extra file or extra line, duplicated instances(to decrease the weights) or
duplicate feature in the same instance to increase the weights(classic
tf-idf)

Robin

On Thu, Sep 30, 2010 at 9:50 PM, Neil Ghosh <[email protected]> wrote:

> So All I have to do is add an extra file containing
>
> LABEL<TAB>problem<TAB>complaint<TAB>problemo
>
> Along with the usual training data in Bayes format ?
>
>
> On Thu, Sep 30, 2010 at 9:44 PM, Robin Anil <[email protected]> wrote:
>
>>
>>>> Or Do I have flexibility to give some other input specific to my problem
>>>> ? Such as if words like "Problem", "Complaint" etc are more likely to 
>>>> appear
>>>> in a text containing grievance.
>>>>
>>>>
>>>>
>>
>>
>>> You can provide a Weight, usually TF-IDF, that often does a good job of
>>>> factoring in the importance of words.  If you have certain sentiment words
>>>> that you think influence things one way or the other, you could consider a
>>>> weighting process that adds weight to those words, I suppose, but I would
>>>> want to experiment with that a bit.
>>>>
>>>
>>> I would first get your data in the bayes format
>> <LABEL><TAB><FEATURE1><SPACE><FEATURE2>......
>>
>> Feature can be words, or pairs of word (word1_word2) or binned numerical
>> values ( 0.1, 0.2.. etc) or enums. (SEX:MALE,  SEX:FEMALE)
>>
>> Give this as input to the classifier and get the output.
>>
>> If you need to add couple words hardcoded into the classifier. Add them as
>> a training instance. Since features are assumed to be independent in bayes.
>> it doesnt matter how you give them
>>
>>  POS<TAB>problem<TAB>complaint<TAB>problemo
>>
>>
>>
>>
>>
>>
>>
>>>
>>>
>>> On Thu, Sep 30, 2010 at 8:55 PM, Robin Anil <[email protected]>wrote:
>>>
>>>> It does that by default for all words. What else do you have in mind?
>>>>
>>>> On Thu, Sep 30, 2010 at 8:07 PM, Neil Ghosh <[email protected]>wrote:
>>>>
>>>>> Does anybody have examples/reference how to use TF-IDF weights in
>>>>> mahout
>>>>> cbayes for particular words and phrases while doing text classification
>>>>> ?
>>>>>
>>>>> --
>>>>> Thanks and Regards
>>>>> Neil
>>>>> http://neilghosh.com
>>>>>
>>>>
>>>>
>>>
>>>
>>> --
>>> Thanks and Regards
>>> Neil
>>> http://neilghosh.com
>>>
>>>
>>>
>>>
>>
>
>
> --
> Thanks and Regards
> Neil
> http://neilghosh.com
>
>
>
>

Reply via email to