I Have a task on text mining.The task is to evaluate and score questions answered by all the candidates attending an interview on a particular skill.I am looking to do this using r.Please let me know how to approach this,or please share codes or examples which are similar to the task.
If you have not done a little research about text mining already, I’d suggest that you do that first.
I am also attaching a link to a PPT of a sentiment analysis using text mining project I did with a couple of my friends not so long ago. Skip the first 9 slides and you can browse through the code there. It might take a while before it starts to make some sense, but I’ll recommend you to perform those operations on your own dataset to get a better idea.
I hope this helps.
Firstly,Thank you so much for responding.
I don’t think sentimental analysis can be used for this particular column.
I need to compare the answers given by candidates against the file which has correct answers.
And i the score should be given based on how many matches are correct.
I have researched and tried sentimental analysis and preference aggregation on my dataset,but could’nt arrive at the solution.
Using Bag of words approach i will remove stopwords and will do stemming on my test data set(file which has answers given by candidates) and compare keywords with the corresponding lines in actual document.
This is my idea,need help in getting the code up and working for this.
I am not suggesting you to perform a sentimental analysis. I shared the link as it contains code snippets to remove stopwords, stemming and other pre-processing for text mining. Have a look from slide #10.
My suggestion is if you know the correct answers then you can use document similarity to score.
i.e how close the answer given is similar to the original answer