Information processing system, information processing method, program, and non-transitory information storage medium
View Patent ↗A difference in tendency of times associated with combinations of content clusters and user clusters among the user clusters is reflected in a result of correspondence between the content cluster and the user cluster. A data acquisition unit acquires association data indicating a combination of a content belonging to the content cluster, a user belonging to one of a plurality of user clusters, and a time relating to a combination of the content and the user. A dividing unit divides, under a condition that the tendency of the times associated with the users in the association data differs among the plurality of user clusters to which the users belong, the content cluster into a plurality of clusters each corresponding to at least one of the plurality of user clusters.
1. An information processing system for analyzing a selected content cluster among a plurality of content clusters, where the selected content cluster includes a group of associated content and is in correspondence with a plurality of respective user clusters, each user cluster including a respective group of associated users, the system comprising a micro-processor operating under the instructions of a computer program, which causes the micro-processor to include:
an association data acquisition unit that acquires association data, which includes respective combinations of the content belonging to the selected content cluster, the users belonging to the plurality of user clusters, and a plurality of times at which the users took action concerning the content of the selected content cluster;
a statistic amount calculation unit that calculates statistics for each of the user clusters based on the plurality of times of the association data, wherein the statistics for each of the user clusters includes: (i) an average (AVG) representing an average of all the times that the users of a given one of the plurality of user clusters took action concerning the content of the selected content cluster, (ii) a standard deviation (STD) representing an standard deviation of all the times that the users of the given one of the plurality of user clusters took action concerning the content of the selected content cluster, and (iii) computing a time scope for each user cluster, where AVG−STD time scope≦AVG+STD;
a determining unit that determines whether or not the selected content cluster is to be divided into at least two content clusters, the determination being based on whether one of the time scopes does not overlap any of the other time scopes, and if so then the determination is affirmative and the selected content cluster is to be divided into the at least two content clusters;
a dividing unit that divides the selected content cluster into at least two content clusters when the determination is in the affirmative, and specifies for each of the at least two content clusters at least one of the user clusters to be in correspondence therewith; and
a setting unit that sets respective groups of the content into each of the at least two content clusters based on the association data.
2. The information processing system according to claim 1 , further comprising an information providing unit that provides, to the user, information relating to a content belonging to a content cluster corresponding to the user cluster to which the user belongs.
3. The information processing system according to claim 1 , wherein the respective times in the association data is one of a time when the user accesses information relating to the content, a time when a comment on the content is received from the user, and a time when the user purchases the content.
4. An information processing method for analyzing a selected content cluster among a plurality of content clusters, where the selected content cluster includes a group of associated content and is in correspondence with a plurality of respective user clusters, each user cluster including a respective group of associated users, the method comprising:
acquiring association data, which includes respective combinations of the content belonging to the selected content cluster, the users belonging to the plurality of user clusters, and a plurality of times at which the users took action concerning the content of the selected content cluster;
calculating statistics for each of the user clusters based on the plurality of times of the association data, wherein the statistics for each of the user clusters includes: (i) an average (AVG) representing an average of all the times that the users of a given one of the plurality of user clusters took action concerning the content of the selected content cluster, (ii) a standard deviation (STD) representing an standard deviation of all the times that the users of the given one of the plurality of user clusters took action concerning the content of the selected content cluster, and (iii) computing a time scope for each user cluster, where AVG−STD time scope≦AVG+STD;
determining whether or not the selected content cluster is to be divided into at least two content clusters, the determination being based on whether one of the time scopes does not overlap any of the other time scopes, and if so then the determination is affirmative and the selected content cluster is to be divided into the at least two content clusters;
dividing the selected content cluster into at least two content clusters when the determination is in the affirmative, and specifying for each of the at least two content clusters at least one of the user clusters to be in correspondence therewith; and
setting respective groups of the content into each of the at least two content clusters based on the association data.
5. A non-transitory computer readable information storage medium storing a program which is to be executed by a computer and cause the computer to execute actions for analyzing a selected content cluster among a plurality of content clusters, where the selected content cluster includes a group of associated content and is in correspondence with a plurality of respective user clusters, each user cluster including a respective group of associated users, the actions comprising:
acquiring association data, which includes respective combinations of the content belonging to the selected content cluster, the users belonging to the plurality of user clusters, and a plurality of times at which the users took action concerning the content of the selected content cluster;
calculating statistics for each of the user clusters based on the plurality of times of the association data, wherein the statistics for each of the user clusters includes: (i) an average (AVG) representing an average of all the times that the users of a given one of the plurality of user clusters took action concerning the content of the selected content cluster, (ii) a standard deviation (STD) representing an standard deviation of all the times that the users of the given one of the plurality of user clusters took action concerning the content of the selected content cluster, and (iii) computing a time scope for each user cluster, where AVG−STD time scope≦AVG+STD;
determining whether or not the selected content cluster is to be divided into at least two content clusters, the determination being based on whether one of the time scopes does not overlap any of the other time scopes, and if so then the determination is affirmative and the selected content cluster is to be divided into the at least two content clusters;
dividing the selected content cluster into at least two content clusters when the determination is in the affirmative, and specifying for each of the at least two content clusters at least one of the user clusters to be in correspondence therewith; and
setting respective groups of the content into each of the at least two content clusters based on the association data.