Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Peer-to-peer online support groups and the discussion forums in these groups can help patients by providing opportunities for increasing their empowerment. Most previous research on online empowerment and online social support uses qualitative methods, or questionnaires to gain insight into the dynamics of online empowerment processes. Objective: The overall goal of this study was to analyze the presence of the empowering processes in the online communication of people affected by cancer, using text mining techniques. Use of these relatively new methods enable to study social processes such as empowerment on large scale and with unsolicited data, avoiding unintended biases caused by research tools. Methods: The sample consisted of 5534 messages in 1708 threads by 2071 by users of a forum for cancer patients and their relatives. Supervised and unsupervised text mining techniques were used to explore the online communication of the users. Supervised text mining techniques were applied to operationalize empowerment: a random subsample of 2051 messages in the forum was manually labeled with five empowerment processes (sharing narrative, asking a question, providing information, emotional support, reference to information), using a coding scheme based on knowledge from previous research. Machine learning classifiers (Linear Support Vector Classification and Logistic regression) were then trained on the content of the messages with their manually assigned labels and the remainder of the sample was automatically labeled with empowerment processes using the generated model. For the unsupervised part, the messages in the sample were automatically labeled with categories from the Linguistic Inquiry and Word Count (LIWC). Subsequently, the supervised and unsupervised data were studied in conjunction. Empowerment was studied in relation to the intensity of online participation in relation to the LIWC categories. Results: The precision of the automatic labeling was 85.6%, which was considered to be sufficient for automatically labeling the complete corpus and doing further analyses on the labeled data. 62.9% of the messages contained a narrative, 23.8% a question, and 27.5% informational support. Emotional support and references to external sources were less frequent. Users with more posts more often referred to an external source, provided informational support, and emotional support (Kendall τ > 0.2, p < 0.001) and less often shared narratives (Kendall τ = -0.297, p < 0.001). A number of LIWC categories are significant predictors for the empowerment processes. Words expressing assent (‘ok’, ‘yes’) and emotional processes (expressions of feelings) are significant positive predictors for emotional support (p = 0.002). Conclusions: Empowerment processes are associated with intensity of online use. The relation between linguistic analyses and empowerment processes indicates that empowerment processes can be identified from the occurrences of specific linguistic cues.