基于R语言利用现有词典匹配分析调研评论的技术问询

Hey there! I’ve got you covered for this dictionary-based survey comment analysis. Let’s break this down into simple, actionable steps with code examples you can adapt right away.

1. 先准备示例数据 #

First, let's set up some sample survey comments and dictionaries to work with. I’ll use relatable categories like positive sentiment, negative sentiment, and service-related terms:

# 示例调研评论
survey_comments <- c(
  "The staff was incredibly friendly and helpful!",
  "Waited 2 hours with no update, terrible experience.",
  "Just okay.",
  "Loved the quick checkout process!",
  "Food was cold, won't be coming back."
)

# 示例词典
positive_dict <- c("friendly", "helpful", "loved", "quick")
negative_dict <- c("terrible", "cold", "won't", "wait")
service_dict <- c("staff", "checkout", "update", "experience")

2. 编写匹配函数 #

We’ll create a helper function that checks if a comment contains any term from a given dictionary. This function will handle case insensitivity (so "Friendly" matches "friendly") and full-word matches (to avoid false positives like "wait" matching "waited" if you don’t want that—we’ll adjust for that too).

# 定义匹配函数：检查评论是否包含词典中的任意词汇
check_dictionary_match <- function(comment, dict, full_word = TRUE) {
  # 构建正则表达式：如果是完整单词匹配，加上词边界\\b
  pattern <- if(full_word) {
    paste0("\\b", paste(dict, collapse = "\\b|\\b"), "\\b")
  } else {
    paste(dict, collapse = "|")
  }
  # 检查匹配，忽略大小写
  grepl(pattern, comment, ignore.case = TRUE)
}

3. 生成结果DataFrame #

Now we’ll apply this function to each dictionary and combine everything into a single dataframe where the first column is the original comment, followed by True/False columns for each dictionary:

# 创建结果数据框
result_df <- data.frame(
  Survey_Comment = survey_comments,
  Positive_Match = sapply(survey_comments, check_dictionary_match, dict = positive_dict),
  Negative_Match = sapply(survey_comments, check_dictionary_match, dict = negative_dict),
  Service_Match = sapply(survey_comments, check_dictionary_match, dict = service_dict),
  stringsAsFactors = FALSE
)

# 查看结果
print(result_df)

4. 输出示例 #

When you run the code above, you’ll get a dataframe that looks like this:

Survey_Comment Positive_Match Negative_Match Service_Match
1 The staff was incredibly friendly and helpful!           TRUE          FALSE          TRUE
2   Waited 2 hours with no update, terrible experience.          FALSE           TRUE          TRUE
3                                      Just okay.          FALSE          FALSE         FALSE
4                Loved the quick checkout process!           TRUE          FALSE          TRUE
5                Food was cold, won't be coming back.          FALSE           TRUE         FALSE

额外调整建议 #

• 如果不需要完整单词匹配（比如想让 "waited" 匹配 "wait"），只需要在调用函数时设置 full_word = FALSE。
• 如果你有很多词典，可以把它们放进一个列表里，然后用 purrr::map_dfc 来批量生成列，这样更高效：

  library(purrr)
  
  # 把词典放进列表
  dict_list <- list(
    Positive = positive_dict,
    Negative = negative_dict,
    Service = service_dict
  )
  
  # 批量生成匹配列
  match_columns <- map_dfc(dict_list, ~sapply(survey_comments, check_dictionary_match, dict = .x))
  
  # 合并到结果
  result_df <- bind_cols(data.frame(Survey_Comment = survey_comments), match_columns)
  

内容的提问来源于stack exchange，提问作者thewrightowns