Systems and methods of log optimization for television advertisements
Embodiments of the present invention provide systems and methods of log optimization for television advertisements. Exemplary method and systems can comprise: receiving, at a computer software platform, (i) the digital log, (ii) constraint parameters, and (iii) viewership predictions; and generating, with the computer software platform, an optimized log based on the received digital log, the constraint parameters, and the viewership predictions.
1. A method for optimizing a digital log of N available advertisements for a television network publisher, the method comprising:
(a) receiving, by a computer, (i) the digital log, (ii) constraint parameters, and (iii) viewership predictions, each for the N available advertisements;
(b) generating, by the computer, a constraint matrix of size N×N, wherein each entry in the constraint matrix corresponds to a separation time between two of the available advertisements in the digital log, the separation time determined based on the constraint parameters of the two available advertisements;
(c) analyzing, by the computer, the constraint matrix using a graph theory process to separate the available advertisements into a number of groups based on the separation times provided in the constraint matrix based on the separation times in the constraint matrix, wherein the number of groups correspond to maximal cliques of the constraint matrix as determine by the group theory process, wherein each group consists of a portion of the available advertisements all having separation times less than a threshold amount;
(d) if the number of groups is greater than a predetermined threshold, applying, by the computer, a further constraint selected from (i) preventing a group of the available advertisements from appearing more than once in a commercial break, or (ii) freezing a portion of the available advertisements to initial advertisement slots, and repeating steps (c) and (d) with the further constraint to further reduce the number of groups until the reduced number of groups is below the predetermined threshold;
(e) generating, with the computer, an optimized log using a linear integer programming (LIP) optimization process based on the reduced groups of available advertisements and the viewership predictions, wherein the optimized log comprises an updated schedule for airing the N available advertisements;
(f) transmitting the optimized log from the computer to an advertisement trafficking server; and
(g) moving and airing, by the advertisement trafficking server, the N available advertisements according to the updated schedule of the optimized log.
2. The method of claim 1 , wherein the digital log is a list of the available advertisements for the television network publisher or an optimized list of the available advertisements for the television network publisher.
3. The method of claim 1 , wherein the constraint parameters include one at least one of: separation constraints, exclusions, and time locks,
wherein the viewership predictions are based on a machine learning model, and
wherein the machine learning model is configured to predict a number of viewers of a certain demographics that will watch a certain program.
4. The method of claim 3 , further comprising:
transmitting, with the advertisement trafficking server, data associated with the moving and airing of the N available advertisements according to the updated schedule of the optimized log to the computer;
retraining, with the computer, the machine learning model with the transmitted data.
5. The method of claim 1 , further comprising:
generating, with the computer, a report including a difference of a first revenue associated with the digital log and a second revenue associated with the optimized log.
6. A system for optimizing a digital log of N available advertisements for a television network publisher, the system comprising:
at least one computer comprising a processor configured to execute instructions, wherein the instructions cause the processor to perform operations comprising:
(a) receiving, by a computer, (i) the digital log, (ii) constraint parameters, and (iii) viewership predictions, each for the N available advertisements,
(b) generating, by the computer, a constraint matrix of size N×N, wherein each entry in the constraint matrix corresponds to a separation time between two of the available advertisements in the digital log, the separation time determined based on the constraint parameters of the two available advertisements;
(c) analyzing, by the computer, the constraint matrix using a graph theory process to separate the available advertisements into a number of groups based on the separation times provided in the constraint matrix based on the separation times in the constraint matrix, wherein the number of groups correspond to maximal cliques of the constraint matrix as determine by the group theory process, wherein each group consists of a portion of the available advertisements all having separation times less than a threshold amount,
(d) if the number of groups is greater than a predetermined threshold, applying, by the computer, a further constraint selected from (i) preventing a group of the available advertisements from appearing more than once in a commercial break, or (ii) freezing a portion of the available advertisements to initial advertisement slots, and repeating steps (c) and (d) with the further constraint to further reduce the number of groups, until the reduced number of groups is below the predetermined threshold, and
(e) generating, with the computer, an optimized log using a linear integer programming (LIP) optimization process based on the groups of available advertisements, wherein the optimized log comprises an updated schedule for airing the N available advertisements; and
an advertisement trafficking server configured to receive the optimized log from the at least one computer, and configured to move and air the N available advertisements according to the updated schedule of the optimized log.
7. The system of claim 6 , wherein the digital log is one of a list of the available advertisements for the television network publisher or an optimized list of the available advertisements for the television network publisher.
8. The system of claim 6 , wherein the constraint parameters include one at least one of separation constraints, exclusions, and time locks, wherein the viewership predictions are based on a machine learning model, wherein the machine learning model is configured to predict a number of viewers of a certain demographics that will watch a certain program.
9. The system of claim 8 , wherein the advertisement trafficking server is configured to transmit data associated with the moving and airing of the N available advertisements according to the updated schedule of the optimized log to the computer, and wherein the computer is configured to retrain the machine learning model with the transmitted data.
10. The system of claim 6 , wherein the computer is further configured to:
generate a report including a difference of a first revenue associated with the digital log and a second revenue associated with the optimized log.