Amazon SageMaker

Input:

• The algorithm works with text data
• Supported content types: 'text/csv'
• Input should be a .csv file. Maximum number of rows should not exceed 30 per batch transform
• Mandatory Column: 'Text'. Each row should contain social media data, with text less than 280 characters

Output:

Result will only contain the text rows which have life events for a person. 'Text' column contains the actual social media data and 'Category' column represents the category of life event detected.

• Supported content types: 'text/csv'
• Following are the categories with their definitions:*
• relationship: Beginning of a romantic relationship
• exam: Participation or completion of a major exam or test
• marriage: Wedding or marriage related events
• writeup: Graduation on the completion of a school or college course
• movement: Moving to new places however in a non-vacation setting
• job: Beginning of employment in a new work position
• vacation: Going for a vacation
• relationship_: Ending of a romantic relationship
• education: Completion of a major academic exercise
• awards: Winning of an award in a non-sporting setting
• birth: Events related to the birth of a child
• birthday: Birthday related celebration
• book: Completion or release of a fiction or non-fiction book authored by the individual
• death: Mourning on passing away of a significant other
• sick: Person suffering from a major ailment, or loss of physical wellbeing
• acting: Individual being involved in an acting role across media
• agreement: Entering into a major legal contract
• courtCase: Individual being involved in a legal proceeding as either a party or a jury member
• medical: Major surgery or illness
• song: Individual completing the making of a song or musical property
• win: Winning in a sports/athletic setting
• events: Important public events
• other: Any other categories that are not a part of the defined list of categories

Resources

• Sample Output

Mphasis HyperGraf
Sample Input File
Sample Jupyter Notebook