A Computational Approach to Politeness with Application to Social Factors (1306.6078v1)

Abstract: We propose a computational framework for identifying linguistic aspects of politeness. Our starting point is a new corpus of requests annotated for politeness, which we use to evaluate aspects of politeness theory and to uncover new interactions between politeness markers and context. These findings guide our construction of a classifier with domain-independent lexical and syntactic features operationalizing key components of politeness theory, such as indirection, deference, impersonalization and modality. Our classifier achieves close to human performance and is effective across domains. We use our framework to study the relationship between politeness and social power, showing that polite Wikipedia editors are more likely to achieve high status through elections, but, once elevated, they become less polite. We see a similar negative correlation between politeness and power on Stack Exchange, where users at the top of the reputation scale are less polite than those at the bottom. Finally, we apply our classifier to a preliminary analysis of politeness variation by gender and community.

• The paper introduces a classifier that approximates human performance in identifying politeness using advanced text parsing and curated lexicons.
• It leverages a newly annotated corpus from Wikipedia and Stack Exchange to extract strategies like gratitude and deference.
• Findings reveal a strong correlation between politeness and social status, with practical implications for automated moderation and social analysis.

A Computational Framework for Identifying Politeness in Linguistic Requests

The paper introduces a comprehensive computational framework aimed at identifying and quantifying the concept of politeness within linguistic requests, presenting a detailed exploration of the intersection between language use and social variables. The authors leverage a newly annotated corpus of requests, focusing on the domain of Wikipedia and Stack Exchange, to extend traditional linguistic theories of politeness, offering a robust classifier that approximates human performance in its predictions.

Significant findings include the identification of key politeness strategies such as gratitude, deference, and the use of specific lexical markers, which are extracted using advanced text parsing methodologies in conjunction with curated lexicons. The classifier's performance, achieving near-human accuracy, underscores the universality of politeness features across domains. This has practical implications for analyzing large textual databases to understand social interactions and status dynamics.

Key results from the paper reveal that politeness correlates with social status and outcomes in digital communities: for instance, Wikipedia users who later achieve administrative roles generally exhibit higher politeness in requests compared to those who fail in such endeavors. Moreover, a discernible decrease in politeness post-status elevation suggests that power dynamics impact the linguistic politeness of interactions, consistent with long-established theories of language and power.

In cross-domain analysis, the politeness classifier demonstrated robust performance, as reflected by its successful application to the Stack Exchange platform, further validating its domain-independence and general applicability for studying politeness in diverse socio-technical contexts.

The theoretical implications of this research extend into realms of pragmatic theory and social computing, offering potential utility in automated moderation systems and enhancing cross-cultural communication platforms. Future developments could explore integrating contextual variations in politeness usage with sentiment analysis tools to better understand nuanced human-computer interactions.

By systematically connecting linguistic features of politeness with tangible social outcomes, the paper advances our understanding of how language can function as both a barometer and a tool of social stratification. Subsequent inquiries should aim to refine predictions by incorporating broader socio-cultural contexts, potentially illuminating further the complex dynamics between language, politeness, and power in digital communication environments.