Abstract:Understanding and being able to react to customer feedback is the most fundamental task in providing good customer service. However, there are two major obstacles for international companies to automatically detect the meaning of customer feedback in a global multilingual environment. Firstly, there is no widely acknowledged categorisation (classes) of meaning for customer feedback. Secondly, the applicability of one meaning categorisation, if it exists, to customer feedback in multiple languages is questionable. In this paper, we extracted representative real world samples of customer feedback from Microsoft Office customers in multiple languages, English, Spanish and Japanese,and concluded a five-class categorisation(comment, request, bug, complaint and meaningless) for meaning classification that could be used across languages in the realm of customer feedback analysis.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how international companies can automatically detect the meaning of customer feedback in a global multilingual environment. Specifically, the paper mainly addresses the following two issues:
1. **Lack of widely - recognized classification of customer feedback meaning**:
- Currently, there is no widely - accepted classification of customer feedback meaning (i.e., categories). Different companies and organizations may use different classification methods, but these classifications are often not public or not applicable to multiple languages.
2. **Applicability issues of multilingual customer feedback classification**:
- Even if there is a certain classification method, its applicability in different languages is also doubtful. Due to differences between languages, a classification method may be effective in some languages but not in others.
To solve these problems, by analyzing multilingual feedback (including English, Spanish and Japanese) from Microsoft Office customers, the paper proposes a general five - category classification method (comment, request, bug, complaint, meaningless), aiming at cross - language application. This method can help international companies more accurately understand and classify customers' intentions when dealing with multilingual customer feedback.
### Main contributions of the paper
- **Proposing a five - category classification method**: Through the labeling and analysis of Spanish and Japanese customer feedback, a general five - category classification method has been summarized, which can be used for the classification of multilingual customer feedback.
- **Verifying the cross - language applicability of the classification method**: By comparing feedback data in different languages, the applicability of this classification method in different languages has been proven.
- **Exploring the impact of machine translation quality**: The impact of machine translation quality on the accuracy of customer feedback classification has been studied, and it has been found that although there are differences in translation quality, generally it will not significantly affect the classification effect.
### Formula representation
In this paper, although no complex mathematical formulas are involved, for the sake of clear expression, the following is a simple formula example regarding the improvement of classification accuracy:
\[ \text{Classification accuracy improvement}=\frac{\text{Number of correct classifications after improvement}-\text{Number of original correct classifications}}{\text{Total number of samples}}\times100\% \]
This formula can help quantify the improvement degree of classification accuracy, especially in different languages and under different translation qualities.
### Summary
By proposing a general five - category classification method and verifying its applicability in a multilingual environment, this paper provides international companies with a better tool to understand and process customer feedback.