NodeMapper takes only the score of 1 element from the list of values to analyze

[vakuzmin]

for (String value : schema.getValues(tableName, colName)) { if (word == null || value == null) { System.out.println("Comparing "+word+" and "+value); System.out.println("In table "+tableName+", column "+colName); } valueNodes.add(new NodeInfo("VN", tableName+"."+colName, WordSimilarity.getSimilarity(word, value, wordNet))); //add every sample value into valueNodes }

even if we have 2K of elements for the table/column pair the similarity score of only the first value is taken into account that couldn't be the best one, that happens due to the how equals/hash code functions for NodeInfo are defined