Tracking Sentiment in Mail: How Genders Differ on Emotional Axes (1309.6347v1)

Abstract: With the widespread use of email, we now have access to unprecedented amounts of text that we ourselves have written. In this paper, we show how sentiment analysis can be used in tandem with effective visualizations to quantify and track emotions in many types of mail. We create a large word--emotion association lexicon by crowdsourcing, and use it to compare emotions in love letters, hate mail, and suicide notes. We show that there are marked differences across genders in how they use emotion words in work-place email. For example, women use many words from the joy--sadness axis, whereas men prefer terms from the fear--trust axis. Finally, we show visualizations that can help people track emotions in their emails.

• The paper describes creating a large word-emotion lexicon via crowdsourcing to enable sentiment analysis in email communication.
• Analysis reveals distinct emotional profiles across different mail types (e.g., joy/trust in love letters, sadness/fear in suicide notes) and significant gender differences in workplace emails, with women using more joy/sadness terms and men more trust/fear terms.
• This research has practical implications for enhancing tools for analyzing interpersonal/organizational dynamics and potentially integrating emotion tracking into email services for personal use and broader applications.

Tracking Sentiment in Mail: Gender Differences on Emotional Axes

The paper presented by Saif M. Mohammad and Tony Yang investigates sentiment analysis in email communication, focusing on gender-based differences in emotional expressions. Their research involves the creation of a comprehensive word-emotion association lexicon via crowdsourcing, which aids in identifying and comparing the presence of emotional indicators in different types of correspondences, such as love letters, hate mail, and suicide notes. The work extends to explore these dynamics within work-place emails, using the Enron email corpus as a resource.

The methodology utilized in the development of the word-emotion lexicon is notably rigorous. The authors employed Amazon's Mechanical Turk to annotate terms with associations to a set of emotions: joy, sadness, fear, trust, anger, disgust, surprise, and anticipation, acknowledging previous semantic work by psychologist Robert Plutchik. This meticulous approach yielded a lexicon of approximately 14,200 words, ensuring reliable tracking of sentiment across extensive textual datasets.

The paper demonstrates clear distinctions between the types of mail analyzed. Love letters were found to be predominately associated with emotions such as joy and trust, while hate mail contained a larger proportion of fear and anger-related terms. Suicide notes presented a unique emotional spectrum closely tied to sadness and fear. The visualizations used—such as pie charts and relative-salience word clouds—enhance comprehension of these distributions and support the overall sentiment analysis.

One of the salient conclusions of this research is the marked gender differences in emotional expressions within workplace emails. Women tend to utilize words related to joy and sadness more frequently, supporting the narrative of women fostering interpersonal relationships through emotional communication. Conversely, men exhibit a preference for trust and fear terms, indicative of a communication style oriented towards social positioning and concern. This finding corroborates existing literature on gendered communication patterns.

The practical implications of this paper are multifaceted. By enabling sophisticated analysis of emotions embedded in everyday communications, this research potentially enhances tools for monitoring and improving interpersonal and organizational dynamics. Furthermore, automatic emotion tracking systems could serve as early indicators of psychological distress in professional settings, thus contributing to a more emotionally intelligent workplace.

For future development, the researchers suggest integrating their sentiment analysis capabilities with email services—such as Gmail—to allow users to track personal emotional communication patterns. This could lead to broader applications in affect-based searches and improved emotional conveyance in digital communication. The authors also propose extending their research to include longitudinal studies, which could analyze emotional trends over time and assess changes resulting from external interventions.

In summary, this paper underscores the significant presence of emotional language within digital communications, highlighting gender-specific tendencies. The insights garnered from this research lay the groundwork for advanced emotion-sensitive applications in both personal and professional domains, offering intriguing possibilities for the evolution of AI-driven sentiment analysis tools.