Those who are verbose have an advantage, but do their comments contain enough positive words/phrases?

The person who requests a chart and has highest total wins a 2.5k HP delegation for a month.

Cheers,

Asher