Information distribution device

Abstract

Provided is an information distributing device capable of distributing information at a timing that makes it easy to guide a user to a facility.　In a recommendation distributing device 100, which is an information distributing device according to the present disclosure, a recommendation distributing unit 104 distributes information at a timing corresponding to a visited facility category (facility information) of a facility visited by a user and a situation of the user, on the basis of visited facility history information indicating the visited facility category (facility information), and the date and time of the visit, and situation history information including the situation of the user and the occurrence date and time thereof. According to this configuration, information can be distributed at a timing that makes it easy to guide the user to a facility. Consequently, distribution of information, such as distribution of recommendations, can be performed at a timing at which conversion (a store visit, usage and awarding of d-points in participating stores), which is a key performance indicator (KPI) of a distribution strategy, is easy to achieve, leading effectively to conversion.

Claims

1. An acquisition unit acquires facility history information indicating the facility information and visit time of the facilities visited by the user, and status history information including the user's status and the time it occurred. A distribution unit that distributes information at a timing appropriate to the facility information and the status based on the aforementioned visit history information and status history information, A calculation unit that calculates information indicating the possibility of visiting in the situation based on the visit time or occurrence time included in the visit history information or the situation history information, A data generation unit generates a matching table for each user, matching the aforementioned situation with the aforementioned facility information, using the aforementioned visit time and the aforementioned occurrence time as keys. Based on the aforementioned matching table, the occurrence count derivation unit derives the number of occurrences of each situation for each time period for each user, Based on the aforementioned matching table, a visit count derivation unit derives the number of visits to the facility for each user under the circumstances that occurred during each time period, A probability calculation unit calculates the likelihood of a visit based on the number of occurrences and the number of visits for each combination of the user, time period, situation, and facility information, Equipped with, The aforementioned occurrence count derivation unit is, For any combination of time periods and circumstances for each user for which the number of occurrences cannot be derived, the number of occurrences in other combinations is used for interpolation. The aforementioned distribution unit, Based on the information indicating the possibility of visiting, the distribution process for information regarding the facility in the aforementioned situation is performed, Based on the aforementioned visitability, the time for information distribution will be determined and the information will be distributed. Information distribution device.

2. The aforementioned occurrence count derivation unit is, A frequency matrix is ​​generated based on the number of occurrences of each time period and situation combination for each user. By performing matrix factorization on the aforementioned occurrence count matrix, an occurrence count user matrix and an occurrence count item matrix relating to combinations of time periods and situations are generated. Based on the aforementioned occurrence count user matrix and occurrence count item matrix, the occurrence count for the undeterminable combinations for each user is interpolated. The information distribution device according to claim 1.

3. The aforementioned occurrence count derivation unit is, If interpolation is not possible based on the user matrix of occurrence counts and the item matrix of occurrence counts, interpolation will be performed based on the status of other users. The information distribution device according to claim 2.

4. An acquisition unit that acquires facility history information indicating facility information and visit time of facilities visited by the user, and status history information including the user's status and the time it occurred, A distribution unit that distributes information at a timing appropriate to the facility information and the status based on the aforementioned visit history information and status history information, A calculation unit that calculates information indicating the possibility of visiting in the situation based on the visit time or occurrence time included in the visit history information or the situation history information, A data generation unit generates a matching table for each user, matching the aforementioned situation with the aforementioned facility information, using the aforementioned visit time and the aforementioned occurrence time as keys. Based on the aforementioned matching table, the occurrence count derivation unit derives the number of occurrences of each situation for each time period for each user, Based on the aforementioned matching table, a visit count derivation unit derives the number of visits to the facility for each user under the circumstances that occurred during each time period, A probability calculation unit calculates the likelihood of a visit based on the number of occurrences and the number of visits for each combination of the user, time period, situation, and facility information, Equipped with, The distribution unit determines the time for information distribution based on the likelihood of visits and then distributes the information. Based on the information indicating the possibility of visiting, the distribution process for information regarding the facility in the aforementioned situation is performed. The aforementioned visit count derivation unit is: For combinations of facility information visited during the aforementioned time period and under the aforementioned circumstances, if the number of visits cannot be derived from a combination, the number of visits from other combinations will be used for interpolation. Information distribution device.

5. The aforementioned visit count derivation unit is: A visit count matrix is ​​generated based on the number of visits to a facility for each user, considering the combination of time of day, situation, and facility information visited. By performing matrix factorization on the aforementioned visit count matrix, a visit count user matrix representing users and a visit count item matrix representing combinations of information about time, situation, and facility are generated. Based on the visit count user matrix and the visit count item matrix, the visit count for each user for the undeterminable combination is interpolated. The information distribution device according to claim 4.

Images