JP3239846U - A system that infers ad slot positions based on communication data - Google Patents

Abstract

A system for inferring the location of an advertisement slot based on communication data.
A system for inferring ad slot locations based on communication data includes a first collection module that obtains a plurality of crowd delineation data from the plurality of communication data and a second collection module that obtains a plurality of point-of-interest data. a module and a processor, wherein the processor utilizes a honeycomb hexagonal spatial indexing algorithm to define accuracies of the crowd depiction data and the point of interest data as first and second accuracies, respectively; Integrate the crowd delineation data and the point of interest data based on and retrieve the integrated data to infer the location of at least one slot that simultaneously fits the crowd delineation and the point of interest.
[Selection drawing] Fig. 1

Description

TECHNICAL FIELD The present invention relates to advertisement distribution technology, and more particularly to a system for estimating the location of advertisement slots based on communication data.

Advertisers and media agencies needed to accurately deliver ads to targeted customers to increase the probability of purchasing a product.

The key to accurate marketing, however, is accurately reaching a demographic with the right demographics for distributed content. For example, if a public broadcasting system is installed near a park, and the description of the crowd near the park, for example, many young people who like exercise, many middle-aged and elderly people with pets, etc. Advertisements can be delivered to the system. The attributes of advertisements to be distributed also differ according to the difference in the crowd. Therefore, generating more business opportunities by gaining empathy from people who see advertisements is a major problem to be solved in order to improve the effectiveness of advertisements.

Crowd description data is generally obtained by communication data, or in cooperation with telecommunications carriers, within a 35*35m square area of a designated point, and how many people are in the crowd each day according to gender, age, and tastes. get a depiction. In addition to obtaining a part of the crowd description from communication data, data analysis and statistical techniques are used to select accurate crowd description information from a large amount of data. There are also certain difficulties in obtaining point of interest (POI) data for point descriptions. For example, how to obtain information on how many banks, how many schools, how many parks, and other facilities are in the vicinity. Apart from the costly and manual collection of Point of Interest data, the most effective method is to obtain it using the Google Point of Interest Application Programming Interface (POI API). Although the above-mentioned two types of data can be acquired with the conventional technology, integrating and using communication data of "crowd description" and point-of-interest data of "point description" is a problem that cannot be solved by the conventional technology. Met. The precision of the communication data was 35* ^35m2 centered on the point, and the range was too narrow when collecting the data with the same precision for collecting the graphic data of the point of interest. So not only did you not know how many points of interest you had, you also had to spend big bucks to get the data into the API. For example, if the area of Taiwan is ^{36197000000m2} , and the communication data is 35*35= ^1225m2 , then it will be 36197000000/1225=29533877, with an accuracy of 35*35(m), the data will be sent to the Google Point of Interest application Enter the programming interface. Therefore, I had to enter nearly 30 million times to enter one type. If you enter 90 types, you have to enter 30 million times * 90 types = 2.7 billion times into the Google Point of Interest application programming interface, which is an astronomical number of times and costs. In addition, it is not always possible to collect complete points of interest, the accuracy of 35*35 (m) is too narrow, and the amount of information each time in the Google points of interest application programming interface is incomplete.

Therefore, the inventor of the present invention thought that the above-mentioned drawbacks could be improved, and as a result of earnest studies, he came to the proposal of the present invention to effectively improve the above-mentioned problems with a rational design.

SUMMARY OF THE INVENTION The present invention has been made through intensive research by the inventor in view of the above problems, and its main purpose is to provide a system for estimating the position of an advertisement slot based on communication data. In other words, UBER's hexagonal hierarchical spatial index (Uber H3) is used to integrate two types of data with different accuracies, and integrate traditional crowd description communication data and crowd description point-of-interest data. It solves the problem of not being able to, and makes the position of the advertisement slot more suitable to the demand.

Another object of the present invention is to provide a system for estimating the location of advertisement slots based on communication data. That is, the Uber H3 honeycomb hexagon spatial indexing algorithm is used to expand the search range from one cell to six adjacent cells.

Yet another object of the present invention is to provide a system for estimating the location of advertising slots based on communication data. That is, the Uber H3 honeycomb hexagonal space index algorithm is used to obtain the index value around the cell, quickly search the surrounding area, and ensure that it is a neighboring point to speed up the search speed.

To solve the above problem, one embodiment of the present invention provides a system for estimating the location of advertisement slots based on communication data. The system for estimating the position of the advertisement slot based on the communication data is installed in a server, and the system for estimating the position of the advertisement slot based on the communication data for estimating the position of the advertisement slot, comprising a plurality of communication data; a first collection module for obtaining a plurality of crowd delineation data; a second collection module for obtaining a plurality of points of interest data; a first database coupled to the first collection module for storing the crowd delineation data; a second database connected to the second collection module to store point-of-interest data; and a processor connected to the first database and the second database, the crowd delineation utilizing a honeycomb hexagonal spatial indexing algorithm. Define the precision of the data as the first precision, define the precision of the point of interest data as the second precision, integrate the crowd depiction data and the point of interest data based on the first precision and the second precision, Inferring the location of at least one slot that simultaneously fits the point of interest and retrieving the crowd depiction data or point of interest data in one of the plurality of cells of first or second precision satisfies the demand. and a processor that, if the position of at least one slot cannot be retrieved, uses that cell as a reference point and searches toward surrounding adjacent cells.

According to one aspect of the present invention, the crowd description data is obtained from communication data obtained from at least one carrier, and includes gender, age, and hobby preferences as the crowd description data.

According to one aspect of the invention, the point of interest data is obtained using a point of interest collection application programming interface and includes public facilities in a particular geographic area as subjects of the point of interest data.

According to one aspect of the invention, a first comparison table corresponding to a first accuracy of group description data is stored in a first database.

According to one aspect of the invention, a second comparison table corresponding to the second precision of the point of interest data is stored in the second database.

According to one aspect of the invention, both the first precision and the second precision divide the map into a honeycomb-like mesh composed of a plurality of regular hexagonal cells.

According to one aspect of the invention, when the processor retrieves crowd depiction data in one of the plurality of cells of the first accuracy, the processor further infers an accuracy of one layer above the first accuracy, and Merge the 6 adjacent cells around to the cell and expand the search range.

According to one aspect of the invention, the processor calculates index values for coordinate points at the first precision or the second precision for any one coordinate point.

Since the present invention is configured as described above, it has the following effects.

1 is a block diagram illustrating a system for estimating the location of an advertisement slot based on communication data according to one embodiment of the present invention; FIG. 4 is a flow chart illustrating a method for estimating the position of an advertisement slot based on communication data according to one embodiment of the present invention; FIG. 2 is a schematic diagram showing the Uber H3 honeycomb hexagon spatial indexing algorithm; FIG. 4 is a schematic diagram of integrating crowd description data and point-of-interest data using the Uber H3 honeycomb hexagonal spatial indexing algorithm according to the present invention; FIG. 4 is a schematic diagram of using the Uber H3 honeycomb hexagon spatial indexing algorithm to find adjacent cells of the present invention;

The following clearly and completely describes the technical means of the embodiments of the present invention together with the accompanying drawings of the embodiments of the present invention. Of course, the described embodiments are only some embodiments of the present invention, but not all embodiments. According to the embodiments of the present invention, all other embodiments obtained by those skilled in the art without any inventive effort shall fall within the protection scope of the present invention.

The terms "comprising" and "including" as used in this specification and the appended utility model claims refer to the presence of the features, wholes, steps, operations, elements and/or members described, but not one. or the presence or addition of a plurality of other features, wholes, steps, operations, elements, members, and/or assemblies thereof.

Also, it should be noted that the terminology used in the specification of the present invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the present invention. As used in the specification of the present invention and the appended utility model claims, the singular forms "one", "one" and "said" unless the context clearly indicates otherwise in the context of the context. also includes multiple forms.

Also, the term "and/or" as used in the specification of the present invention and the attached utility model claims refers to any and all possible combinations of one or more of the associated claims and It should be noted that this includes combinations of

The present invention provides a system for inferring the location of advertising slots based on communication data. FIG. 1 is a block diagram showing a system for estimating the location of advertisement slots based on communication data according to the present invention. A system 10 for estimating the location of advertisement slots based on communication data of the present invention is installed on a server 12 . A system 10 for inferring ad slot locations based on this communication data includes a first collection module 13, a second collection module 14, a first database 15, a second database 16, and a processor 18. Configured. The first collection module 13 is signal-connected to the host computer of the carrier 20, and collects the gender, age, and hobby preferences (e.g., what type of web pages are viewed using the mobile phone) of the carrier's 20 customers. is browsing), obtain communication data for the carrier 20, including crowd delineation data such as footsteps of movement. The first collection module 13 is connected to the first database 15 and the collected crowd description data is stored in the first database 15 . The second acquisition module 14 is signal-connected to a host computer of an Internet service provider 22, such as the Google POI application programming interface, which is a public facility (i.e. point of interest) such as a bank, school, park, etc. Get information about a point description, including its location. The second collection module 14 is connected to the second database 16 and the collected point-of-interest data is stored in the second database 16 . Processor 18 is connected to first database 15 and second database 16 . The processor 18 utilizes UBER's hexagonal hierarchical spatial index (Uber H3) algorithm to define the accuracy of the crowd delineation data and point of interest data, and integrates the crowd delineation data and point of interest data.

Please also refer to FIG. 2, which is a flowchart illustrating a method for estimating the position of an advertisement slot based on communication data according to the present invention. In step S10, a plurality of crowd depiction data and a plurality of points of interest data are obtained from the plurality of communication data. At step S12, the processor 18 utilizes the honeycomb hexagon spatial indexing algorithm to define the precision of the crowd descriptive data and associates the crowd descriptive data with the first precision of the honeycomb hexagonal spatial indexing algorithm. The areas covered by the first accuracy and the second accuracy are different. Then, at step S14, the processor 18 defines the precision of the point of interest data, maps the point of interest data to a second precision of the honeycomb hexagon spatial indexing algorithm, and determines the crowd size based on the first precision and the second precision. Integrate descriptive data and point-of-interest data. Finally, in step S16, after integrating the crowd delineation data and the point of interest data, the processor 18 searches for keywords to integrate the data and locates at least one slot that simultaneously fits the crowd delineation and the point of interest. Guess.

The honeycomb hexagonal spatial indexing algorithm used in this invention is the Uber H3 algorithm, which divides the earth's space into identifiable cells, and encodes the latitude and longitude into a hexagonal mesh index with H3, which is shown in Figure 3. , a plurality of cells 30 form a honeycomb mesh. Taking a cell 30 with a target point (lat, lon) as an example, this cell 30 is surrounded by 6 adjacent cells 30, these 7 cells 30 forming a larger hexagon (Fig. 4 reference). The H3 index value of each point in the world is unified, and it is divided into 16 levels (levels), each level has a different accuracy. Let 15 be the smallest diameter (side length is 0.5m), the smaller the diameter, the higher the accuracy, so step 15 is the most accurate. Index values are extrapolated upwards and downwards, eg, the area covered by a hexagon in stage 10 is 7 times that of stage 11 (1884*7=13188). The cell 30 index values (cell id) are mutually inferred, and if the step 11 index value is 8b4ba010e0a1fff, the H3 algorithm infers the step 10 index value as 8a4ba010e0a7fff. Therefore, the Uber H3 algorithm mutually infers and mixes data of two different precisions as data of the same precision.

Based on the Uber H3 principles discussed above, the processor 18 matches the precision of the crowd description data to the first precision 32 of the honeycomb hexagon spatial index algorithm, and the precision of the point of interest data to the honeycomb hexagon spatial index algorithm, as shown in FIG. It corresponds to the second precision 34 of the angular space indexing algorithm and integrates two different precision points of interest data and crowd description data. Since the accuracy of the communication data is 35 m ² , it is most preferable to correspond to the H3 index value of matching stage 11 (approximately 25 m radius). Therefore, the first accuracy 32 is step 11. In addition, the precision of the most important point of interest data is step 8 (radius of about 460m) with second precision 34, which greatly reduces the number of calls to the Google Point of Interest application programming interface. Since the area of Taiwan is 36197000000m ² , the area of each H3 level 8 hexagon is 646367m ² , which is 36197000000/646367=56000 times. In other words, it takes only 56000*90=5040000 times to collect all 90 classifications until they are full, which is 2.7 billion times as described in the prior art, and 5 million times after using the H3 algorithm. , the quality of the collected point-of-interest data does not decrease. This solves the problem of mixing and using data with two different precisions. Processor 18 computes index values for coordinate points at first precision 32 or second precision 34 for any one coordinate point and infers the position of at least one slot that simultaneously fits the crowd description and the point of interest. . In addition, the present invention applies the H3 algorithm to further save the number of calls to the point-of-interest application programming interface.

The idea of this embodiment is that the first database 15 stores the crowd depiction data, and the crowd depiction data also stores the first comparison table 152 corresponding to the first accuracy 32 in the first database 15 . Similarly, in addition to storing the point-of-interest data, the second database 16 also stores a second comparison table 162 corresponding to the second accuracy 34 in the second database 16 . The first comparison table 152 includes columns such as crowd depiction data points, index values, and first accuracy. The second comparison table 162 includes columns such as points of point of interest data, index value, and second accuracy.

In addition, the index value of the H3 algorithm is estimated from a small index value to a large index value, thereby performing the index overlay function. When the processor 18 retrieves crowd depiction data in one cell 30 of the first precision 32, it further infers the precision of one layer above the first precision 32, and in this way the 6 cells surrounding the cell 30. Merge two neighboring cells into cell 30 to expand the search range. For example, if we are currently using stages 8 to define this communication data, then of course going one layer up to stage 7 will increase the diameter even further, and we will get a wider range, and the surrounding six stages 8 cell data into a wider range of data indices. Taking FIG. 5 as an example, if only one convenience store is found in one center cell 36, the quantity is insufficient, so the next level is moved up and the surrounding 6 adjacent cells are also merged. and a larger upper cell 38, allowing more convenience stores near the target point (lat, lon) to be searched.

Similarly, the characteristics of the surrounding indices can be known based on the H3 algorithm, and if the index value of the center cell 36 is known, the index values of the six surrounding cells can also be known. Thus, when processor 18 retrieves crowd depiction data or point-of-interest data in one cell 30 of either first precision 32 or second precision 34, it searches cell 30 if it fails to retrieve a slot location that meets the demand. Search toward the surrounding adjacent cells as a reference point. This method ensures that the retrieved data are neighboring points, eliminates the need to calculate the linear distance between points of interest by yourself, and further avoids the drawback of scanning every point of interest in the prior art. . This method accelerates the indexing of the data and greatly improves the search speed.

The idea of this embodiment is that the slot device is an Android TV Box, a digital signage, a display screen at a bus stop or a convenience store, a screen in an elevator, etc., and can deliver a specified advertisement at a specified time. The present invention uses the Uber H3 algorithm to integrate the accuracy of the crowd description data and the point of interest data, and then quickly finds the position of the desired advertisement slot. For example, if the desired slot location is a point where there are many young men who like movies, is near a shopping area, and is a point where there is a department store, this condition is determined based on the communication data provided by the present invention. Applies to systems and methods for estimating the position of

In summary, the system and method for estimating the location of advertising slots based on communication data of the present invention integrates crowd description data and point-of-interest data with different accuracies through the index results of Uber H3 algorithm, Effectively save the number of times of data purchase and collection. Also, just by confirming the geographic coordinates (lat, lon) of one point on the earth, we can know all the index values of that coordinate in stages 1 to 15, thereby integrating various data. The data of various stages can be extracted, mixed and used.

Although the embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and design changes and the like are included within the scope of the present invention.

REFERENCE SIGNS LIST 10 system for estimating position of advertisement slot based on communication data 12 server 13 first collection module 14 second collection module 15 first database 152 first comparison table 16 second database 162 second comparison table 18 processor 20 carrier 22 Internet Service Provider 30 Cell 32 First Precision 34 Second Precision 36 Center Cell 38 Upper Cell

Claims (10)

A system installed on a server for estimating the position of an advertisement slot based on communication data for estimating the position of the advertisement slot,
a first collection module that obtains a plurality of crowd depiction data from a plurality of communication data;
a second collection module that obtains multiple points of interest data;
a first database connected to said first collection module for storing said crowd depiction data;
a second database connected to the second collection module to store these points of interest data;
a processor connected to the first database and the second database, wherein a precision of the crowd description data is defined as a first precision using a honeycomb hexagon spatial index algorithm; defining an accuracy as a second accuracy; integrating said crowd delineation data and said point of interest data based on said first and said second accuracies; inferring slot locations and retrieving said crowd descriptive data or said point of interest data in one of said plurality of cells of said first precision or said second precision of said at least one slot meeting demand; A system for estimating the position of an advertisement slot based on communication data, comprising: a processor that, when a position cannot be searched, uses the cell as a reference point and searches for surrounding adjacent cells. . 2. A system for inferring advertising slot locations based on communication data as recited in claim 1, wherein said crowd descriptive data are obtained among said communication data obtained from at least one carrier. 2. The system for inferring advertising slot locations based on communication data of claim 1, wherein the crowd description data includes gender, age, and hobby preferences. 2. The system of claim 1, wherein said points of interest data are obtained using a point of interest collection application programming interface. 2. A system for estimating the position of an advertisement slot based on communication data according to claim 1, wherein the target of said point of interest data includes public facilities in a specific area. 2. Inferring advertising slot locations based on communication data according to claim 1, wherein a first comparison table corresponding to said first accuracy of said crowd depiction data is stored in said first database. system to do. 2. A second comparison table corresponding to said second accuracy of said point of interest data is stored in said second database for determining ad slot locations based on communication data according to claim 1. guessing system. 2. The first precision and the second precision both divide the map into a honeycomb-like mesh composed of the plurality of cells of regular hexagons. guessing system. When the processor retrieves the crowd depiction data in one of the plurality of cells of the first precision, the processor further infers a precision of one higher layer of the first precision, and 9. The system of claim 8, wherein some six adjacent cells are merged into the cell to expand the search range. 2. The processor calculates an index value of the coordinate point at the first precision or the second precision for any one of the coordinate points. A system for estimating location.

Images