No models other than naiveBayes work with text categorization

39 views Asked by At

I have a text classification problem with a text feature 'DESCRIPTION' and output variable 'TYPE', which has three different values.

I perform the preprocessing steps and then run the model but I get the following error when I run the confusion matrix (at the end of the code).

Error in confusionMatrix.default(clothing_reviews_test$TYPE, as.factor(pred2)) : the data cannot have more levels than the reference

What is the problem in the code which cause this error?
 
Code is below


  d=read.csv("SONAR_RULES.csv", stringsAsFactors = TRUE)
  d$DESCRIPTION= as.character(d$DESCRIPTION)
  
  d= d[, !names(d) %in% c("REMEDIATION_GAP_MULT", "REMEDIATION_FUNCTION", "REMEDIATION_BASE_EFFORT") ]
  d$REMEDIATION_FUNCTION=NULL
  d$DEF_REMEDIATION_GAP_MULT=NULL
  d$REMEDIATION_BASE_EFFORT=NULL
  glimpse(d)
  
  
  p= function(x) {sum(is.na(x))/length(x)*100}  ## checking missing values % in columns
  apply(clothing_reviews_train, 2, p)   ## checking missing values % in columns
  
head(d)
  set.seed(42)
  idx <- createDataPartition(d$TYPE, 
                             p = 0.7, 
                             list = FALSE, 
                             times = 1)
  clothing_reviews_train <- d[ idx,]
  clothing_reviews_test  <- d[-idx,]
  
  stem_tokenizer <- function(x) {
    lapply(word_tokenizer(x), 
           SnowballC::wordStem, 
           language = "en")
  }
  stop_words = tm::stopwords(kind = "en")
  # create prunded vocabulary
  vocab_train <- itoken(clothing_reviews_train$DESCRIPTION, 
                        preprocess_function = tolower, 
                        tokenizer = stem_tokenizer,
                        progressbar = FALSE)
  
  v <- create_vocabulary(vocab_train, 
                         stopwords = stop_words)
  
  pruned_vocab <- prune_vocabulary(v, 
                                   doc_proportion_max = 0.99, 
                                   doc_proportion_min = 0.01)
  
  vectorizer_train <- vocab_vectorizer(pruned_vocab)

  # preprocessing function
  create_dtm_mat <- function(text, vectorizer = vectorizer_train) {
    
    vocab <- itoken(text, 
                    preprocess_function = tolower, 
                    tokenizer = stem_tokenizer,
                    progressbar = FALSE)
    
    dtm <- create_dtm(vocab, 
                      vectorizer = vectorizer)
    
    tfidf = TfIdf$new()
    fit_transform(dtm, tfidf)
  }
  dtm_train2 <- create_dtm_mat(clothing_reviews_train$DESCRIPTION)  

  dtm_test2 <- create_dtm_mat(clothing_reviews_test$DESCRIPTION)
  str(dtm_train2)  

  
  xgb_model2 <- xgb.train(params = list(max_depth = 10, 
                                        eta = 0.2, 
                                        objective = "binary:logistic",
                                        eval_metric = "error", nthread = 1),
                          data = xgb.DMatrix(as.matrix(dtm_train2), 
                                             label = clothing_reviews_train$TYPE == "1"),
                          nrounds = 500)
  
  
  
  
  pred2 <- predict(xgb_model2, dtm_test2)
  
  confusionMatrix(clothing_reviews_test$TYPE,
                  as.factor(pred2))  
  
Glimpse of data

glimpse(d)
Rows: 1,819
Columns: 14
$ PLUGIN_RULE_KEY             <fct> InsufficientBranchCoverage, InsufficientLineCo~
$ PLUGIN_CONFIG_KEY           <fct> , , , , , , , , , , S1120, , , , StringEqualit~
$ PLUGIN_NAME                 <fct> common-java, common-java, common-java, common-~
$ DESCRIPTION                 <chr> "An issue is created on a file as soon as the ~
$ SEVERITY                    <fct> MAJOR, MAJOR, MAJOR, MAJOR, MAJOR, MAJOR, MINO~
$ NAME                        <fct> "Branches should have sufficient coverage by t~
$ DEF_REMEDIATION_FUNCTION    <fct> LINEAR, LINEAR, LINEAR, LINEAR_OFFSET, LINEAR,~
$ DEF_REMEDIATION_GAP_MULT    <fct> 5min, 2min, 2min, 10min, 10min, 10min, , , , ,~
$ DEF_REMEDIATION_BASE_EFFORT <fct> , , , 10min, , , 5min, 5min, 15min, 15min, 1mi~
$ GAP_DESCRIPTION             <fct> "number of uncovered conditions", "number of l~
$ SYSTEM_TAGS                 <fct> "bad-practice", "bad-practice", "convention", ~
$ IS_TEMPLATE                 <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0~
$ DESCRIPTION_FORMAT          <fct> HTML, HTML, HTML, HTML, HTML, HTML, HTML, HTML~
$ TYPE                        <fct> CODE_SMELL, CODE_SMELL, CODE_SMELL, CODE_SMELL~
0

There are 0 answers