Reach Estimation Using Pairwise Overlap Approximation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the reach of information contact, such as television commercials and digital advertisements, lack accuracy and efficiency in processing load, particularly when considering multiple overlapping information sources.
Innovation Solution
An information processing apparatus and method that acquires individual reaches for multiple information types and calculates a total reach by considering overlapping relationships between two sources while ignoring complex interactions among three or more, using predicted overlap coefficients to enhance accuracy with reduced computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to calculate total reach by considering all overlapping relationships among multiple information sources, then measurement precision is improved, but processing load increases significantly
Solution Approach 1:
The patent segments the calculation process into two distinct parts: (1) exact calculation of two-source overlaps using inclusion-exclusion principle, and (2) approximation of three-source-or-more overlaps as zero. This segmentation allows the system to focus computational resources on the most significant overlap component while simplifying the treatment of higher-order overlaps, thereby resolving the contradiction between precision and processing load.
Solution Approach 2:
The patent applies partial action by selectively calculating overlap components based on their expected magnitude. Two-source overlaps are calculated exactly because they represent the dominant contribution to total reach, while three-source-or-more overlaps are approximated as zero, representing a deliberate omission of minor components. This partial calculation approach maintains sufficient accuracy while dramatically reducing processing requirements.
2Measurement precision
If all overlap components including three-source interactions are considered in reach calculation, then reach accuracy is improved, but arithmetic operation amount increases
Solution Approach 1:
The patent implements partial action by calculating only the necessary overlap components (two-source overlaps) while deliberately omitting higher-order overlaps (three-source and above) from exact calculation. The formula R_total = ΣR_i - ΣR_ij + ΣR_ijk - ... is truncated to R_total ≈ ΣR_i - ΣR_ij, where the omitted terms are assumed to be negligible. This approach achieves acceptable accuracy while significantly reducing the arithmetic operations required.
Solution Approach 2:
The patent treats three-source-or-more overlap components as disposable approximations rather than requiring their exact calculation. By assuming these higher-order overlaps contribute negligibly to the total reach and setting them to zero, the system sacrifices minimal accuracy in exchange for substantial computational efficiency, analogous to using a cheap approximation when high precision is not critical for those specific components.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
An information processing apparatus, an information processing method, and an information processing program capable of acquiring the scale of the number of persons who contact information (reach) with high accuracy and a small processing load (for example, arithmetic operation amount) are provided. An information processing apparatus (1) includes: an individual reach acquisition unit (10) that acquires a first reach that is a scale of the number of persons who contact first information, a second reach that is a scale of the number of persons who contact second information,... and an N-th reach that is a scale of the number of persons who contact N-th information, when N is defined as a natural number of three or more; and a total reach acquisition unit (20) that acquires a total reach that is a union of the first reach, the second reach,... and the N-th reach, wherein the total reach acquisition unit (20) considers a component in which two of the first to N-th reaches have an overlapping relationship, while ignoring a component in which three or more reaches have an overlapping relationship, as an incremental reach that is a scale of the number of persons who do not contact one information and contact the other information.