"opinion mining/sentiment analysis-rapidminer5"

lina

hi!
I would appreciate your giving me any piece of information!it is really important to me!
i have created an excel file,filled with comments about a specific topic!
now,i am trying to classify these comments(in fact the comments are short sentences from various sources via the net)
into positive,neutral and negative!
how can i proceed?
please let me inform you that all comments are written in greek.i hope there is no problem with it!
since i am new to this topic i would be really grateful for any help!
thanks in advance, i am looking forward to your reply!

el_chief

see my blog vancouverdata.blogspot.com

i have a five part video series on text mining, including how to do classification (sentiment analysis) in the 5th part

good luck

neil

lina

Neil McGuigan wrote:
see my blog vancouverdata.blogspot.com i have a five part video series on text mining, including how to do classification (sentiment analysis) in the 5th part good luck neil

thank you very much,neil!i'm going to visit your blog and watch the videos!!

lina

hi!
i'm still working on opinion mining but i have few problems.
i have watched the videos from vancouverdata.btw,i found them really helpfull,thanks neil!
First of all,the language i use is greek so i want to create a text for the operator: Filter Stopword.Does anyone know how the text should be like? I've created a text like this: "word1|word2...."but unfortunately it is not recognized. Any idea, please?
Also, there is not a stem operator for my language.How can i create one as it seems to be very important?
Apart from these problems, i have followed the method which is showed in the 5th part of the video series but i also have a problem. The operator naive bayes : "cannot check whether input example set has special attribute "label""
What about this?Should i specify a label or an attribute in the file i use?Specifically, i use an excel file instead of database which is used in the video.
Sorry for the long post.
我期待你的回答和你的帮助!!

B_

Lina

Filter Stopwords by Dictionary allows you to create your own stoplist - it reads from a file that you create.

You can try using regular expressions to create a basic stemmer if the endings of Greek words are consistent for cases and gender.

Here is a simple classifer you can adapt:
http://rapid-i.com/rapidforum/index.php/topic,2993.0.html

Remove the N-Gram operator and change input to Excel. The column that contains the opinion should be set as Label in the Set Role operator.

B.

lina

thank you so much B.i do appreciate yor help!

Filter Stopwords by Dictionary allows you to create your own stoplist - it reads from a file that you create.

i'm trying to create this file but it is not recognized by RapidMiner.what should the form of this file be like?
i've tried something like : "word1|word2..." but it doesn't work!any idea about it?
regarding the classifier and the example given,i'm going to check it out and i hope i manage to classify my own documents!

B_

In Windows it's a txt file.

In rmstop.txt
one
two
three

<运营商激活= " true " class = "文本:process_document_from_data" compatibility="5.1.001" expanded="true" height="76" name="Process Documents from Data" width="90" x="313" y="165">

fritmore

lina wrote:

thank you so much B.i do appreciate yor help! Filter Stopwords by Dictionary allows you to create your own stoplist - it reads from a file that you create. i'm trying to create this file but it is not recognized by RapidMiner.what should the form of this file be like? i've tried something like : "word1|word2..." but it doesn't work!any idea about it? regarding the classifier and the example given,i'm going to check it out and i hope i manage to classify my own documents!

create an ascii file with txt or csv extension
sample of the file data structure:

attrib1,attrib2,attrib3
apple,monkey,brick
orange,monkey,stick