1
$\begingroup$

I have a dataset of 40K reddit posts in Italian, and I have a sentiment-based dictionary of 9K unique words and phrases, which classifies words into positive or negative. I would like to measure sentient per reddit post across time and I noticed that are several methods to compute sentiment per post. I am currently using the following equation:

enter image description here

I wonder if it has any obvious downsides? The main advantage of course is interpretability, which straightforward with this method as it produces a score with a theoretical scale between −100 points (extremely negative) to 100 points (extremely positive).

$\endgroup$

1 Answer 1

0
$\begingroup$

The formula:

$$\frac{\text{positive}-\text{negative}}{\text{total}}$$

is intuitive and simple. The only problem is that it assumes that all sentiment -related words are of equal strength.

For example, the words good and great are both positive but do not express sentiment of the same strength.

So one improvement would be to assign weights to sentiment words (in the dictionary) and adjust the formula to take that into account.

$$\sum_i{p_i}-\sum_j{n_j}$$

$p_i$ is weight of positive word in text $n_j$ is weight of negative word in text

$\endgroup$
3
  • 1
    $\begingroup$ This is a great suggestion! Do you know of any papers/guides that operationalize this technique in research? $\endgroup$ Apr 13, 2023 at 11:22
  • 1
    $\begingroup$ I don't have some paper handy, but you can search, I am sure it has been used/suggested allready $\endgroup$
    – Nikos M.
    Apr 13, 2023 at 11:55
  • 1
    $\begingroup$ Thanks, will do! I have noticed some useful methodology papers that operationalize this but I am trying to find an econ or polsci paper that performs this. Thanks for the suggestion! $\endgroup$ Apr 13, 2023 at 12:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.