Download Latest Version Mechaglot_Alpha3.zip (9.5 MB)
Email in envelope

Get an email when there's a new version of Mechaglot, Calculate Semantic Similarity

Home
Name Modified Size InfoDownloads / Week
RAW DATA 2014-09-23
README.txt 2014-10-07 2.9 kB
Mechaglot_Alpha3.zip 2014-10-07 9.5 MB
Mechaglot_Alpha2.zip 2014-10-01 8.0 MB
Mechaglot_alpha.zip 2014-09-26 6.6 MB
Totals: 5 Items   24.0 MB 0
The version has been updated with the Neural Net model which has been trained with the sample-size of 4127 rows to

Correlation coefficient                  0.9978
Mean absolute error                      0.0318
Root mean squared error                  0.0347
Relative absolute error                 16.8578 %
Root relative squared error             15.4341 %
Total Number of Instances             4127    


IMPORTANT: This version is in conflict with Java 1.8, so please use 1.7 instead 
(the imported Neural Nets from a lib folder do not work with JAVA 1.8). Furthermore, I am using Eclipse Luna Release (4.4.0)


The test-run gives the following results:

---------START-----------
-Using the Neural Net model method
Similarity between the sentences:
Pete and Rob have found a dog near the station.
Pete and Rob have never found a dog near the station.
is: 1.0
--------------------
Similarity between the sentences:
Patricia found a dog near the station.
It was a dog who found Pete and Rob under the snow.
is: 0.68698204
--------------------
Similarity between the sentences:
Patricia found a dog near the station.
I am fine, thanks!
is: 0.29550347
--------------------
Similarity between the sentences:
Hello there, how are you?
I am fine, thanks!
is: 0.37528354
The overall time to compute the examples using this method was: 2333 nanoseconds.
--------------------
--------------------
-Using the RBF model method
Similarity between the sentences:
Pete and Rob have found a dog near the station.
Pete and Rob have never found a dog near the station.
is: 1.0
--------------------
Similarity between the sentences:
Patricia found a dog near the station.
It was a dog who found Pete and Rob under the snow.
is: 0.68698204
--------------------
Similarity between the sentences:
Patricia found a dog near the station.
I am fine, thanks!
is: 0.29550347
--------------------
Similarity between the sentences:
Hello there, how are you?
I am fine, thanks!
is: 0.37528354
The overall time to compute the examples using this method was: 94 nanoseconds.
--------------------
--------------------
-Using the fast method
Similarity between the sentences:
Pete and Rob have found a dog near the station.
Pete and Rob have never found a dog near the station.
is: 0.98825234
--------------------
Similarity between the sentences:
Patricia found a dog near the station.
It was a dog who found Pete and Rob under the snow.
is: 0.7043551
--------------------
Similarity between the sentences:
Patricia found a dog near the station.
I am fine, thanks!
is: 0.30880186
--------------------
Similarity between the sentences:
Hello there, how are you?
I am fine, thanks!
is: 0.38521636
The overall time to compute the examples using this method was: 33 nanoseconds.
---------END-----------


=============================
Source: README.txt, updated 2014-10-07