R：构建文本分类器

Question

我的内容集必须根据一些规则进行分类。

示例数据：

    1     chin jeffrey hong kong  wednesday  october     global business reporting cc subramanian raghuveer   kumar m santhosh   antoo ramesh subject request  obtain global icis data dear team appreciate   can distribute   monthly basis  latest global icis data    ramesh antoo upon  availability regards jeffrey chin associate business risk strategy  efficiency brse  asia international  institutional banking australia  new zealand banking group limited f three exchange square  connaught place central hong kong phone    voice net   email jeffreychinanzcom brse   key business risk interface team within iib providing global support  strategic perspectives  policy procedures  reporting  includes risk appetite credit process quality assurance interlock  key stakeholders  well  iib support  key projects   project glue cacheorion working groups efficiency initiatives  business risk forums p please consider  environment  printing  email
    2      beren stuart vanuatu  monday  october     g s venkatesh ramesh sandeep talanki   h r nagaraj subject please approve  qlikview gpa access hi please process  following form  gpa access please email  requestor  line manager   access  granted raj can  please add  given  user  qlikview workgroup    gpa access form  requestors name lim tek kon vanuatu address lini high way port vila efate title  relationship manager emerging corporates employee id  lan id limtk bsbcc  authorising manager beren stuart vanuatu  yes   read  use  gpa dashboard business technical reason na  
    3     kumar m santhosh   behalf  relationshipbankingfinancesupport  friday  october     g s venkatesh cc global business reporting subject fw please approve  qlikview gpa access regards santhosh   faunt daniel png  wednesday  october     relationshipbankingfinancesupport cc amet sova subject fw please approve  qlikview gpa access hi   unable  approve  excel due  macro issues please process   amet sova  monday  october     faunt daniel png subject please approve  qlikview gpa access hello can  please review  attached form  click line manager approval  approve 
    4     thomson owen tonga  thursday  october     g s venkatesh ramesh sandeep talanki   h r nagaraj subject please approve  qlikview gpa access hi please process  following form  gpa access please email  requestor  line manager   access  granted raj can  please add  given  user  qlikview workgroup    gpa access form  requestors name hia viliami address head office fakafanua centre maufanga vuna road nukualofa tongatapu tonga nukualofa tongatapu title  nfc  amu manager employee id  lan id hiav bsbcc   authorising manager thomson owen tonga  yes   read  use  gpa dashboard business technical reason  
    5     kumar rajesh fiji  tuesday  october     g s venkatesh ramesh sandeep talanki   h r nagaraj subject please approve  qlikview gpa access hi please process  following form  gpa access please email  requestor  line manager   access  granted raj can  please add  given  user  qlikview workgroup    gpa access form  requestors

这只是一行值，我必须实时处理超过 500 - 10000 行，这里我提取了将要使用的词

> O
$text
$text[[1]]
 [1] "qlikview" "gpa"      "access"   "gpa"      "access"   "access"   "qlikview" "gpa"      "access"  "gpa"     

$text[[2]]
 [1] "report"   "qlikview" "gpa"      "access"   "qlikview" "gpa"      "access"   "qlikview" "gpa"     
[10] "access"  

$text[[3]]
 [1] "qlikview" "gpa"      "access"   "gpa"      "access"   "access"   "qlikview" "gpa"      "access"  
[10] "gpa"     

$text[[4]]
 [1] "qlikview" "gpa"      "access"   "gpa"      "access"   "access"   "qlikview" "gpa"      "access"  
[10] "gpa"     

$text[[5]]
 [1] "report"   "qlikview" "gpa"      "access"   "access"   "gpa"      "access"   "qlikview" "gpa" "access"   "access"   "gpa"      "qlikview" "gpa"      "access"   "qlikview" "gpa"  "access"  

$text[[6]]
 [1] "report"   "qlikview" "access"   "access"   "report"   "qlikview" "access"   "access"   "gpa"     
[10] "qlikview" "access"   "access"   "qlikview" "access"   "access"  

$text[[7]]
 [1] "report" "report" "access" "access" "report" "report" "report" "report" "report" "report" "data"  "data"   "report" "access" "report" "report"

$text[[8]]
[1] "report"   "qlikview" "gpa"      "access"   "gpa"      "access"  

$text[[9]]
 [1] "report" "gpa"    "access" "access" "gpa"    "gpa"    "gpa"    "gpa"    "gpa"    "access" "gpa"   "gpa"    "gpa"    "report"

$text[[10]]
 [1] "report" "gpa"    "gpa"    "access" "gpa"    "access" "gpa"    "access" "gpa"    "gpa"    "report" "gpa"    "gpa"    "access" "gpa"    "gpa"    "gpa"    "gpa"    "gpa"

现在我必须使用 if 条件 while 在此基础上建立规则，如何表示列表以循环每个文本并检查 "access" 是否存在然后检查 "gpa" 或 "qlikview" 存在，然后 return 作为行值的 ACCESS（示例数据）否则，如果 "report" 存在，则检查 "pfi" 或 "Regional" return REPORT

我已将上面的列表转换为如下所示的数据框

代码：

maxl <- max(sapply(O,length))
out <- do.call(cbind, lapply(O,function(x) x[1:maxl]))
out <- as.data.frame(out) 
    text
1   c("qlikview", "gpa", "access", "gpa", "access", "access", "qlikview", "gpa", "access", "gpa")
2   c("report", "qlikview", "gpa", "access", "qlikview", "gpa", "access", "qlikview", "gpa", "access")
3   c("qlikview", "gpa", "access", "gpa", "access", "access", "qlikview", "gpa", "access", "gpa")
4   c("qlikview", "gpa", "access", "gpa", "access", "access", "qlikview", "gpa", "access", "gpa")
5   c("report", "qlikview", "gpa", "access", "access", "gpa", "access", "qlikview", "gpa", "access", "access", "gpa", "qlikview", "gpa", "access", "qlikview", "gpa", "access")
6   c("report", "qlikview", "access", "access", "report", "qlikview", "access", "access", "gpa", "qlikview", "access", "access", "qlikview", "access", "access")

如何删除此数据框中的列表类型 c()？

尝试了一些代码：

 #Rule Classifier-----
rule <- function(out)
{
  for(i in out)
  {
    for(j in out[i])
    {
      if(x[j]=="Access")
      {
        if(x[j]=="gpa" | x[j]=="qlikview")
        {
          return("Access")
        }
      }
      else if(x[j]=="Report")
      {
        if(x[j]=="pfi" | x[j]=="data" )
        {
          return("Report")
        }
      }
    }
  }
}

预期输出:

1      Access
2       Access
3       Access
4       Access
5       Access
6       Access
7       Report/Data
8       Access
9       Access
10      Access
11     Report/Data
12     Report/Data
13     Report/Data
14     Report/Data

dput(O$text[1:10])

list(c("qlikview", "gpa", "access", "gpa", "access", "access", 
"qlikview", "gpa", "access", "gpa"), c("report", "qlikview", 
"gpa", "access", "qlikview", "gpa", "access", "qlikview", "gpa", 
"access"), c("qlikview", "gpa", "access", "gpa", "access", "access", 
"qlikview", "gpa", "access", "gpa"), c("qlikview", "gpa", "access", 
"gpa", "access", "access", "qlikview", "gpa", "access", "gpa"
), c("report", "qlikview", "gpa", "access", "access", "gpa", 
"access", "qlikview", "gpa", "access", "access", "gpa", "qlikview", 
"gpa", "access", "qlikview", "gpa", "access"), c("report", "qlikview", 
"access", "access", "report", "qlikview", "access", "access", 
"gpa", "qlikview", "access", "access", "qlikview", "access", 
"access"), c("report", "report", "access", "access", "report", 
"report", "report", "report", "report", "report", "data", "data", 
"report", "access", "report", "report"), c("report", "qlikview", 
"gpa", "access", "gpa", "access"), c("report", "gpa", "access", 
"access", "gpa", "gpa", "gpa", "gpa", "gpa", "access", "gpa", 
"gpa", "gpa", "report"), c("report", "gpa", "gpa", "access", 
"gpa", "access", "gpa", "access", "gpa", "gpa", "report", "gpa", 
"gpa", "access", "gpa", "gpa", "gpa", "gpa", "gpa"))
rule(out)
#this is throwing some error -  Error in `[.default`(out, i) : invalid subscript type 'list'

我知道这很天真，但我是新手，如果我哪里出错了请纠正我。

Answer 1

您似乎假设 for ... in 循环将使用整数进行迭代。它们只是遍历列表中的对象，因此当它是列表时，您正试图使用 i 作为索引。但是 ?lapply 将向您展示使用列表的更好方法。

text <- list(c("qlikview", "gpa", "access", "gpa", "access", "access",
            "qlikview", "gpa", "access", "gpa"), 
          c("report", "qlikview", 
            "gpa", "access", "qlikview", "gpa", "access", "qlikview", "gpa", 
            "access"), 
          c("qlikview", "gpa", "access", "gpa", "access", "access", 
            "qlikview", "gpa", "access", "gpa"), 
          c("qlikview", "gpa", "access", 
            "gpa", "access", "access", "qlikview", "gpa", "access", "gpa"), 
          c("report", "qlikview", "gpa", "access", "access", "gpa", 
            "access", "qlikview", "gpa", "access", "access", "gpa", "qlikview", 
            "gpa", "access", "qlikview", "gpa", "access"), 
          c("report", "qlikview", 
            "access", "access", "report", "qlikview", "access", "access", 
            "gpa", "qlikview", "access", "access", "qlikview", "access", 
            "access"), 
          c("report", "report", "access", "access", "report", 
            "report", "report", "report", "report", "report", "data", "data", 
            "report", "access", "report", "report"), 
          c("report", "qlikview", 
            "gpa", "access", "gpa", "access"), 
          c("report", "gpa", "access", 
            "access", "gpa", "gpa", "gpa", "gpa", "gpa", "access", "gpa", 
            "gpa", "gpa", "report"), 
          c("report", "gpa", "gpa", "access", 
            "gpa", "access", "gpa", "access", "gpa", "gpa", "report", "gpa", 
            "gpa", "access", "gpa", "gpa", "gpa", "gpa", "gpa")) 
O <- as.data.frame(cbind(text))

rule <- function(out) {   
  vapply(out$text, function (row) {
    if ("access" %in% row && ("gpa" %in% row || "qlikview" %in% row)) {
      return("Access")
    } else if ("report" %in% row && ("pfi" %in% row || "data" %in% row)) {
      return("Report/Data")
    } else {
      return("NA")
    }   
  }, "") 
}

rule(O)
#  [1] "Access"      "Access"      "Access"      "Access"      "Access"      "Access"      "Report/Data" "Access"     
#  [9] "Access"      "Access"

R：构建文本分类器

R: building text Classifier

regex

r

rules

document-classification