TWI522955B - A method for determining optimal business type and location using internet data - Google Patents

Description

Ways to use network data to determine the type and location of the most suitable business

The present invention relates to an analysis method, and more particularly to an analysis method capable of recommending a shop type according to a store location or recommending a store location according to a store type.

In this era of highly circulated information, the open data of various regions is becoming more and more abundant, including the types of businesses, government agencies, public facilities, and real estate information of a certain location. It is a very convenient source of information for the general public or foreign tourists. Of course, this is also true for those who want to open a store.

However, when the industry is faced with the choice of opening a store, it is necessary to consider the location and type of the store. Often it is already a good place to choose, but I don't know what type of store is suitable for the location; or I have decided to open a certain type of store, but I am not sure where to open it.

Therefore, the present invention proposes a method for providing an ideal store location ranking or an optimal store type ranking for an operator who wants to open a store by analyzing the open data, as a reference for the establishment to make a store decision.

In accordance with the above needs, the present invention provides a method for determining an optimal business type and location using network data, including obtaining a plurality of first environmental parameters corresponding to a store location from a public repository, and the first environmental parameters After performing a weight adjustment, establishing a first vector space, and then obtaining at least one indicator instance in at least one store type from the public database, and a plurality of second environment parameters corresponding to each indicator instance, and according to the After the second environment parameter establishes a second vector space, a similar value between the first vector space and the second vector spaces is separately calculated.

Wherein, the location of the store can be specified by a user, or the location and location of the store are given by the user, and the corresponding store location is searched.

In an embodiment, the method for determining the optimal business type and location by using the network data further includes arranging the similar values according to the size to obtain a median value, and setting the median value to correspond to the type of the store type. Moderate scores, and are ranked according to the size of each suitability score to determine the ranking of the store type.

In an embodiment, the method for determining the optimal business type and location by using the network data before obtaining the first environmental parameters includes obtaining a preset shop type and a region from the user, and obtaining multiple regions in the region. Set the location of the store.

For example, in the embodiment, where the indicator examples correspond to the preset shop type, the method for determining the optimal business type and location by using the network data further includes arranging the similar values according to the size to obtain a median value. And setting the median value as a suitability score of the corresponding store location, and ranking according to the suitability score size to determine the ranking of the store locations.

In one embodiment, the public database is a government public information database or a Google local information database.

In an embodiment, the indicator example is an example of all the stores belonging to the type of the store from the public database, and the surplus is ranked in order of the surplus, and the surplus is ranked in the previous example as an example of the indicator of each store type. .

In an embodiment, the weight adjustment of each environmental parameter is performed by a term frequency-inverse document frequency (TF-IDF) algorithm.

In an embodiment, the similarity value is a cosine value of the first vector space and the second vector space (ie, a product value in the unit vector).

Therefore, the present invention can obtain an environmental parameter of an environmental parameter and an indicator example of a store location from a public database, and obtain a store type or a store location by comparing environmental parameters of the store location with environmental parameters of the indicator example. Appropriate rankings to help users determine the best store type or best location.

10‧‧‧Analysis system

12‧‧‧ Client

14‧‧ ‧ public database

140‧‧‧Environmental Database

142‧‧‧Exemplary database

B‧‧ Bank

C‧‧‧Convenience Store

L‧‧‧ Location

M‧‧‧MRT exports

P‧‧‧Parking

R‧‧‧ range

S‧‧Shopping Mall

Figure 1 is a schematic diagram showing the module connection relationship of the analysis system to which the present invention is applied.

Figure 2 is a reference diagram of the service facilities within a predetermined range around a location.

Figure 3 is a flow chart of the method for determining the most suitable store type by using the network material to determine the optimal business type and location.

Figure 4 is a flow chart of the method for determining the optimal business type and location by using the network data in the present invention to determine the optimal location of the store.

Please refer to FIG. 1 to FIG. 2 . FIG. 1 is a schematic diagram showing the module connection relationship of the analysis system applied in the present invention, and FIG. 2 is a reference diagram of the service facilities in a predetermined range around a place. The method for determining the optimal business type and location using the network data is performed by an analysis system 10 that is coupled to a client 12 at one end and to a public repository 14 at the other end. The client 12 is a user interface for inputting or reading information, such as an application installed on a smart phone, a tablet or a personal computer.

The public database 14 may be a public information database provided by a local government unit or a Google local information database. Those skilled in the art of the present invention know that the Google Local Information Library can access data through the Google Places API (Google Place API), and this will not be repeated for the Google Local Information Service interface.

The public repository 14 can include an environmental repository 140 and an example repository 142. The environmental database 140 is used to store information about the types of all service facilities (or types of services), the name of the service facility (or type of business), the location of the service facility (or type of business), and real estate information (eg, number of pings, Face width or house age); that is, when the user wants to know the environmental conditions in the vicinity of a certain store location, he can obtain all the service facilities in a range near the store location from the environmental database 140. Information, where the range can be set by the user or preset by the system (for example, within a radius of 300 to 500 meters).

Suppose the user wants to check the service in a range R near the store location L. Facilities, you may see that there are two parking lots P, a convenience store C, a bank B, a MRT exit M and a shopping mall S near the location L. Of course, the names and quantities of these service facilities are It can be converted into a parameter value to clearly indicate a plurality of environmental parameters near the location L of the store, such as: P(2), C(1), B(1), M(1), and S(1).

The sample database 142 is an installation record and business information of the storage store, and is used as an example of the analysis shop placement point of the present invention. The location of the store can correspond to the location in the aforementioned environmental database 140, so that the nearby environment can be known. The number of conditions or service facilities. The business information includes the revenue of the sample store. Generally speaking, the indicator value of revenue is Profits. The higher the surplus, the better the revenue of the sample, and the more relevant the information is. The value of the reference can be used as an example of the indicator.

In the present invention, the analysis system 10 is configured to receive user input from the client 12 and obtain the required information from the public database 14 according to the method of determining the optimal business type and location using the network data according to the present invention. Then, the information is weighted, the vector space, the sorting operation is established, and the operation result is transmitted to the client 12 for presentation to the user through the client 12.

Please refer to FIG. 1 to FIG. 3 together. FIG. 3 is a flow chart of the method for determining the optimal business type and location by using the network data to determine the optimal store type. As shown in the figure, the method for determining the optimal business type and location by using the network data of the present invention includes the step S20: obtaining a plurality of first environment parameters from a public database in step S20, wherein the plurality of first environmental parameters are obtained from the user. Environmental parameter system Corresponding to the location of the store, S22 performs a weight adjustment on each of the first environmental parameters, and establishes a first vector space, and S23 obtains at least one indicator of at least one store type from the public database, and corresponding indicators. According to the plurality of second environment parameters of the example, S24 establishes a second vector space according to the second environment parameters, and S25 respectively calculates a similarity value between the first vector space and each second vector space, and S26 similarizes After the values are arranged according to the size, a median value is obtained, S27 sets the fitness value of the corresponding store type for the median value, and S28 is arranged according to the size of each suitability score to determine the store type. Appropriate ranking, in which the first place is the most suitable store type for this store location, the second place is the second most suitable store type for this store location, and so on.

In the present invention, when the user wants to know that a certain store location is suitable for opening a specific store type, the user can input the location of the store to be inquired through the device of the user terminal 12, and the analysis system 10 will Read the location of the store from the user terminal 12, and obtain, according to step S21, a plurality of first environmental parameters from a public repository 14 about a predetermined range of the location of the store, wherein each of the first environments The parameter is the environmental parameter corresponding to the location of the store. When the user inputs the address of the store location L, the analysis system 10 obtains, for example, the parking lot P (2) and the convenience store C (1) from the public database 14 . Environmental parameters such as Bank B (1), MRT Exit M (1) and Shopping Mall S (1) to indicate the type and number of service facilities in the vicinity of the store location. In addition, the analysis system 10 also obtains information such as the number of pings, the width of the face, or the age of the house as part of the environmental parameters based on the real estate information of the store location L in the public database 14. For example, if the location of the store L occupies the floor The number is 20 pings, then The analysis system 10 will obtain the number of pings (20) as its parameters.

After obtaining the first environmental parameters of the location of the store, the analysis system 10 performs the weighting of the environment parameters of the store location by the TF-IDF (term frequency-inverse document frequency) algorithm according to step S22. Adjustment, that is, reducing the weight of environmental parameters that are not highly discriminating among all environmental parameters, and increasing the weight of environmental parameters with higher discriminating power, for example, there will be almost fire hydrants near each individual store location in a region. Therefore, it can be seen that if the location of a certain fire-fighting shackle is not meaningful, the weight of the fire hydrant environmental parameter will be reduced; otherwise, if there is only a few shops in a certain area, there is a parking lot nearby. That means that if there is a parking lot near the location of the store, the weight of the parking lot will be increased. Then, the analysis system 10 can establish a first vector space corresponding to the location of the store according to the weighted first environment parameters, that is, vectorize the first environment parameters to facilitate subsequent calculations. .

Obtaining each first environmental parameter and the first vector of the location of the store After the space, the analysis system 10 obtains at least one indicator example of each of the store types from the public database 14 according to step S23. Wherein, as described above, in the public database 14, the store type is stored in an indexed manner, and each of the examples is stored in the store type index according to the store type of the store, and the examples are based on The surpluses are arranged in order from the most, and according to the rankings of the examples, the examples in the ranking are the examples of the indicators, that is, the examples of the higher the surplus in the store type. In this embodiment, the examples of the indicators are It is an example of the top 3%. Accordingly, the present invention That is to say, by using these examples of indicators, the suitability of each corresponding store type for the location of the store is compared.

When the analysis system 10 obtains an example of each of the set of stores, a plurality of second environmental parameters of the sample of the indicators are also obtained, that is, environmental parameters corresponding to each of the indicator examples, and these are second The environment parameter establishes a second vector space. After the analysis system 10 establishes or obtains the second vector space, the similarity between the first environment parameter and each of the second environment parameters may be compared according to step S25. In the example, the similarity is calculated by performing the cosine operation (ie, the product value in the unit vector) of the first vector space and the second vector space of each index example respectively, and the remaining chord values are corresponding to the single indicator example. One of the locations of the store is similar. Among them, the higher the similarity value, the higher the similarity between the surrounding environment of the store location and the surrounding environment of this indicator example. If the store type corresponding to this indicator example is opened at the store location, it will be quite Suitable for.

However, in order to prevent the calculation result of the present invention from being affected by some extreme cases, the analysis system 10 of the present invention will belong to the same design according to step S26 after obtaining the similarity value of each indicator example of the store type. The sample indicators of the store type are sorted according to their similarity values, and after sorting, the median value is taken according to the number of indicator examples, that is, the similarity value is an example of all indicators in the store type. Intermediate rankings to provide more objective data. Next, the analysis system 10 sets the median value to the suitability score of the store type according to step S27, and then arranges the power according to the suitability score of each store type to pass through the aforementioned client. 12 will rank the rankings of these store types Present to the user.

Please refer to FIG. 1 , FIG. 2 and FIG. 4 together. FIG. 4 is a flow chart of the method for determining the optimal business type and location by using the network data in the invention to determine the optimal location of the store. In the present embodiment, in addition to the above, the present invention can provide a user to input a store location to query a suitable store type, and can also provide the user with the desired store type to find at least one of the regions suitable for the store type. The store location is set and ranked, that is, the store type in the embodiment is obtained by the analysis system 10 according to the input of the user through the user terminal 12, and the analysis system 10 is based on the store type. Analyze the suitability of multiple store locations in a region to provide a user reference for calculating the suitability ranking of each store location.

Therefore, the method for determining the optimal business type and location by using the network data of the present invention includes the steps S30: obtaining a preset shop type and a region from the user, S31 obtaining a plurality of store locations in the region, and S32 self-published materials. The library obtains a plurality of first environment parameters, wherein the first environment parameters correspond to the store location, S33 performs a weight adjustment on each first environment parameter, and then establishes a first vector space, and the S34 self-published database respectively Obtaining at least one indicator example of at least one store type, and a plurality of second environment parameters corresponding to each indicator instance, wherein the indicator instances correspond to a preset store type, and S35 establishes a second environment parameter according to the second environment parameter The second vector space, S36 respectively calculates a similarity value between the first vector space and each of the second vector spaces, and S37 arranges the similar values according to the size, and obtains a median value, and S38 sets the median value. A suitability score for the corresponding store location and S39 are arranged according to the size of the suitability score to determine the location of the store. Ranking.

For example, if the user wants to find a place in a certain area suitable for opening a coffee shop, the user can enter the name of the area in the area of the area through the user terminal 12 (for example, a new store area), and When the coffee shop is entered in the field of the store type, the analysis system 10 will read from the public database 14 a plurality of store locations and stores in the area that can be set up according to the input of the user. Locating a plurality of first environmental parameters in a range, and establishing a first vector space according to the first environmental parameters, and then obtaining a plurality of second environmental parameters of the indicator instance from the public database 14 according to step S34. The examples of these indicators correspond to the preset shop type input by the user, that is, the sample type of the shop type is a coffee shop. Similarly, as described above, the analysis system 10 also establishes a second environment parameter for the second environment parameter. The second vector space.

Then, after having the first vector space of each store location and the second vector space of each indicator instance, the analysis system 10 can compare the first environmental parameters and the second environmental parameters according to step S36. Similarity, by performing the cosine operation on each of the first vector spaces and each of the second vector spaces, respectively, the similar values of the respective store locations for each indicator example can be obtained. Among them, the higher the similarity value, the higher the similarity of the surrounding environment of this store location with the surrounding environment of this indicator example, the location of this store should be quite suitable for the store corresponding to this indicator example. Types of.

Similarly, in order to prevent the calculation result of the present invention from being affected by some extreme cases, the analysis system 10 of the present embodiment obtains the similar values. After that, each indicator example is sorted separately, and the median value of the number of each indicator example is taken after sorting, so that the median value is set as the suitability score of the store type. Next, the analysis system 10 sets the suitability scores corresponding to the cafes in the store locations to the suitability scores of the store locations according to step S39, and determines the location of the store locations according to the size. Suitability ranking.

For example, if the store type is a coffee shop, the examples are Starbucks, Braun and Yike, and the location of a store is 98 points for Starbucks and 62 for Braun. Yi Ke's suitability score is 80 points. After sorting, the order should be Starbucks (98 points), Chake (80 points) and Burlang (62 points). The median value should be Yi Ke (80 points). That means that the location of this store is 80 points for the coffee shop.

After calculating the suitability scores for the cafes at each store location, the analysis system 10 sorts the suitability scores based on these suitability scores, typically presented to the user in descending order. In other words, the user may first see the address of the store in the area that is most suitable for opening a coffee shop, and then view the address of the lower-selling store location in order of the score, so that it is clear Let the user know which store location is most suitable.

Therefore, the present invention can obtain an environmental parameter of an environmental parameter and an indicator example of a store location from a public database, and obtain a store type or a store location by comparing environmental parameters of the store location with environmental parameters of the indicator example. The suitability ranking is to provide a reference for the user to determine the best store type or the best store location.

Claims (9)

A method for determining an optimal business type and location by using network data, comprising: obtaining a plurality of first environmental parameters from a public database, wherein the first environmental parameters correspond to a store location; and the first environment After the parameter is adjusted by a weight, a first vector space is established; at least one indicator instance of at least one store type is obtained from the public database, and a plurality of second environment parameters corresponding to each indicator instance are obtained; a second environment parameter, establishing a second vector space; and separately calculating a similarity value between the first vector space and each of the second vector spaces, wherein the similarity values are the first vector space and each of the first The cosine of the two vector spaces. The method for determining the optimal business type and location according to the use of the network data described in the first application of the patent scope includes: arranging the similar values according to the size to obtain a median value; setting the median value to correspond a suitability score of the store type; and ranking according to each of the suitability scores to determine the ranking of the store types. The method for determining the optimal business type and location by using the network data as described in the first application of the patent application, before obtaining the first environmental parameters, includes: obtaining a predetermined shop type and a region from a user; and Get multiple locations in the area. The method for determining the optimal business type and location by using the network data according to item 3 of the patent application scope, wherein the indicator examples correspond to the preset shop type. The method for determining the optimal business type and location by using the network data according to item 4 of the patent application scope includes: selecting the similar values according to the size to obtain a median value; setting the median value to correspond a suitability score of the location of the store; and a ranking according to the size of the suitability score to determine the ranking of the locations of the stores. The method of determining the optimal business type and location by using the network data as described in item 1 of the patent application scope, wherein the location of the store is set by a user. The method for determining the optimal business type and location by using the network data according to the first application of the patent application scope, wherein the public database is a government public information database or a Google local information database. A method for determining an optimum business type and location according to the use of network data as described in claim 1 of the scope of the patent application, wherein each of the indicator examples is at least one example of all types belonging to the store from the public database, The profit of this example The remainder is arranged in order, and the examples ranked above are examples of each indicator. The method for determining the optimal business type and location by using the network data according to claim 1 of the patent application scope, wherein the weight adjustment of the first environmental parameters is performed by a TF-IDF algorithm.