Reputation system
Reputation systems have been used to promote trust between participants in activities including online activities such as online market places. Existing online market places provide a reputation system which is a simple cumulative registry of user ratings on a given market place member. However, this simple system is open to abuse in situations where, for example, many positive ratings are given in a fraudulent manner. By modeling both reputation of participants and required reputation of participants an improved reputation system is provided. The required reputation may be thought of as a threshold, referred to herein as a required threshold, which may be used in determining how to update an indication of the reputation of a participant in the activity. The reputation system is able to learn information about required reputation and reputation of participants using an update process which is robust to participants who consistently give feedback of a particular type.
1. A computer-implemented method comprising:
for each participant in an activity, modeling, by one or more processors, a reputation of the participant relating to the activity and modeling, by the one or more processors, a required reputation of the participant relating to a reference value based on a reporting threshold determined by actions of the participant as a reviewer of other participants' behavior in the activity;
receiving, by the one or more processors, information about an interaction between participants in the activity, the information comprising positive, negative, or neutral feedback given by a first participant about a second participant's behavior in the interaction;
accessing, by the one or more processors, information stored in memory about the first participant's required reputation and about the second participant's reputation; and
updating, by the one or more processors, the accessed information on the basis of the received information and storing the updated information in memory, wherein updating the accessed information is carried out using an update rule which comprises a relationship between the first participant's required reputation and the second participant's reputation, wherein the update rule comprises a partial update process when the received information is neutral feedback, and wherein the partial update process comprises:
updating the accessed information with pre-specified feedback comprising positive feedback or negative feedback prior to storing the updated information in memory, and
interpolating a standard deviation value between a standard deviation of the stored information and a standard deviation of the updated information and replacing the standard deviation of the stored information with the interpolated standard deviation value, and interpolating a mean value between a mean value of the stored information and a mean value of the updated information and replacing the mean value of the stored information with the interpolated mean value.
2. The computer-implemented method as claimed in claim 1 wherein modeling the reputation and the required reputation comprises using a probability distribution.
3. The computer-implemented method as claimed in claim 2 wherein accessing the stored information comprises accessing a mean and a standard deviation value of a belief distribution associated with the first participant's required reputation or with the second participant's reputation.
4. The computer-implemented method as claimed in claim 1 wherein updating the accessed information comprises using a Bayesian statistical update mechanism.
5. The computer-implemented method as claimed in claim 1 which further comprises, accessing information about required reputation and reputation for two participants proposed for a potential interaction together in the activity, and using the accessed information to predict an outcome of that proposed interaction.
6. A computer-implemented method comprising:
for each of at least two participants in an activity, modeling, by one or more processors, a reputation of the participant relating to the activity and modeling, by the one or more processors, a required reputation of the participant, relating to a reference value based on a reporting threshold determined by actions of the participant as a reviewer of other participants' behavior in the activity;
receiving, by the one or more processors, information about an interaction between participants in the activity, the information comprising positive, negative, or neutral feedback given by a first participant about a second participant's behavior in the interaction;
accessing, by the one or more processors, information stored in memory about the first participant's required reputation and about the second participant's reputation; and
updating, by the one or more processors, the accessed information on the basis of the received information and storing the updated information in memory, wherein accessing the stored information comprises accessing one or more statistics relating to a belief distribution associated with the first participant's required reputation and accessing one or more statistics relating to a belief distribution associated with the second participant's reputation, wherein updating the accessed information is carried out using an update rule which comprises a relationship between the first participant's required reputation and the second participant's reputation, wherein the update rule comprises a partial update process when the received information is neutral feedback, and wherein the partial update process comprises:
updating the accessed information with pre-specified feedback comprising positive feedback or negative feedback prior to storing the updated information in memory, and
interpolating a standard deviation value between a standard deviation being one of the one or more statistics of the stored information and a standard deviation of the updated information and replacing the standard deviation of the stored information with the interpolated standard deviation value, and interpolating a mean value between a mean value being one of the one or more statistics of the stored information and a mean value of the updated information and replacing the mean value of the stored information with the interpolated mean value.
7. The computer-implemented method as claimed in claim 6 wherein modeling the reputation and the required reputation comprises using a probability distribution.
8. The computer-implemented method as claimed in claim 6 wherein updating the accessed information comprises using a Bayesian statistical update mechanism.
9. One or more computer-readable storage media, the one or more computer-readable storage media being hardware, storing computer-readable instructions that, when executed on a processor, configure the processor to perform acts comprising:
modeling, for each participant in an activity, a reputation of the participant relating to the activity and modeling a required reputation of the participant, relating to a reference value based on a reporting threshold determined by actions of the participant as a reviewer of other participants' behavior in the activity;
receiving information about an interaction between a first and a second participant in the activity, the information comprising positive, negative, or neutral feedback given by the first participant about the second participant's behavior in the interaction;
accessing information stored in memory about the first participant's required reputation and about the second participant's reputation;
updating the accessed information on the basis of the received information and storing the updated information in memory, wherein updating the accessed information is carried out using an update rule which comprises a relationship between the first participant's required reputation and the second participant's reputation, wherein the update rule comprises a partial update process when the received information is neutral feedback, and wherein the partial update process comprises:
updating the accessed information with pre-specified feedback comprising positive feedback or negative feedback prior to storing the updated information in memory,
interpolating a standard deviation value between a standard deviation of the stored information and a standard deviation of the updated information and replacing the standard deviation of the stored information with the interpolated standard deviation value, and interpolating a mean value between a mean value of the stored information and a mean value of the updated information and replacing the mean value of the stored information with the interpolated mean value.
10. The one or more computer-readable storage media as claimed in claim 9 , wherein updating the accessed information comprises using a Bayesian statistical update mechanism.