I am working on a project where a movie review is given and I've to check whether the movie is good or bad by analyzing the review. Using Language processing I've extracted the percentage of supported words(Good) and unsupported words(Bad) in review. Now, I have to get the ratings out of 10 using these percentage.
Suppose: The product has good and bad percentages.
Good: 70% Bad: 30% or
Good: 40% Bad: 50%
and all possible combinations.
How can I make a general rating out of 10?
1: Extremely bad
.
.
5: Average (When Good:50% and Bad: 50%)
.
.
10: Extremely Good
I've tried some formulas made by myself but those were not good.
i.e (Good/bad) + 4 it can work good if Good is greater or equal to Bad.
Good(50)/Bad(50) + 4 = 5
Good (70) / bad(30) + 4 = 6
if Good is lesser than Bad then it always results in 4. Because of bad formula.
Good(30) / bad (60) + 4 = 4.5
Good(20) / bad (80) + 4 = 4.25
I need a formula or method. Please help.
How about
$$\text{Rating} = 10\cdot \frac{\text{Good}}{\text{Good} + \text{Bad}}?$$
Examples.
$\text{Good} = 40, \text{Bad} = 60$. Then $\text{Rating = 4}$.
$\text{Good} = 20, \text{Bad} = 60$. Then $\text{Rating = 2.5}$.
$\text{Good} = 80, \text{Bad} = 20$. Then $\text{Rating = 8}$.
The only problem is that you can get a rating of $0$ if you have $0\%$ in Good.