Table 3.
Main sources of information for health personalization in the Social Web
| Sources | Examples of Information That Can Be Extracted for Health Personalization |
| Personal health records[16] | Personal health information (eg, diagnoses and treatment) Demographic information Genetic information (eg, rare mutations) [33] |
| Textual content | Textual content is present in most of the Web content, and it can contain information about the authors or about the content itself (eg, description of a video). |
| User profiles in online communities | Health risk behaviors (eg, smoking) Demographic information [12,13,31] User preferences (eg, topics of interest) [34] |
| Forum posts and comments | Personal health information (eg, diagnoses and treatments) [32] Emotional/mental status of users [35] Type of content (eg, informational or conversational) [36] |
| Search queries | User interests [37] |
| Tags | Topics of tagged content and users interests [38] |
| Audio | Users emotional status [39,40] Diagnosis (eg, depression) [41] |
| Facial photos | Emotions [42], gender [43], and age [20] |
| Videos | Diagnosis (eg, neurological diseases) [44] Characteristics of videos (eg, topic and style) [45] |
| Ratings | Users preferences and similarities [46] |
| Social networks and links | Community discovery [47,48] Characteristics of Web content [24,49] |
| Web usage data | Classification of users based on navigation patterns (eg, clicks and browsing data) [50] |