“使用跨越距离两个文档相似性”

asafwat

1. 我正在使用rapidminer比较similarity between two text fields in two sheets in same excel file using cross distance, as i want to compart one request will all referernce to return the similarity value by cosine similarity, the problem is the distance returns as question mark '?' without knowing the reason

<参数键= value =“transform_to瞧wer case"/>



<参数键= "目录" value = " /用户/电脑/ Downloads/WordNet-3.0/dict"/>


































< portSpacing port="source_document" spacing="0"/>
< portSpacing port="sink_document 1" spacing="0"/>
< portSpacing port="sink_document 2" spacing="0"/>















































































































<参数键= value =“transform_to瞧wer case"/>



<参数键= "目录" value = " /用户/电脑/ Downloads/WordNet-3.0/dict"/>


































< portSpacing port="source_document" spacing="0"/>
< portSpacing port="sink_document 1" spacing="0"/>
< portSpacing port="sink_document 2" spacing="0"/>







































































< portSpacing port="source_input 1" spacing="0"/>
< portSpacing port="sink_result 1" spacing="0"/>
< portSpacing port="sink_result 2" spacing="0"/>
Read Requirements Document
Read Requirements Change Requests

lionelderkrikor

Hi@asafwat,

I think I found elements of answers (now calculated distances/similarities have numerical values) :

In the documentation of theCross-Distancesoperator it is said that :

"Please note that both input ExampleSets should havethe same attributesand in the same order".

So you have to use aSuperset(cf documentation of this operator) operator to feed thereqandrefports of theCross-Distancesoperator with 2 datasets which have strictly the same attributes.

Moreover, I made some modifications in your process :

- in theProcess Documents from Dataoperators : vector creation ->Term Occurences.

- in theTokenize运营商:模式- >non letters.

The process :

<参数键= "目录" value = " /用户/电脑/ Downloads/WordNet-3.0/dict"/>



















< portSpacing port="source_document" spacing="0"/>
< portSpacing port="sink_document 1" spacing="0"/>
< portSpacing port="sink_document 2" spacing="0"/>








































<参数键= "目录" value = " /用户/电脑/ Downloads/WordNet-3.0/dict"/>

















< portSpacing port="source_document" spacing="0"/>
< portSpacing port="sink_document 1" spacing="0"/>
< portSpacing port="sink_document 2" spacing="0"/>







































< portSpacing port="source_input 1" spacing="0"/>
< portSpacing port="sink_result 1" spacing="0"/>
< portSpacing port="sink_result 2" spacing="0"/>
Read Requirements Document
Read Requirements Change Requests

I hope it helps,

Regards,

Lionel

lionelderkrikor

Hi (one more time ...)@asafwat,

Just a (last ?) little advice, you don't need to specify that an attribute is "regular" in theSet Roleoperator :

By default, RapidMiner set automatically an attribute as "regular"...

Regards,

Lionel

lionelderkrikor

Hi@asafwat,

Are your attributes "numerical" ?

Can you share your dataset(s) in order we can reproduce what you observe ?

Regards,

Lionel

asafwat

Sure, here is it, i have changed it to csv in order to attach it

lionelderkrikor

Hi again@asafwat,

I have difficulties with your CSV file, can you send me your original Excel file by :

- zipping it, then, attaching it to this post

- sending your Excel file on Google Drive and then copy and share the link here in the forum

Regards,

Lionel

lionelderkrikor

再次你好(再一次)@asafwat,

Can you send me your Wordnet dictionnary too (by zipping it for example).

Regards,

Lionel

asafwat

@lionelderkrikorwooow it works, great efforts, really you made my day. much apperciated

Thanks a lot