Anti-cyberbullying systems and methods
Some embodiments use text and/or image processing methods to determine whether a user of an electronic messaging platform is subject to an online threat such as cyberbullying, sexual grooming, and identity theft, among others. In some embodiments, a text content of electronic messages is automatically harvested and aggregated into conversations. Conversation data are then analyzed to extract various threat indicators. A result of a text analysis may be combined with a result of an analysis of an image transmitted as part of the respective conversation. When a threat is detected, some embodiments automatically send a notification to a third party (e.g., parent, teacher, etc.).
1. A parental control method comprising employing at least one hardware processor of a computer system to:
determine an aggressiveness score, a friendliness score, and a sexual content score of a conversation according to a text content of the conversation, the conversation comprising a sequence of electronic messages exchanged between a first user and a second user, the sequence ordered according to a timestamps associated with each message, wherein the aggressiveness score indicates a level of aggressiveness of the conversation, the friendliness score indicates a level of friendliness of the conversation, the sexual content score indicates whether the conversation comprises sexually-explicit language, and wherein at least one of the aggressiveness score and the friendliness score is determined according to multiple messages of the conversation;
determine a first sentiment score according to messages of the conversation sent by the first user, the first sentiment score indicative of an inferred sentiment of the first user;
determine a second sentiment score according to message of the conversation sent by the second user, the second sentiment score indicative of an interred sentiment of the second user;
determine whether the first user is bullied by the second user according to the aggressiveness score, and friendliness score, sexual content score, and further according to the first and second sentiment scores; and
in response, when the first user is bullied, transmit a parental notification to a parental reporting device identified from a plurality of devices according to the first user, the parental notification indicating that the first user is bullied.
2. The method of claim 1 , comprising determining the aggressiveness score according to a count of messages of the conversation that include aggressive language.
3. The method of claim 1 , comprising determining the aggressiveness score according to a count of consecutive messages separating two messages within the conversation that both include aggressive language.
4. The method of claim 1 , comprising determining that the first user is not bullied when the aggressiveness and friendliness scores indicate that the conversation is both aggressive and friendly.
5. The method of claim 1 , comprising determining that the first user is bullied when the aggressiveness score indicates that a language of the second user is substantially more aggressive than a language of the first user.
6. A computer system comprising at least one hardware processor configured to execute a conversation analyzer and a parental notification dispatcher, wherein: the conversation analyzer is configured to:
determine an aggressiveness score, a friendliness score, and a sexual content score of a conversation according to a text content of the conversation, the conversation comprising a sequence of electronic messages exchanged between a first user and a second user, the sequence ordered according to a timestamps associated with each message, wherein the aggressiveness score indicates a level of aggressiveness of the conversation, the friendliness score indicates a level of friendliness of the conversation, the sexual content score indicates whether the conversation comprises sexually-explicit language, and wherein at least one of the aggressiveness score and the friendliness score is determined according to multiple messages of the conversation;
determine a first sentiment score according to messages of the conversation sent by the first user, the first sentiment score indicative of an inferred sentiment of the first user;
determine a second sentiment score according to message of the conversation sent by the second user, the second sentiment score indicative of an interred sentiment of the second user; and
determine whether the first user is bullied by the second user according to the aggressiveness score, friendliness score, sexual content score, and further according to the first and second sentiment scores; and
the parental notification dispatcher is configured, in response to the conversation analyzer determining that first user is bullied, to transmit a parental notification to a parental reporting device identified from a plurality of devices according to the first user, the notification message indicating that the first user is bullied.
7. The computer system of claim 6 , wherein the conversation analyzer is configured to determine the aggressiveness score according to a count of messages of the conversation that include aggressive language.
8. The computer system of claim 6 , wherein the conversation analyzer is configured to determine the aggressiveness score according to a count of consecutive messages separating two messages within the conversation that both include aggressive language.
9. The computer system of claim 6 , wherein the conversation analyzer is configured to determine that the first user is not bullied when the aggressiveness and friendliness scores indicate that the conversation is both aggressive and friendly.
10. The computer system of claim 6 , wherein the conversation analyzer is configured to determine that the first user is bullied when the aggressiveness score indicates that a language of the second user is substantially more aggressive than a language of the first user.
11. A non-transitory computer-readable medium storing instructions which, when executed by at least one hardware processor of a computer system, cause the computer system to form a conversation analyzer and a parental notification dispatcher, wherein:
the conversation analyzer is configured to:
determine an aggressiveness score, a friendliness score, and a sexual content score of a conversation according to a text content of the conversation, the conversation comprising a sequence of electronic messages exchanged between a first user and a second user, the sequence ordered according to a timestamps associated with each message, wherein the aggressiveness score indicates a level of aggressiveness of the conversation, the friendliness score indicates a level of friendliness of the conversation, the sexual content score indicates whether the conversation comprises sexually-explicit language, and wherein at least one of the aggressiveness score and the friendliness score is determined according to multiple messages of the conversation;
determine a first sentiment score according to messages of the conversation sent by the first user, the first sentiment score indicative of an inferred sentiment of the first user;
determine a second sentiment score according to message of the conversation sent by the second user, the second sentiment score indicative of an interred sentiment of the second user, and
determine whether the first user is bullied by the second user according to the aggressiveness score, friendliness score, sexual content score, and further according to the first and second sentiment scores; and
the parental notification dispatcher is configured, in response to the conversation analyzer determining that first user is bullied, to transmit a parental notification to a parental reporting device identified from a plurality of devices according to the first user, the notification message indicating that the first user is bullied.
12. The method of claim 1 , comprising determining that the first user is not bullied when the aggressiveness score and the sexual content score indicate that the conversation is both aggressive and sexual.
13. The computer system of claim 6 , wherein the conversation analyzer is configured to determine that the first user is not bullied when the aggressiveness score and the sexual content score indicate that the conversation is both aggressive and sexual.
14. The computer system of claim 1 , wherein the conversation analyzer is configured to determine that the first user is not bullied when the aggressiveness score indicates that both the first and second users use aggressive language within the conversation.
15. the computer system of claim 6 , wherein the conversation analyzer is configured to determine that the first user is not bullied when the aggressiveness score indicates that both the first and second users use aggressive language within the conversation.
16. The method of claim 1 , further comprising constructing the conversation in preparation for determining the aggressiveness and friendliness scores, wherein constructing the conversation comprises:
selecting a message from a plurality of electronic messages surreptitiously harvested from a communication device used by the first user; and
including the selected message into the conversation according to whether the message was exchanged between the first user and the second user, and further according to a timestamp associated with the selected message.
17. The computer system of claim 6 , further comprising a message aggregator configured to construct the conversation in preparation for determining the aggressiveness and friendliness scores, wherein constructing the conversation comprises:
selecting a message from a plurality of electronic messages surreptitiously harvested from a communication device used by the first user; and
including the selected message into the conversation according to whether the message was exchanged between the first user and the second user, and further according to a timestamp associated with the selected message.