first edit of readme

README.md

@@ -1 +1,7 @@
-# PerceptualAttributes
+# Perceptual Attributes
+The goal of this work is to use distinct tools to define and display groups of people in user generated content. 
+This is, to find groups that by analyzing their reviews on an entity, (such as a movie, product or service) agree on the importance of certain features. 
+Features can be extracted explicitly by use of a Natural Language Processing and Sentiment Analysis approach, or in a more abstract way such as Vector Embeddings (doc2vec). 
+Then the groups are generated via clustering algorithms, possibly Subspace Clustering, to deal with the high-dimentionality of the data.
+The parameters of the algorithms should then be finely tuned by a domain expert to assure that the groups represent the actual perceptual agreement of users for the given entity.
+Finaly, this groups should be shown in a manner similar to data cubes, with drill-down, slicing and dicing capabilities, etc.; with the end goal of capturing the information contained in many reviews in a useful way.

Reviews_Python/OpenSubspace/README.txt

@@ -0,0 +1,19 @@
+- use run.bat/run.sh for Weka Explorer
+
+- command line (e.g. Proclus):
+java -Xmx1024m -cp i9-weka.jar;weka.jar;i9-subspace.jar;Jama.jar;jsc.jar;commons-math-1.1.jar;vecmath.jar;j3dcore.jar;j3dutils.jar weka.subspaceClusterer.Proclus
+hint: list all parameters with -h, set input file through -t 
+
+java -Xmx1024m -cp i9-weka.jar;weka.jar;i9-subspace.jar;Jama.jar;jsc.jar;commons-math-1.1.jar;vecmath.jar;j3dcore.jar;j3dutils.jar weka.subspaceClusterer.Proclus -t data\Databases\synth_dbsizescale\S1500.arff > output\proclus_S1500.txt
+
+- From within Java use 
+
+SubspaceClusterer clusterer = new .... // e.g. new Proclus();
+String[] options = ..... //set options, especially input file e.g. -t c:\data\pendigits.arff
+String eval = SubspaceClusterEvaluation.evaluateClusterer(clusterer, options);
+
+//Access list of clusters
+List<Cluster> cluster = clusterer.getSubspaceClustering();
+
+//Print evaluation/cluster results (different measures set through option -m)
+System.out.println(eval);

Reviews_Python/OpenSubspace/Readme.md

@@ -0,0 +1 @@
+Readme

Reviews_Python/OpenSubspace/run.bat

@@ -0,0 +1 @@
+java -Xmx2048m -cp i9-weka.jar;weka.jar;i9-subspace.jar;Jama.jar;jsc.jar;commons-math-1.1.jar;vecmath.jar;j3dcore.jar;j3dutils.jar weka.gui.explorer.Explorer

Reviews_Python/OpenSubspace/run.sh

@@ -0,0 +1,2 @@
+#!/bin/sh
+java -Xmx1024m -cp i9-weka.jar:weka.jar:i9-subspace.jar:Jama.jar:jsc.jar:commons-math-1.1.jar:vecmath.jar:j3dcore.jar:j3dutils.jar weka.gui.explorer.Explorer

Reviews_Python/OpenSubspace/run2.sh

@@ -0,0 +1 @@
+java -Xmx1024m -cp i9-weka.jar:weka.jar:i9-subspace.jar:Jama.jar:jsc.jar:commons-math-1.1.jar:vecmath.jar:j3dcore.jar:j3dutils.jar weka.subspaceClusterer.Proclus -t pendigits.arff -M Entropy:F1Measure:Coverage -c last

Reviews_Python/Readme.md

@@ -0,0 +1 @@
+Readme