JP6908308B2 - Sales support equipment and sales support program - Google Patents

Description

The present invention relates to a sales support device and a sales support program for analyzing corporate management information, predicting trends in real estate transactions, and supporting sales.

As a system for analyzing the management information of a company, a settlement analysis system such as Patent Document 1 is known. The system of Patent Document 1 morphologically analyzes financial statement-related information such as financial statements, and automatically extracts the cause of a specific event from a pattern in which each element is labeled.

Japanese Patent No. 6155409

However, although the system of Patent Document 1 does not describe the handling of real estate-related information, even if real estate-related information can be analyzed as a specific event, is there a possibility of real estate transaction demand simply by extracting the factors? It's hard to predict. In the past, it took a lot of time and effort to find a company with potential real estate transaction demand as a sales destination, and we had to rely on the experience and intuition of skilled salespeople. In consideration of such circumstances, the present invention is a sales support device and sales support capable of predicting a company having a possibility of real estate transaction demand by analyzing management information and automatically selecting and displaying it. The purpose is to provide a program.

In order to solve the above problems, the sales support device of the present invention includes a storage unit that stores specific keywords related to real estate transactions, an information acquisition unit that acquires management information of a plurality of companies, and an information acquisition unit. It is determined by the keyword judgment unit that determines whether or not the acquired management information includes the specific keyword stored in the storage unit, and the keyword judgment unit that the management information includes the specific keyword. It includes a company selection unit that selects a company as a business partner company candidate, and an output unit that outputs display data that displays the business partner company candidate.

According to the sales support device of this aspect, real estate transaction demand is obtained by acquiring management information of a plurality of companies and performing an analysis to determine whether or not a specific keyword related to real estate transaction is included in the management information. Can predict potential companies. Furthermore, companies that are determined to include specific keywords in management information are selected as business partner company candidates and display data is output, so companies with potential real estate transaction demand are automatically selected as business partner company candidates. It can be displayed on a terminal device or the like. As a result, candidates for sales positions can be selected without relying on the experience and intuition of skilled sales staff. Therefore, the time and effort required to select a candidate for a business partner can be significantly reduced, and sales activities can be made more efficient.

In a preferred embodiment of the present invention, the management information includes different types of management information, the specific keyword includes different types of specific keywords, and the keyword determination unit changes the type of the specific keyword according to the type of management information. To determine a specific keyword. According to this aspect, by changing the type of a specific keyword according to the type of management information, the analysis result by the specific keyword judgment of the management information is more effective than the case where the same specific keyword is used for different types of management information. It is possible to enhance the effect of selecting candidates for business partners. Therefore, it is possible to improve the accuracy of selecting business partner company candidates.

In a preferred embodiment of the present invention, the management information includes the first management information including the management plan information and the second management information which is a type of management information different from the first management information and includes the officer information, and is specified. The keywords include the first specific keyword for predicting the trend of real estate transactions and the second specific keyword for searching for a company having an officer who is a key person in real estate transactions. The specific keyword is determined by the first specific keyword, and the second management information is determined by the second specific keyword.

According to this aspect, it is possible to select a business partner company candidate whose approach is changed according to the type of management information. Since the first management information including the management plan information is judged by the first specific keyword for predicting the trend of real estate transactions, it is possible to select a business partner company candidate by the approach of whether or not the company has real estate transaction demand. On the other hand, the second management information including officer information is judged by the second specific keyword for searching for a company that has an officer who is a key man in real estate transactions, so it is easy for the company to understand the necessity of real estate liquidation. Candidates for business partners can be selected by different approaches, including whether or not there is a real estate company. As a result, it is possible to greatly enhance the effect of linking the analysis result by the specific keyword judgment according to the type of management information to the selection of a promising business partner company candidate. Therefore, the accuracy of selecting business partner company candidates can be significantly improved.

In a preferred embodiment of the present invention, the display data includes data for separately displaying the business partner company candidate selected by the first management information and the business partner company candidate selected by the second management information. According to this aspect, since the business partner company candidate based on the first management information and the business partner company candidate based on the second management information are displayed separately, it becomes easy to understand which management information selected the business partner company candidate. This makes it easier to formulate sales strategies for companies that are candidates for business partners.

In a preferred embodiment of the present invention, the storage unit stores in advance a plurality of keyword groups in which specific keywords are grouped and weight coefficients associated with them according to the characteristics of corporate management related to real estate transactions. The keyword determination unit is provided with and displays a score acquisition unit that calculates and acquires the score of the management information based on the number of appearances of the specific keyword included in the management information and the weighting coefficient of the keyword group including the specific keyword. The data includes data for displaying business partner company candidates in descending order of management information score.

According to this aspect, a plurality of keyword groups in which specific keywords are grouped and a weighting coefficient associated with these are stored in advance according to the characteristics of corporate management related to the real estate transaction. It is possible to analyze management information that reflects the high level of interest in real estate transactions from the insight of related corporate management. This makes it easier to select influential companies as business partners. Moreover, the display data makes it possible to display the business partner company candidates on the terminal device or the like in descending order of the score of the management information. Therefore, it is possible to display companies so that the more promising business partner candidates with a high possibility of real estate transaction demand are, the higher the ranking.

In a preferred embodiment of the present invention, the storage unit stores a learned model learned to output a score by inputting information on the number of occurrences of a specific keyword, and the keyword determination unit is included in the management information. It is equipped with a score acquisition unit that acquires the score output from the learned model as the score of the management information by inputting the information of the number of occurrences of the specific keyword to be displayed. Contains data to display candidates. According to this aspect, by using the trained model for the specific keyword determination process of the management information and inputting the information of the number of occurrences of the specific keyword, the weight is optimized by the trained model without setting the weight coefficient in advance. You can get the converted score. As a result, the analysis results of management information using specific keywords can be effectively linked to the selection of promising business partner candidates. Therefore, it is possible to improve the accuracy of selecting business partner company candidates.

In a preferred embodiment of the present invention, the storage unit stores a learned model learned to output a score by inputting management information and information of a specific keyword, and the keyword determination unit identifies the management information and the specific keyword. It has a score acquisition unit that acquires the score output from the learned model as the score of the management information by inputting the keyword information, and the display data is the data that displays the business partner company candidates in descending order of the management information score. including. According to this aspect, the trained model can be used for the specific keyword determination process of the management information, and the score can be obtained from the output of the trained model by inputting the management information and the specific keyword information. This makes it possible to simplify the specific keyword determination process.

In a preferred embodiment of the present invention, the storage unit includes a score database for accumulating the scores of management information for each company, and the company selection unit has the score of the management information acquired by the score acquisition unit equal to or higher than a predetermined threshold. Select a company with a higher score than the previous score obtained from the score database as a candidate for a business partner. In a company, when the management policy changes, the keywords used in the management information are likely to change. Therefore, since it can be inferred that the management policy has changed due to the change in the score of the management information, the change in the trend of real estate transactions can also be predicted. Companies with a higher score than the previous score are considered to be the best time to open a business because the number of specific keywords related to real estate transactions is increasing.

According to this aspect, a company whose management information score is equal to or higher than a predetermined threshold value and is larger than the immediately preceding score is selected as a candidate for a business partner. You can even guess if there was a change in. Moreover, companies whose management policy has changed in the direction of increasing the score can be preferentially extracted and displayed. Therefore, the time and effort required to select a candidate for a business partner can be significantly reduced, and sales activities can be made more efficient than when a company is selected based only on the latest management information score.

In a preferred embodiment of the present invention, the storage unit includes a score database that accumulates the score of the management information for each company and the number of appearances of the specific keyword of the management information for each keyword group. The score of the management information acquired by the acquisition department is equal to or higher than a predetermined threshold, and the number of appearances of the specific keyword of the management information acquired by the score acquisition department is higher than the number of appearances of the specific keyword of the management information immediately before the acquisition from the score database. Select a company with a large keyword group as a business partner candidate. As described above, the keyword group divides specific keywords according to the management characteristics related to real estate transactions. Therefore, it is possible to know in which direction the management policy was changed depending on which keyword group appeared more frequently.

According to this aspect, a company having a keyword group whose management information score is equal to or higher than a predetermined threshold value and whose number of appearances of a specific keyword is larger than the number of appearances immediately before is selected as a candidate for a business partner, so that the management policy changes. Not only can you guess if it happened, but you can also guess the direction of change in management policy. Therefore, it becomes easier to formulate an appropriate sales strategy that matches the direction of the management policy as well as the timing of sales.

In a preferred embodiment of the present invention, the management information includes a first management information including the management plan information and a second management information which is a type of management information different from the first management information and includes the officer information, and is open for business. previous company candidate includes a first operating partners candidate and the second operating partners candidates, companies selecting unit, the score of the first management information is equal to or greater than the predetermined threshold acquired by the score acquisition unit, the score acquisition unit The first business partner company is a company that has a keyword group in which the number of occurrences of the first specific keyword of the first management information acquired in is larger than the number of appearances of the first specific keyword of the first management information immediately before being acquired from the score database. The score database is the number of occurrences of the second specific keyword of the second management information immediately before being selected as a candidate and the score of the second management information acquired by the score acquisition department is equal to or higher than a predetermined threshold and immediately before being acquired by the score acquisition department. A company having a keyword group larger than the number of appearances of the second specific keyword in the second management information obtained from is selected as a candidate for the second business partner.

According to this aspect, a company having a keyword group in which the score of the first management information is equal to or higher than a predetermined threshold value and the number of appearances of the first specific keyword is larger than the number of appearances immediately before is selected as the first business partner company candidate. Not only can you guess whether there has been a change in management policy, but you can also guess the direction of change in management policy. Moreover, since the change in the management policy can be inferred from the first management information including the management plan in which the trend of the real estate transaction is easily described, the trend of the real estate transaction can be predicted. Therefore, it becomes easier to formulate an appropriate sales strategy that matches the direction of the management policy as well as the timing of sales. In addition, since a company having a keyword group in which the score of the second management information is equal to or higher than a predetermined threshold value and the number of appearances of the second specific keyword is larger than the number of appearances immediately before is selected as the candidate for the second business partner, the management policy Not only can you guess if there has been a change, but you can also guess the direction of change in management policy. Moreover, since it can be inferred from the second management information including the officer information, it is possible to infer whether or not a human resource who is easy to accept real estate transactions has become an officer. Therefore, it becomes easier to formulate an appropriate sales strategy that matches the direction of the management policy as well as the timing of sales.

In order to solve the above problems, the sales support program of the present invention acquires management information of a plurality of companies, and the acquired management information includes specific keywords related to real estate transactions stored in the storage unit. A specific keyword judgment is performed to determine whether or not a specific keyword is included, a company determined to include a specific keyword in the management information is selected as a business partner company candidate, and display data for displaying the business partner company candidate is output. Let the computer do it. By executing the program of this aspect, the computer can function as a sales support device.

According to the present invention, it is possible to analyze the management information of a company and automatically select and display a company having a possibility of real estate transaction demand as a business partner company candidate, so that a promising business partner company candidate can be selected. Without relying on experience and intuition, the time and effort required before sales can be greatly reduced, and sales activities can be streamlined.

It is a figure which shows the structure of the sales support system of 1st Embodiment. It is a block diagram which shows the specific configuration example of the sales support system of 1st Embodiment. It is a figure which shows the configuration example of a keyword database. It is a figure which shows the configuration example of the score database. It is a figure which shows the specific example of the display screen. It is a flowchart which shows the specific example of the management information analysis processing. It is a flowchart which shows the specific example of the specific keyword determination processing. It is a block diagram which shows the specific configuration example of the sales support system of 2nd Embodiment. It is a figure which shows the specific example of a trend keyword. It is a figure which shows the specific example of a key man keyword. It is a figure which shows the specific example of the equipment status keyword. It is a figure which shows the specific example of the 1st score information. It is a figure which shows the specific example of the 2nd score information. It is a flowchart which shows the specific example of 1st management information analysis processing. It is a flowchart which shows the specific example of the 1st specific keyword determination processing. It is a flowchart which shows the specific example of the 2nd management information analysis processing. It is a flowchart which shows the specific example of the 2nd specific keyword determination processing. It is a figure which shows the specific example of the display screen of 2nd Embodiment. It is a figure which shows another specific example of the display screen of 2nd Embodiment. It is a block diagram which shows the specific configuration example of the sales support system of 3rd Embodiment. It is a figure which shows the structural example of the trained model of 3rd Embodiment. It is a figure which shows the other configuration example of the trained model of 3rd Embodiment. It is a block diagram which shows the specific configuration example of the sales support system of 4th Embodiment. It is a figure which shows the structural example of the 1st trained model. It is a figure which shows the structural example of the 2nd trained model. It is a figure which shows the other structural example of the 1st trained model. It is a figure which shows the other configuration example of the 2nd trained model. It is a figure which shows the other configuration example of the trained model of 4th Embodiment. It is a flowchart which shows the specific example of the business destination company candidate selection process of 5th Embodiment. It is a figure which shows the comparative example of the transition of the 1st score. It is a figure which shows the specific example of the transition of the 1st score. It is a flowchart which shows the specific example of the 1st business partner company candidate selection process which concerns on 1st modification. It is a figure which shows the display example of the 1st score transition which concerns on 1st modification. It is a figure which shows the comparative example of the transition of the number of times of keyword appearance in 1st management information. It is a figure which shows the specific example of the transition of the number of times of keyword appearance in the 1st management information. It is a flowchart which shows the specific example of the 1st business partner company candidate selection process which concerns on 2nd modification. It is a figure which shows the display example of the transition of the number of appearances of a keyword which concerns on the 2nd modification. It is a figure which shows the comparative example of the transition of the number of times of keyword appearance in the 2nd management information. It is a figure which shows the specific example of the transition of the number of times of keyword appearance in the 2nd management information. It is a flowchart which shows the specific example of the 2nd business destination company candidate selection process which concerns on 3rd modification. It is a figure which shows the display example of the transition of the number of times of appearance of a keyword which concerns on 3rd modification.

<First Embodiment>
Hereinafter, the sales support system 100 according to the first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing a configuration of a sales support system 100 according to the first embodiment. As shown in FIG. 1, the sales support system 100 includes a sales support device 10 and a terminal device 20. The sales support device 10 of the first embodiment selects a company whose management information such as a management plan includes a specific keyword related to real estate transactions as a business partner company candidate, and displays the selected company information. The display data is output to the terminal device 20. Here, the sales support device 10 exemplifies a case where the sales support device 10 is configured by a server computer whose client is the terminal device 20.

The sales support device 10 may be configured to perform distributed processing by a plurality of devices, or may be configured by a plurality of virtual machines provided in one server device. Further, the sales support device 10 may be configured by a personal computer or a cloud server. The sales support device 10 and the terminal device 20 are configured to be able to communicate with each other via a network N such as the Internet.

The sales support device 10 is configured to be able to communicate with the external management information providing server 30 via the network N. The management information providing server 30 is a server that provides management information of a company such as a management plan and a securities report. The management information providing server 30 may be a server that provides management information on a company's homepage or the like, or may be a server of a business operator that operates a management information site or the like that provides management information of a plurality of companies. The network N may be composed of an intranet in the company that connects the sales support device 10 and the terminal device 20, and an Internet that connects the sales support device 10 and the management information providing server 30.

The terminal device 20 is an information processing device used by the user. The terminal device 20 is, for example, a mobile terminal such as a smartphone, a tablet, a PDA (Personal Digital Assistant), a desktop personal computer, a notebook personal computer, or the like. A plurality of terminal devices 20 may be connected to the network N.

FIG. 2 is a block diagram showing a specific configuration example of the sales support system 100 according to the first embodiment. The sales support device 10 shown in FIG. 2 includes a communication unit 11, a control unit 12, and a storage unit 14. The communication unit 11, the control unit 12, and the storage unit 14 are each connected to the bus line 10L, and information (data) can be exchanged with each other.

The communication unit 11 is connected to the network N by wire or wirelessly, and transmits / receives information (data) between the terminal device 20 and the management information providing server 30. The communication unit 11 functions as a communication interface for the Internet and an intranet, and can perform communication using, for example, TCP / IP.

The control unit 12 comprehensively controls the entire sales support device 10. The control unit 12 is composed of an integrated circuit such as an MPU (Micro Processing Unit). The control unit 12 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), and a ROM (Read Only Memory). The control unit 12 loads a necessary program into the ROM and executes the program using the RAM as a work area to perform various processes (such as management information analysis process described later).

The storage unit 14 stores various programs executed by the control unit 12 and data (specific keywords, weighting factors, etc.) used by these programs. The storage unit 14 is composed of a storage device such as a hard disk or an optical disk. The configuration of the storage unit 14 is not limited to these, and the storage unit 14 may be configured by a semiconductor memory such as a RAM or a flash memory. For example, the storage unit 14 may be configured by SSD (Solid State Drive).

The storage unit 14 includes a program storage unit 141, a company information database 142 (company information DB), a management information database 143 (management information DB), a keyword database 144 (keyword DB), a score database 145 (score DB), and the like.

The program storage unit 141 stores various programs executed by the control unit 12. The control unit 12 reads a necessary program from the program storage unit 141 and executes various processes such as a management information analysis process described later.

The corporate information database 142 stores corporate information such as the company name of a listed company, a securities code (stock code), and a listed market (Tokyo Stock Exchange, Nagoya Stock Exchange, Fukuoka Stock Exchange, etc.). In addition, the company information database 142 can store information useful for selling real estate transactions to each company. For example, basic corporate information such as capital, industry, business type, number of employees, settlement date, and location may be stored in the corporate information database 142.

The management information database 143 stores management information for each company, which is subjected to analysis processing for selecting a business partner company candidate for real estate transactions by the sales support device 10. Examples of management information include IR (Investor Relations) information for investors such as management plans (medium-term management plans, etc.), securities reports, quarterly reports, and financial results briefing reports. The management information is not limited to IR information, but also includes management-related information published on, for example, a company's homepage.

The keyword database 144 stores specific keywords and weighting coefficients used for analyzing management information. Specific keywords are keywords related to real estate transactions. In the keyword database 144, one or more specific keywords are preset according to the type of management information to be analyzed.

The specific keyword is preferably a keyword that can predict the possibility of real estate transactions. For example, a specific keyword has a high sales success rate by setting a keyword that may be included in the management information of a company that has real estate transaction demand (including demand such as buying and selling or renting real estate and its potential demand). You can select a company. Such specific keywords include, for example, "trend prediction keywords" for predicting trends in real estate transactions of companies, and "keyman keywords" for finding companies that have executives who can easily understand the necessity of real estate liquidation. , "Facilities status keywords" for finding companies that own real estate and real estate equipment that are likely to be in demand for real estate transactions.

By analyzing how many of these specific keywords are included in the management information, it is possible to efficiently select business partners with a high possibility of successful sales. Companies that include these specific keywords in their management information are likely to dig up demand for real estate transactions, so it is highly valuable to select them as potential business partners. For example, "trend forecast keywords" include "asset efficiency improvement" and "interest-bearing debt reduction". Selling and organizing real estate contributes to asset efficiency and debt reduction, so if management information includes specific keywords such as "asset efficiency" and "interest-bearing debt reduction," dig up demand for real estate sales. Cheap.

By the way, for example, many specific keywords can be considered as "trend prediction keywords" for predicting the trend of real estate transactions of a company, and there is a high possibility that the sales success rate will change depending on the type of specific keyword (word type). Therefore, the first embodiment includes a plurality of keyword groups in which a plurality of specific keywords representing various characteristics of corporate management related to real estate transactions are grouped according to the characteristics. A plurality of specific keywords representing common characteristics are grouped in the same keyword group, and a plurality of specific keywords representing different characteristics are grouped in different keyword groups.

Examples of such a keyword group include "efficiency-related keywords" related to asset efficiency and the like, and "finance-related keywords" related to financial improvement. The "efficiency-related keywords" include specific keywords such as the above-mentioned "asset efficiency". “Financial keywords” include, for example, the above-mentioned “interest-bearing debt reduction”. In this way, the characteristics of each keyword group are different, and the degree of interest in real estate transactions of a company also differs depending on how a specific keyword is used in the management information of the company. Therefore, there is a high possibility that the sales success rate will change depending on which keyword group specific keywords are included in the management information.

Based on such knowledge, in the first embodiment, a weighting coefficient is set for each of these keyword groups. As a result, different weighting factors can be given to specific keywords that represent different characteristics of corporate management related to real estate transactions. In addition, the same weighting coefficient can be given to specific keywords having common characteristics. In this way, by setting the weighting coefficient for each keyword group, it is possible to associate the weighting coefficient according to the characteristics of corporate management related to real estate transactions. This makes it possible to analyze management information that reflects the high level of interest in real estate transactions from the perspective of corporate management related to real estate transactions, making it easier to select a leading company as a business partner.

The specific keyword and the weighting coefficient are associated with each keyword group and stored in the keyword database 144. Although the case where the specific keywords included in the trend prediction keyword are grouped is illustrated, the specific keywords included in the above-mentioned keyman keyword and equipment status keyword may also be grouped and associated with the weighting coefficient.

Here, a configuration example of the keyword database 144 will be described with reference to FIG. An example is given when the specific keywords included in the above-mentioned "trend prediction keywords" are divided into n keyword groups (KW groups). The configuration of the keyword database 144 is not limited to that shown in FIG.

The keyword database 144 (keyword DB) of FIG. 3 includes n keyword groups (KW groups) GA1, GA2, ..., GAn, and n weight coefficients WA1, WA2, ... Associated with each keyword group. , WAN and a specific keyword (specific KW) included in each keyword group are stored.

The specific keywords KA11, KA12, ... Are included in the keyword group GA1, and the specific keywords KA21, KA22, ... Are included in the keyword group GA2. The notation such as "KA11" is a conceptual expression of a character string of a specific keyword. The keyword group GA1 is, for example, an "efficiency-type keyword" used when improving the efficiency of assets or the like in management information, and the specific keyword KA11 included in this GA1 is, for example, the above-mentioned "asset efficiency improvement". The keyword group GA2 is, for example, a “financial keyword” used when it is desired to promote financial improvement, and the specific keyword KA21 included in this GA2 is, for example, the above-mentioned “interest-bearing debt reduction”.

The weighting factor of the keyword group GA1 is WA1, and the weighting factor of the keyword group GA2 is WA2. For example, when the specific keyword of the keyword group GA2 appears in the management information, the weight coefficient WA2 is higher than the weight coefficient WA1 when the possibility of business success is higher than the case where the specific keyword of the keyword group GA1 appears. Is also set to be large.

The score database 145 stores the number of appearances and scores (index values) of specific keywords associated with each management information for each company. The score is calculated for each management information, and is used as approach information, for example, to select a candidate for a business partner and determine the display order. The score is calculated based on the number of occurrences of a specific keyword in each keyword group and the weighting coefficient. Specifically, the score is calculated from the word hit information and the weighting coefficient obtained from the number of occurrences of the specific keyword for each keyword group. The word hit information is information for each keyword group, which is an index of how much each specific keyword included in the keyword group is included in the management information (whether it is a hit). According to the word hit information, it is possible to know how many specific keywords of which keyword group appear in the management information.

The word hit information is, for example, the sum of the number of occurrences of all specific keywords included in the keyword group. However, the present invention is not limited to this, and if even one of the specific keywords included in the keyword group is included in the management information, the word hit information may be set to 1 (hit). Further, the word hit information may be a numerical value obtained by normalizing or standardizing the number of occurrences of each specific keyword included in the keyword group.

By calculating the score from the word hit information and the weighting coefficient of each keyword group in this way, it is possible to change the score depending on how many specific keywords of which keyword group appear in the management information. In this respect, the score can also be understood as the word coverage rate of a specific keyword. By listing the business partner company candidates in descending order of score, the company list can be displayed in descending order of probability of successful sales.

Here, a configuration example of the score database 145 will be described with reference to FIG. Take as an example the case where the management plan of company A is acquired as management information and analyzed by the specific keyword shown in FIG. The configuration of the score database 145 is not limited to that shown in FIG.

The score database 145 (score DB) of FIG. 4 stores the number of appearances HA11, HA12, ... And the score Sa of the specific keyword (specific KW) that analyzes the business plan of company A. As shown in FIG. 4, the score database 145 stores basic information of the management plan such as the company ID, the type of management information, and the company name. As shown in FIG. 4, the score database 145 may store information such as a keyword group, a weighting coefficient, and a specific keyword used for analyzing management information.

The number of appearances HA11 in FIG. 4 is the number of appearances of the specific keyword KA11 of the keyword group (KW group) GA1, and the number of appearances HA21 is the number of appearances of the specific keyword KA21 of the keyword group GA2. The word hit information of each keyword group in FIG. 4 is, for example, the total number of occurrences of each specific keyword included in the keyword group. For example, HGA1 is the sum of all the appearance times HA11, HA12, ... Of the specific keyword in the keyword group GA1. When there are n keyword groups in FIG. 4, the score Sa from the word hit information HGA1 to HGAn and the weighting coefficients WA1 to WAN can be expressed by the following mathematical formula (1). However, the formula for calculating the score Sa is not limited to the following formula (1).

Sa = HGA1 x WA1 + HGA2 x WA2 + ... + HGAn x WAN
... (1)

The weighting factor of each keyword group can be set freely. For example, if a skilled salesperson sets a weighting coefficient in advance based on his / her own experience, the score can be made higher as the management information of a leading business partner company with a high possibility of real estate transaction demand becomes higher. As described above, in the first embodiment, specific keywords are grouped according to common characteristics and the weighting coefficient is set in advance, so that the characteristics and importance of the idea of corporate management can be reflected in the score of management information. can. As a result, it is possible to obtain a score according to the possibility of real estate transaction demand.

The control unit 12 shown in FIG. 2 includes an information acquisition unit 121, a keyword determination unit 122, a company selection unit 124, and an output unit 126. Each component of the control unit 12 may be composed of a physical circuit or a program that can be executed by the CPU. The configuration of the control unit 12 is not limited to the configuration shown in FIG.

The information acquisition unit 121 acquires management information of a plurality of companies from the management information providing server 30 via the communication unit 11. For example, the information acquisition unit 121 scrapes the management information providing site operated by the management information providing server 30, and acquires the document data of the management information of a plurality of companies. The information acquisition unit 121 can automatically acquire document data of management information of various companies by periodically crawling a specific management information providing site.

If the acquired management information is document data (PDF, XML, XBRL, etc.) that can be searched for character strings (text search), the information acquisition unit 121 stores the acquired management information in the management information database 143 as it is, and the document data (documents) that cannot be searched for character strings. In the case of (PDF captured as an image, etc.), it is converted into data that can be searched for character strings and stored in the management information database 143. XBRL (eXtensible Business Reporting Language) is a report description language based on XML (XML (eXtensible Markup Language)).

The document data of the management information obtained by another route can be added to the management information database 143 by the information acquisition unit 121. Specifically, the information acquisition unit 121 receives the document data of the management information in response to the user operation from the terminal device 20 and adds it to the management information database 143. Thereby, for example, even when the document data of the management information of the unlisted company or the slightly private management information is obtained, it can be added to the management information database 143.

The keyword determination unit 122 performs a specific keyword determination for determining whether or not the management information acquired by the information acquisition unit 121 includes a specific keyword in the keyword database 144, and acquires a score as an analysis result of the management information. The keyword determination unit 122 reads the document data of the management information from the management information database 143 and performs a specific keyword determination. The keyword determination unit 122 performs a specific keyword determination for each management information acquired by the information acquisition unit 121, counts the number of appearances of the specific keyword in each keyword group, and stores the counted number of appearances in the score database 145.

The keyword determination unit 122 includes a score acquisition unit 123. The score acquisition unit 123 acquires a score as an analysis result of management information by a specific keyword determination. Specifically, the score acquisition unit 123 calculates and acquires the score of the management information from the number of appearances of the specific keyword in each keyword group and the weighting coefficient. The keyword determination unit 122 stores the score acquired by the score acquisition unit 123 in the score database 145 in association with the management information.

The company selection unit 124 selects a company determined by the keyword determination unit 122 that the specific keyword is included in the management information as a business partner company candidate. The above-mentioned score of management information is used to select a candidate for a business partner. For example, the company selection unit 124 selects a company whose management information score stored in the score database 145 is equal to or higher than a predetermined threshold value as a business partner company candidate. According to this, since the number of business partner company candidates can be adjusted by adjusting the predetermined threshold value, for example, it is possible to prevent too many business partner company candidates having a low score.

For management information that does not include any specific keyword, the number of occurrences of the specific keyword is zero, so the score is also zero. The score of management information that includes even one specific keyword is 1 or more. Therefore, when there are few business partner company candidates, for example, a company whose management information score is a threshold value of 1 or more may be selected as the business partner company candidate. As a result, a company that includes at least one specific keyword can be selected as a sales position company candidate, so that the number of sales position company candidates can be increased.

The output unit 126 generates and outputs display data for displaying the company list of the business partner company candidates selected by the company selection unit 124. The output unit 126 generates display data for displaying the company list in descending order of the score of the management information. The display data is transmitted to the terminal device 20 via the communication unit 11. The terminal device 20 displays a list of business partner company candidates based on the display data. The display data may be Web screen data. Specifically, the sales support device 10 displays a list of business partner company candidates on a Web screen. The terminal device 20 receives the Web screen data and displays it on the browser.

Next, a configuration example of the terminal device 20 will be described with reference to FIG. The terminal device 20 shown in FIG. 2 includes a communication unit 21, a control unit 22, a storage unit 24, an input unit 25, and a display unit 26. The communication unit 21, the control unit 22, the storage unit 24, the input unit 25, and the display unit 26 are each connected to the bus line 20L, and information (data) can be exchanged with each other.

The communication unit 21 is connected to the network N by wire or wirelessly, and transmits / receives information (data) to / from the sales support device 10. The communication unit 21 functions as a communication interface for the Internet and an intranet, and can perform communication using, for example, TCP / IP.

The control unit 22 comprehensively controls the entire terminal device 20. The control unit 22 is composed of an integrated circuit such as an MPU (Micro Processing Unit). The control unit 22 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), and a ROM (Read Only Memory). The control unit 22 performs various processes by loading a necessary program into the ROM and executing the program using the RAM as a work area.

The storage unit 24 stores various programs executed by the control unit 22 and data used by these programs. The storage unit 24 is composed of a storage device such as a hard disk or an optical disk. The configuration of the storage unit 24 is not limited to these, and the storage unit 24 may be configured by a semiconductor memory such as a RAM or a flash memory. For example, the storage unit 24 may be configured by SSD (Solid State Drive).

The input unit 25 includes a keyboard, a mouse, and the like, receives an operation input from the user, and transmits a control signal corresponding to the operation content to the control unit 22. The input unit 25 may include a touch panel.

The display unit 26 is a liquid crystal display, an organic EL display, or the like, and displays various information according to instructions from the control unit 22. The control unit 22 displays a company list of business partner company candidates on the display unit 26 based on the display data received from the sales support device 10 via the communication unit 21.

Here, a specific example of the display screen SC displayed on the display unit 26 is shown in FIG. On the display screen SC, the business partner company candidates selected by the sales support device 10 are displayed in a list as approach information. Candidates for business partners are displayed at the top as the score is higher. As a result, companies with a high probability of successful sales can be displayed at the top. By scrolling the approach information on the display screen SC of FIG. 5, lower-level companies can also be displayed.

The display items of the approach information are "date", "company name", "stock code", "sales information", "score", "remarks" and the like. The display items of approach information are not limited to those shown in Fig. 5, for example, "listed market", "capital", "industry", "business format", "number of employees", "closing date", "location", etc. You may add items. A search period item may be provided so that the period can be entered in the item, and the display can be narrowed down to the companies selected by the analysis of the management information within the entered period. In addition, an item of "evaluation" may be provided. In the "evaluation" item, it is possible to enter a numerical value that evaluates whether or not it was useful for sales in multiple stages such as 3 stages and 5 stages. A link may be pasted to the characters displayed in the item of "management information" so that the document data of the management information of the management information database 143 is displayed. When the characters of the business plan are clicked with the mouse of the input unit 25 or the like, the control unit 22 causes the display unit 26 to display the document data of the business plan of the business information database 143.

In the "remarks", a character string (for example, "important WORD HIT" shown in FIG. 5) indicating a specific keyword of the keyword group having a weighting coefficient of a predetermined value or more is displayed. In addition, the "remarks" is displayed. Is a character string indicating the characteristics of a keyword group that includes a specific keyword (for example, "efficiency WORD HIT" if an efficiency keyword is included, "finance WORD HIT" if a finance keyword is included, etc.). It may be displayed. Displaying these makes it easier to formulate a sales strategy for the company.

Next, the management information analysis process performed by the sales support device 10 according to the first embodiment will be described with reference to FIG. FIG. 6 is a flowchart showing a specific example of the management information analysis process. The management information analysis process is executed by reading a necessary program from the program storage unit 141 by the control unit 12 (information acquisition unit 121, keyword determination unit 122, score acquisition unit 123, company selection unit 124, output unit 126, etc.). Will be done.

First, in step S110 shown in FIG. 6, the control unit 12 acquires the management information of each company. Specifically, the information acquisition unit 121 acquires the document data of the management information of each company from the management information providing server 30 via the communication unit 11 and stores it in the management information database 143. If the acquired management information is document data (PDF including text, XML, XBRL, etc.) that can be searched for a character string, the information acquisition unit 121 stores the acquired management information in the management information database 143 as it is, and the document data that cannot be searched for a character string. In the case of (PDF in which the document is captured as an image), it is converted into data that can be searched for a character string and stored in the management information database 143.

Next, in step S120, the control unit 12 determines a specific keyword for each of the document data of the management information acquired by the information acquisition unit 121. In the specific keyword determination, it is determined whether or not the specific keyword of each keyword group is included in the document data of the management information, and the number of appearances of the specific keyword and the score of the management information are acquired. The specific keyword determination is performed as shown in FIG. 7, for example.

FIG. 7 is a diagram showing a specific example of the specific keyword determination process. In step S121 shown in FIG. 7, the control unit 12 acquires the number of occurrences of each specific keyword (specific KW) for each keyword group. Specifically, the keyword determination unit 122 reads out the document data of the management information of the management information database 143, and determines whether or not the document data includes a specific keyword associated with each keyword group in the keyword database 144. , Count the number of occurrences of a specific keyword. The keyword determination unit 122 stores the counted number of appearances in the score database 145 in association with the management information.

Next, in step S122, the control unit 12 calculates the score of the management information based on the number of appearances of the specific keyword associated with the management information and the weighting coefficient. Specifically, the score acquisition unit 123 calculates a score from word hit information and a weighting coefficient obtained from the number of occurrences of a specific keyword for each keyword group, and stores the score in the score database 145 in association with the management information. The score in step S122 is calculated for the document data of all the management information acquired by the information acquisition unit 121. This score is the result of analysis of management information by specific keyword judgment.

Next, in step S130 shown in FIG. 6, the control unit 12 selects a business partner company candidate from the analysis result of the management information by the specific keyword determination. Specifically, the company selection unit 124 selects a company whose management information score stored in the score database 145 is equal to or higher than a predetermined threshold value as a business partner company candidate. The process of step S130 is performed on the scores of the latest management information for all the companies stored in the score database 145. As a result, candidates for business partners can always be selected from the latest management information.

Next, in step S140, the control unit 12 causes the terminal device 20 to display the company selected in step S130. Specifically, the output unit 126 generates display data and transmits it to the terminal device 20, so that the company list selected by the company selection unit 124 is displayed on the terminal device 20 as approach information. For example, the output unit 126 generates display data so that the selected companies are displayed at the top in descending order of score. The process of step S140 may be executed by the output unit 126 by receiving a display request for approach information from the terminal device 20. On the other hand, the processes of steps S110 to S130 may be automatically executed by periodic crawling of a specific management information providing site, or may be executed by a request from the terminal device 20. ..

According to the sales support device 10 according to the first embodiment as described above, an analysis of acquiring management information of a plurality of companies and determining whether or not a specific keyword related to a real estate transaction is included in the management information. By doing this, it is possible to predict companies that may have real estate transaction demand. Furthermore, since a company determined to include a specific keyword in the management information is selected as a business partner company candidate, a company with a possibility of real estate transaction demand can be automatically selected as a business partner company candidate. As a result, candidates for sales positions can be selected without relying on the experience and intuition of skilled sales staff. Therefore, the time and effort required to select a candidate for a business partner can be significantly reduced, and sales activities can be made more efficient.

<Second Embodiment>
A second embodiment of the present invention will be described. The components having substantially the same functional configuration in each of the embodiments illustrated below are designated by the same reference numerals, and duplicate description will be omitted. In the second embodiment, a case where the type of the specific keyword is changed according to the type of the management information is illustrated in the specific keyword determination as to whether or not the management information disclosed by the company includes the specific keyword related to the real estate transaction. The types of management information and the types of specific keywords are not limited to those illustrated below.

The types of specific keywords here are the "trend prediction keyword", "keyman keyword", "equipment status keyword", etc. exemplified in the first embodiment. "Trend forecast keyword" is a specific keyword for predicting the trend of a company's real estate transaction. The "keyman keyword" is a specific keyword for finding a company that has executives who can easily understand the necessity of real estate liquidation. The "equipment status keyword" is a specific keyword for finding a real estate or a company that owns real estate equipment that is likely to have demand for real estate transactions.

By the way, various information is included in the management information such as the management plan published by the company. However, since the content included differs depending on the type of management information, the specific keywords that are likely to appear also differ depending on the type of management information.

For example, in the medium-term management plan, "trend forecast keywords" that can predict trends in real estate transactions such as asset efficiency improvement and interest-bearing debt reduction are likely to appear. By analyzing the medium-term management plan with such specific keywords, the analysis results can be effectively linked to the selection of business partner company candidates.

However, the medium-term management plan often does not include detailed information on officers. Therefore, it is unlikely that the "Keyman keyword" will appear, and even if it does appear, it is highly likely that it happened to hit an item that has nothing to do with Keyman. Analysis results using such specific keywords cannot be effectively linked to the selection of business partner candidates with key people.

On the other hand, since the securities report contains detailed information on officers and equipment status, the "keyman keywords" and "equipment status keywords" that appear in the securities report can be used to search for keymen and real estate equipment. Probability is high. Therefore, the analysis results using such specific keywords can be effectively linked to the selection of business partner candidates with key persons and real estate facilities.

Therefore, in the second embodiment, the specific keyword determination is performed by changing the type of the specific keyword according to the type of management information. Specifically, for example, the keyword determination unit 122 of the second embodiment uses the first management information (medium-term management plan, quarterly report, financial results briefing report, etc.) including the management plan information as the first specific keyword (trend forecast). Judging by (keyword), the second management information (securities report, etc.), which is a type of management information different from the first management information and includes officer information and equipment information, is a second type different from the first specific keyword. Judgment is made by a specific keyword (keyman keyword or equipment status keyword).

According to this, it is possible to change the approach of selecting business partner company candidates according to the type of management information. For example, in the case of a medium-term management plan that includes management plan information, "trend forecast keywords" are used, so it is possible to select business partner candidates based on the approach of whether or not the company has real estate transaction demand (here, referred to as the first approach). ). On the other hand, in the case of securities reports that include officer information and equipment information, "keyman keywords" and "equipment status keywords" are used instead of "trend forecast keywords." In this case, candidates for business partners are selected by different approaches, including whether the company has a key person who can easily understand the necessity of real estate liquidation, and whether the company owns equipment that is in demand for real estate transactions. Can be selected (referred to here as the second approach). By changing the type of specific keyword according to the type of management information in this way, compared to the case where the same specific keyword is used for different types of management information, the analysis result by the specific keyword judgment of the management information is a powerful business partner. The effect that leads to the selection of company candidates can be greatly enhanced. Therefore, it is possible to improve the accuracy of selecting business partner company candidates.

FIG. 8 is a block diagram showing a specific configuration example of the sales support system 100 according to the second embodiment. The management information database 143 (management information DB) of FIG. 8 stores the first management information 143A (for example, a medium-term management plan) and the second management information 143B (for example, a securities report). The keyword database 144 (keyword DB) of FIG. 8 stores the trend prediction keyword 144A, the keyman keyword 144B, and the equipment status keyword 144C. The trend prediction keyword 144A is an example of the first specific keyword for determining the first management information 143A, and the keyman keyword 144B and the equipment status keyword 144C are the second specific keywords for determining the second management information 143B, respectively. It is an example.

Examples of the trend forecast keyword 144A include “asset efficiency improvement” and “interest-bearing debt reduction”. Since selling and organizing real estate contributes to asset efficiency, it is easy to find demand for real estate sales when management information includes a specific keyword such as "asset efficiency". Since selling and organizing real estate is directly linked to finance, it is easy to find demand for real estate sales when management information includes specific keywords such as "interest-bearing debt reduction." Therefore, a company in which these trend prediction keywords 144A are included in the management information is promising as a business partner company candidate.

The keyman keyword 144B is, for example, "bank". You can search for a company that has an officer from a bank depending on whether or not "bank" is included in the officer column of the securities report. Companies with executives from banks are promising candidates for business because it is easy for them to understand the need for real estate liquidation. The "equipment status keyword" is, for example, "rental". You can search for a company that owns real estate equipment that you are renting, depending on whether or not "rental" is included in the equipment column of the securities report. A company that owns real estate equipment for rent is a promising candidate for a business partner because there is a possibility of demand for reorganization of the real estate equipment.

Here, a configuration example of the keyword database 144 (keyword DB) of the second embodiment will be described with reference to FIGS. 9 to 11. FIG. 9 is a diagram showing a specific example of the trend prediction keyword 144A, FIG. 10 is a diagram showing a specific example of the keyman keyword 144B, and FIG. 11 is a diagram showing a specific example of the equipment status keyword 144C. Here, an example will be given in which each specific keyword of the trend prediction keyword 144A, the keyman keyword 144B, and the equipment status keyword 144C is divided into a plurality of keyword groups (KW groups). The configurations of the trend prediction keyword 144A, the keyman keyword 144B, and the equipment status keyword 144C are not limited to those shown in the figure. For example, the keyman keyword 144B and the equipment status keyword 144C do not necessarily have to be divided into keyword groups.

The trend prediction keywords 144A in FIG. 9 include n keyword groups (KW groups) GA1, GA2, ..., GAn, and n weight coefficients WA1, WA2, ..., WAN associated with each keyword group. , The specific keyword (specific KW) included in each keyword group is stored.

The specific keywords KA11, KA12, ... Are included in the keyword group GA1, and the specific keywords KA21, KA22, ... Are included in the keyword group GA2. The keyword group GA1 is, for example, an "efficiency-type keyword" used when improving the efficiency of assets or the like in management information, and the specific keyword KA11 included in this GA1 is, for example, the above-mentioned "asset efficiency improvement". The keyword group GA2 is, for example, a “financial keyword” used when it is desired to improve finance, and the specific keyword KA21 included in this GA2 is, for example, the above-mentioned “interest-bearing debt reduction”.

The keyman keywords 144B in FIG. 10 include n keyword groups (KW groups) GB1, GB2, ..., GBn, and n weighting coefficients WB1, WB2, ..., WBn associated with each keyword group. A specific keyword (specific KW) included in each keyword group is stored. The specific keywords KB11, KB12, ... Are included in the keyword group GB1, and the specific keywords KB21, KB22, ... Are included in the keyword group GB2. The keyword group GB1 is a "bank-born keyword" used to search for a company having an officer from a bank, and the specific keyword KB11 included in this GB1 is, for example, "bank".

The equipment status keywords 144C in FIG. 11 include n keyword groups (KW groups) GC1, GC2, ..., GCn and n weighting coefficients WC1, WC2, ..., WCn associated with each keyword group. , The specific keyword (specific KW) included in each keyword group is stored. The specific keywords KC11, KC12, ... Are included in the keyword group GC1, and the specific keywords KC21, KC22, ... Are included in the keyword group GC2. The keyword group GC1 is a “rental real estate equipment keyword” used to search for a company that owns rental real estate equipment, and the specific keyword KC11 included in this GC1 is, for example, “rental”.

The score database 145 (score DB) of FIG. 8 stores the first score information 145A and the second score information 145B. The first score information 145A includes information on the number of occurrences of a specific keyword associated with each first management information 143A for each company and information on a first score (index value) as an analysis result of the first management information. The first score is calculated for each first management information. The first score is used, for example, as the first approach information for selecting a candidate for a business partner and determining the display order. The first score is calculated based on the number of appearances of a specific keyword in each keyword group of the trend prediction keyword 144A and the weighting coefficient.

The second score information 145B includes information on the number of appearances of the specific keyword associated with each second management information 143B for each company and the information of the second score (index value) which is the analysis result of the second management information. The second score is calculated for each second management information. The second score is used, for example, as the second approach information for selecting a candidate for a business partner and determining the display order. The second score is calculated based on the number of occurrences of the specific keyword and the weighting coefficient in each keyword group of the keyman keyword 144B and the equipment status keyword 144C.

Here, specific examples of the first score information 145A and the second score information 145B will be described with reference to FIGS. 12 and 13. FIG. 12 is a diagram showing a specific example of the first score information 145A, and FIG. 13 is a diagram showing a specific example of the second score information 145B.

The first score information 145A in FIG. 12 is the number of appearances of the specific keyword (specific KW) of the trend prediction keyword 144A for analyzing the medium-term management plan of company A (example of the first management information 143A) HA11, HA12, ... And the information of the first score Sa1. As shown in FIG. 12, the first score information 145A includes basic information such as a company ID, a type of management information, and a company name. As shown in FIG. 12, the first score information 145A may include information such as a keyword type, a keyword group, a weighting coefficient, and a specific keyword used in the analysis.

The number of appearances HA11 in FIG. 12 is the number of appearances of the specific keyword KA11 in the keyword group (KW group) GA1, and the number of appearances HA21 is the number of appearances of the specific keyword KA21 in the keyword group GA2. The word hit information of each keyword group in FIG. 12 is the total number of occurrences of each specific keyword included in the keyword group. For example, HGA1 is the sum of all the appearance times HA11, HA12, ... Of the specific keyword in the keyword group GA1. When there are n keyword groups in FIG. 12, the first score Sa1 from the word hit information HGA1 to HGAn and the weighting coefficients WA1 to WAN can be expressed by the following mathematical formula (2). However, the formula for calculating the first score Sa1 is not limited to the following formula (2).

Sa1 = HGA1 x WA1 + HGA2 x WA2 + ... + HGAn x WAN
... (2)

The second score information 145B in FIG. 13 includes the number of appearances of the specific keyword (specific KW) of the key man keyword 144B for analyzing the securities report of company A (example of the second management information 143B) HB11, HB12, ... , The number of appearances of the specific keyword (specific KW) of the equipment status keyword 144C HC11, HC12, ..., And the second score Sa2 are provided. As shown in FIG. 13, the second score information 145B includes basic information such as a company ID, a type of management information, and a company name. As shown in FIG. 13, the second score information 145B may include information such as a keyword type, a keyword group, a weighting coefficient, and a specific keyword used in the analysis.

Further, the second score information 145B includes information on the presence / absence of a "key man" or "equipment". The column of "keyman" in FIG. 13 is, for example, "HIT" when there is a keyman, and "NONE" when there is no keyman. If there is a key man, it may be set to "1", and if there is no key man, it may be set to "0". The "equipment" column in FIG. 13 is, for example, "HIT" when there is real estate equipment, and "NONE" when there is no real estate equipment. If there is real estate equipment, it may be set to "1", and if there is no real estate equipment, it may be set to "0".

The number of appearances HB11 in FIG. 13 is the number of appearances of the specific keyword KB11 of the keyword group (KW group) GB1, and the number of appearances HB21 is the number of appearances of the specific keyword KB21 of the keyword group GB2. The number of appearances HC11 is the number of appearances of the specific keyword KC11 of the keyword group (KW group) GC1, and the number of appearances HC21 is the number of appearances of the specific keyword KC21 of the keyword group GC2.

The word hit information of each keyword group in FIG. 13 is the total number of occurrences of each specific keyword included in the keyword group. For example, HGB1 is the sum of all the appearance times HB11, HB12, ... Of the specific keyword of the keyword group GB1. When there are n keyword groups of the keyman keyword 144B and the equipment status keyword 144C in FIG. 13, the word hit information HGB1 to HGBn, HGC1 to HGCn and the weighting coefficients WB1 to WBn, WC1 to WCn to the second score Sa2 are the following mathematical formulas. It can be represented by (3). However, the formula for calculating the second score Sa2 is not limited to the following formula (3).

Sa2 = HGB1 x WB1 + HGB2 x WB2 + ... + HGBn x WBn
+ HGC1 x WC1 + HGC2 x WC2 + ... + HGCn x WCn
... (3)

The weighting coefficient of the keyword group in FIGS. 12 and 13 can be freely set. By setting the weighting factor based on the experience of a skilled salesperson, the first score and the second score should be higher for the management information of a leading business partner company with a high possibility of real estate transaction demand. Can be done. By grouping specific keywords in this way, the weighting coefficient can be changed according to the possibility of real estate transaction demand.

Next, the management information analysis process performed by the sales support device 10 according to the second embodiment will be described with reference to FIGS. 14 to 17. FIG. 14 is a flowchart showing a specific example of the first management information analysis process, and FIG. 15 is a flowchart showing a specific example of the first specific keyword determination in the first management information analysis process. FIG. 16 is a flowchart showing a specific example of the second management information analysis process, and FIG. 17 is a flowchart showing a specific example of the second specific keyword determination in the second management information analysis process.

The approach method for selecting a business partner company candidate is different between the first management information analysis process of FIG. 14 and the second management information analysis process of FIG. In the first management information analysis process, the medium-term management plan of each company (example of the first management information 143A) is judged by the trend prediction keyword 144A (example of the first specific keyword), and the business partner company candidate is selected from the analysis result. do. That is, in the first management information analysis process, a medium-term management plan that tends to include many specific keywords that can predict the trend of real estate transactions is targeted, and a promising business partner company candidate is selected according to how much the trend prediction keyword 144A is included. Yes (first approach).

On the other hand, in the second management information analysis process, the securities report of each company (example of the second management information 143B) is determined by the keyman keyword 144B and the equipment status keyword 144C (example of the second specific keyword), and the analysis result is obtained. Select a business partner candidate from. In other words, it is possible to select business partner candidates based on human resources and real estate equipment that can be key players in real estate transactions, targeting securities reports that tend to include specific keywords related to executive information and equipment information (second approach).

The first management information analysis process and the second management information analysis process are performed by the control unit 12 (information acquisition unit 121, keyword determination unit 122, score acquisition unit 123, company selection unit 124, output unit 126, etc.) from the program storage unit 141. The required program is read and executed.

First, the first management information analysis process shown in FIG. 14 will be described in detail. In step S210, the control unit 12 acquires the first management information 143A of each company. Specifically, the information acquisition unit 121 acquires the medium-term management plan of each company from the management information providing server 30 via the communication unit 11 as the first management information 143A and stores it in the management information database 143. If the acquired medium-term management plan is document data (text search) that can be searched by character strings (PDF including text), the information acquisition unit 121 stores the acquired document data as it is in the management information database 143, and document data (documents) that cannot be searched by character strings. In the case of (PDF, etc. captured as an image), it is converted into data that can be searched for character strings and stored in the management information database 143.

Next, in step S220, the control unit 12 performs the first specific keyword determination for each of the document data of the medium-term management plan acquired by the information acquisition unit 121. In the first specific keyword determination, the number of appearances of the first specific keyword included in the medium-term management plan is acquired, and the first score is acquired from the number of appearances and the weighting coefficient. The first specific keyword here is a specific keyword included in each keyword group of the trend prediction keyword 144A.

The first specific keyword determination is executed, for example, by the procedure shown in FIG. In step S221, the control unit 12 acquires the number of occurrences of each first specific keyword (first specific KW) for each keyword group. Specifically, the keyword determination unit 122 reads the document data of the medium-term management plan of the management information database 143 and determines whether or not the document data includes the first specific keyword associated with each keyword group. .. The keyword determination unit 122 counts the number of appearances of the first specific keyword, associates the counted number of appearances with the management information, and stores it in the score database 145 as the first score information 145A.

Next, in step S222, the control unit 12 calculates the first score of the medium-term management plan based on the number of appearances and the weighting coefficient of the first specific keyword associated with the medium-term management plan. Specifically, the score acquisition unit 123 calculates the first score from the word hit information and the weighting coefficient obtained from the number of occurrences of the first specific keyword for each keyword group, and associates it with the medium-term management plan in the score database 145. It is stored as the first score information 145A. The first score of step S222 is acquired for all the document data of the medium-term management plan acquired by the information acquisition unit 121. This first score is the analysis result of the first management information 143A by the first specific keyword determination.

Next, in step S230 shown in FIG. 14, the control unit 12 selects a business partner company candidate from the analysis result of the first management information 143A by the first specific keyword determination. Specifically, the company selection unit 124 selects a company whose first score of the medium-term management plan stored in the score database 145 is equal to or higher than a predetermined threshold value (first threshold value) as a business partner company candidate. The process of step S230 is performed for all the companies stored in the score database 145 with respect to the first score of the latest medium-term management plan. As a result, candidates for business partners can always be selected from the latest medium-term management plan.

Next, in step S240, the control unit 12 causes the terminal device 20 to display the company selected in step S230. Specifically, the output unit 126 generates display data and transmits it to the terminal device 20, so that the company list selected by the company selection unit 124 is displayed on the terminal device 20 as the first approach information. For example, the output unit 126 generates display data so that the selected companies are displayed at the top in descending order of the first score. The process of step S240 may be executed by the output unit 126 by receiving a display request for approach information from the terminal device 20. On the other hand, the processes of steps S210 to S230 may be automatically executed by periodic crawling of a specific management information providing site, or may be executed by a request from the terminal device 20. ..

Next, the second management information analysis process shown in FIG. 16 will be described in detail. In step S310, the control unit 12 acquires the second management information 143B of each company. Specifically, the information acquisition unit 121 acquires the securities report of each company from the management information providing server 30 via the communication unit 11 as the second management information 143B and stores it in the management information database 143. If the acquired securities report is document data (PDF including text, XML, XBRL, etc.) that can be searched for a character string, the information acquisition unit 121 stores it in the management information database 143 as it is and cannot search the character string. In the case of document data (PDF in which the document is captured as an image), it is converted into data that can be searched for a character string and stored in the management information database 143.

Next, in step S320, the control unit 12 determines the second specific keyword for each of the document data of the securities report acquired by the information acquisition unit 121. In the second specific keyword determination, the number of appearances of the second specific keyword included in the securities report is acquired, and the second score is acquired from the number of appearances and the weighting coefficient. The second specific keyword here is a specific keyword included in each keyword group of the keyman keyword 144B and the equipment status keyword 144C.

The second specific keyword determination is executed, for example, by the procedure shown in FIG. In step S321, the control unit 12 determines whether or not there is a keyman including the specific keyword of the keyman keyword 144B. Specifically, the keyword determination unit 122 refers to the second score information 145B in step S321, and when a specific keyword of the keyman keyword 144B appears, it is assumed that there is a keyman in step S322. Specifically, as shown in FIG. 13, the keyman column of the second score information 145B is set to "HIT" or "1". If the specific keyword of the keyman keyword 144B does not appear in step S321, the keyman column of the second score information 145B is set to "NONE" or "0" as no keyman.

Next, in step S323, the control unit 12 determines whether or not there is equipment including the specific keyword of the equipment status keyword 144C. Specifically, the keyword determination unit 122 refers to the second score information 145B in step S323, and when a specific keyword of the equipment status keyword 144C appears, it is assumed that the equipment is installed in step S324. Specifically, as shown in FIG. 13, the equipment column of the second score information 145B is set to "HIT" or "1". If the specific keyword of the equipment status keyword 144C does not appear in step S324, the equipment column of the second score information 145B is set to "NONE" or "0" as no equipment.

Next, in step S325, the control unit 12 acquires the number of occurrences of each second specific keyword (second specific KW) for each keyword group. Specifically, the keyword determination unit 122 reads the document data of the securities report of the management information database 143 and determines whether or not the document data includes the second specific keyword associated with each keyword group. .. The keyword determination unit 122 counts the number of appearances of the second specific keyword, associates the counted number of appearances with the management information, and stores it in the score database 145 as the second score information 145B.

Next, in step S326, the control unit 12 calculates the second score of the securities report based on the number of appearances and the weighting coefficient of the second specific keyword associated with the securities report. Specifically, the score acquisition unit 123 calculates the second score from the word hit information and the weighting coefficient obtained from the number of occurrences of the second specific keyword for each keyword group, and associates it with the securities report in the score database 145. It is stored as the second score information 145B. The acquisition of the second score in step S326 is performed on the document data of all the securities reports acquired by the information acquisition unit 121. This second score is the analysis result of the second management information 143B by the second specific keyword determination.

When performing the second specific keyword determination, the area of the second management information 143B may be specified and determined. For example, the area of the column related to officers in the securities report is specified to determine whether or not the keyman keyword 144B is included, and the area of the column related to equipment in the securities report is specified to determine whether the equipment status keyword 144C is included. Judge whether or not. As a result, it is possible to prevent a specific keyword from being hit in a column separate from the officers and equipment, which makes it easier to find a company with officers and equipment.

Next, in step S330 shown in FIG. 16, the control unit 12 selects a business partner company candidate from the analysis result of the second management information 143B by the second specific keyword determination. Specifically, the company selection unit 124 selects a company whose second score of the securities report stored in the score database 145 is equal to or higher than a predetermined threshold value (second threshold value) as a business partner company candidate. The process of step S330 is performed for all the companies stored in the score database 145 with respect to the second score of the latest securities report. As a result, it is possible to always select business partner candidates from the latest securities report.

Next, in step S340, the control unit 12 causes the terminal device 20 to display the company selected in step S330. Specifically, the output unit 126 generates display data and transmits it to the terminal device 20, so that the company list selected by the company selection unit 124 is displayed on the terminal device 20 as the second approach information. For example, the output unit 126 generates display data so that the selected companies are displayed at the top in descending order of the second score. The process of step S340 may be executed by the output unit 126 by receiving a display request for approach information from the terminal device 20. On the other hand, the processes of steps S310 to S330 may be automatically executed by periodic crawling of a specific management information providing site, or may be executed by a request from the terminal device 20. ..

FIG. 18 is a diagram showing a specific example of a display screen displayed on the display unit 26 of the terminal device 20. On the display screen SC1 of FIG. 18, the first business partner company candidate selected by the sales support device 10 is displayed in a list as the first approach information, and the second business partner company candidate selected by the sales support device 10 is the second. Listed as approach information.

The display screen SC1 of FIG. 18 illustrates a case where the first approach information and the second business partner company candidate are displayed at the same time, but the present invention is not limited to this, and either the first approach information or the second business partner company candidate is used. One of the displays may be selectable. The sales support device 10 generates display data for displaying both or one of the first approach information and the second business partner company candidate in response to the request from the terminal device 20, and transmits the display data to the terminal device 20.

The display items of the first approach information are "date", "company name", "stock code", "first business information", "first score", "remarks" and the like. The display items of the second approach information are "date", "company name", "stock code", "second business information", "second score", "keyman", "equipment" and the like. The higher the first score, the higher the first business partner company candidate is displayed, and the higher the second score, the higher the second business partner company candidate is displayed. As a result, companies with a high probability of successful sales can be displayed at the top.

The display items of the first approach information and the second approach information are not limited to those shown in FIG. For example, items such as "listed market", "capital", "industry", "business type", "number of employees", "closing date", and "location" may be provided. An item of the search period may be provided so that the period can be entered in the item, and the companies selected by the analysis of the first management information 143A or the second management information 143B within the entered period may be narrowed down and displayed.

Further, an item of "evaluation" may be provided. In the item of "evaluation", as in the case of the first embodiment, it is possible to input a numerical value for evaluating whether or not it was useful for sales in a plurality of stages such as 3 stages and 5 stages. In addition, a link is attached to the characters displayed in the item of "first management information" as in the first embodiment so that the document data of the first management information 143A of the management information database 143 is displayed. You may. A link may be attached to the characters displayed in the item of "second management information" so that the document data of the second management information 143B of the management information database 143 is displayed.

FIG. 19 is a diagram showing another specific example of the display screen displayed on the display unit 26 of the terminal device 20. The display screen SC2 of FIG. 19 illustrates a case where comprehensive approach information is displayed in addition to the first approach information and the second approach information similar to those of FIG. The comprehensive approach information displays a list of business partner company candidates that are ranked by comprehensively ranking the first approach information and the second approach information. In the comprehensive approach information, business partner candidates are listed in descending order of overall score.

The total score is calculated by adding the first score and the second score for each company, for example. Specifically, Company A in FIG. 19 has a first score of "70" and a second score of "80", so the total score is "150", and Company B has a first score of "80" and is the first. Since the 2 scores are "50", the total score is "100". Therefore, in the comprehensive approach information, the company A having a high overall score is displayed higher overall than the company B. The method of calculating the total score is not limited to the example. For example, the total score may be calculated by adding the first score and the second score to which the weighting coefficient is added. The first approach information, the second approach information, and the comprehensive approach information can be scrolled on the display screen SC1 of FIG. 18 and the display screen SC2 of FIG. 19 so that lower-level companies are also displayed.

According to the sales support device 10 according to the second embodiment as described above, as in the first embodiment, a company having a possibility of real estate transaction demand can be automatically selected as a business partner company candidate from the management information of the company. .. Therefore, it is possible to significantly reduce the time and effort required to select a business partner company candidate, and to improve the efficiency of sales activities. Further, in the second embodiment, when the same specific keyword is used for different types of management information by changing the type of the specific keyword according to the type of management information such as the first management information 143A and the second management information 143B. In comparison with, it is possible to significantly enhance the effect of linking the analysis result of management information by specific keyword judgment to the selection of promising business partner company candidates. Therefore, it is possible to greatly improve the accuracy of selecting a promising business partner candidate. Further, as shown in FIGS. 18 and 19, the business partner company candidate based on the first management information 143A and the business partner company candidate based on the second management information 143B are displayed separately, so that the business partner company candidate is selected according to which management information. It becomes easy to understand whether it was done. This makes it easier to formulate sales strategies for companies that are candidates for business partners.

<Third Embodiment>
A third embodiment of the present invention will be described. In the specific keyword determination process (step S120 of FIG. 6) of the first embodiment, a case where the score of the management information is calculated based on the number of appearances and the weighting coefficient of the specific keyword included in the document data of the management information is illustrated. On the other hand, in the specific keyword determination process of the third embodiment, a case where a score of management information is acquired by a learned model (AI: Artificial Intelligence) is illustrated. FIG. 20 is a block diagram showing a specific configuration example of the sales support system 100 according to the third embodiment, and corresponds to FIG. The storage unit 14 of FIG. 20 differs from FIG. 2 in that it stores the trained model 146.

FIG. 21 is a diagram showing a configuration example of the trained model 146. The trained model 146 of FIG. 21 is trained to output the score of the management information by inputting the appearance count information of the specific keyword included in the document data of the management information. For example, the specific keyword is each specific keyword included in each of the keyword groups in the trend prediction keyword.

The appearance count information of the specific keyword is, for example, information on the appearance count of the specific keyword associated with each keyword group. Specifically, the appearance count information of the specific keyword is the word hit information HGA1 to HGAn of each keyword group shown in FIG. 4, and these are separately input to the trained model 146. However, the appearance count information of the specific keyword is not limited to the word hit information, and for example, the appearance count HA11, HA12, ..., HA21, HA22, ... Of each specific keyword included in each keyword group shown in FIG. 4 are separate. May be input to the trained model 146.

In this way, information on the number of occurrences of a specific keyword associated with each of the keyword groups representing different characteristics of corporate management related to real estate transactions (for example, word hit information) is separately input to the trained model 146. .. As a result, it is possible to construct a learned model 146 in which the characteristics of corporate management related to real estate transactions are more likely to be reflected than when the information on the number of occurrences of a specific keyword is collectively input. By selecting a business partner candidate from the output score of such a learned model 146, it is possible to improve the accuracy of selecting a leading business partner company.

FIG. 21 illustrates a trained model 146 composed of a neural network trained by machine learning. The trained model 146 includes an input layer, an intermediate layer, and an output layer. Each layer is composed of multiple nodes. For example, in the trained model 146 in which the word hit information of the n keyword groups shown in FIG. 4 is input as the appearance count information of the specific keyword and the score of the management information is output, the number of nodes in the input layer is n and the output layer. The number of nodes in is 1. The number of nodes included in each layer and the number of layers in the intermediate layer are not limited to those shown in the figure. The trained model 146 may be constructed by CNN (Convolutional Neural Network) or may be configured by RNN (Recurrent Neural Network).

The trained model 146 is a deep learning model constructed by supervised learning using, for example, a deep learning algorithm. Specifically, the trained model 146 learns using teacher data in which the number of appearances information of a specific keyword included in the document data of management information and the score (correct answer label) of the management information are set as one set. In the trained model 146, when the appearance count information of a specific keyword included in the teacher data is input to the input layer, the output of the output layer corresponding to the score approaches the score indicating the correct answer label included in the teacher data. Perform processing. In the learning process, the trained model 146 optimizes data such as coefficients and thresholds of functions that define predetermined operations performed on inputs.

The management information analysis process performed by the sales support device 10 according to the third embodiment is the same as that shown in FIG. In the third embodiment, the score acquisition method (step S122 in FIG. 7) is different from that in the first embodiment. In step S122 of the specific keyword determination process shown in FIG. 7, in the first embodiment, the score is calculated and acquired based on the number of occurrences of the specific keyword and the weighting coefficient, whereas in the third embodiment, learning is performed. The difference is that the score is obtained from the output of the finished model 146. Specifically, in the third embodiment, in step S122, the keyword determination unit 122 acquires a score from the output of the learned model 146 obtained by inputting the number of times the specific keyword appears.

In this way, by using the learned model 146 as shown in FIG. 21 for the specific keyword determination process of the management information and inputting the specific keyword appearance count information, the weighting coefficient of the specific keyword is preliminarily set as in the first embodiment. The weight-optimized score can be obtained by the trained model 146 without setting. As a result, it is possible to effectively connect the analysis result of the management information by the specific keyword judgment to the selection of a promising business partner company candidate. As the number of occurrences information used for input, normalized or standardized information can be used as teacher data.

FIG. 22 is a diagram showing another configuration example of the trained model 146. The trained model 146 of FIG. 22 is trained to output a score by inputting management information and specific keyword information. The trained model 146 of FIG. 22 is also composed of a neural network trained by machine learning as in FIG. Specifically, the trained model 146 of FIG. 22 is trained using teacher data in which management information, specific keyword information, and a score (correct answer label) are set as one set.

In the trained model 146 of FIG. 22, when the management information included in the teacher data and the specific keyword information are input to the input layer, the output of the output layer corresponding to the score is used as a score indicating the correct answer label included in the teacher data. Perform learning processing that approaches.

The management information input to the learned model 146 is, for example, numerical values (for example, vectorization) of keywords (words, etc.) included in document data such as a business plan and a securities report by a known method. The specific keyword information is the numerical value (for example, vectorization) of the specific keyword associated with each keyword group by a known method. For example, each specific keyword KA11, KA12, ..., KA21, KA22, ... Included in each of the keyword groups shown in FIG. 4 is quantified. As the management information and the specific keyword information, for example, digitized data that is normalized or standardized can be used as input data or teacher data.

In this way, by using the learned model 146 as shown in FIG. 22 for the specific keyword determination process of the management information, the score can be obtained from the output of the learned model 146 by inputting the management information and the specific keyword information. This makes it possible to simplify the specific keyword determination process.

When the trained model 146 of FIGS. 21 and 22 is trained, the score indicating the correct answer label included in the teacher data is set to the score obtained from the knowledge and experience of the skilled salesman, so that the knowledge of the skilled salesman can be obtained. And experience can be learned. As a result, the trained model 146 can be constructed so as to output the optimum score as calculated by a skilled sales person, so that it is possible to efficiently find a promising business partner candidate who has a demand for real estate transactions. The accuracy of the score can be improved by additionally learning the trained model 146.

<Fourth Embodiment>
A fourth embodiment of the present invention will be described. In the specific keyword determination process of the fourth embodiment, the case where the first score of the first management information 143A and the second score of the second management information 143B of the second embodiment are acquired by separate learned models (AI) is used. Illustrate. FIG. 23 is a block diagram showing a specific configuration example of the sales support system 100 according to the fourth embodiment, and corresponds to FIG. The storage unit 14 of FIG. 23 differs from FIG. 8 in that it stores two first trained models 146A and a second trained model 146B. Further, in order to make the explanation easy to understand, the specific keywords stored in the keyword database 144 (keyword DB) are the first specific keyword 144a and the second specific keyword 144b.

FIG. 24 is a diagram showing a configuration example of the first trained model 146A, and FIG. 25 is a diagram showing a configuration example of the second trained model 146B. The first trained model 146A of FIG. 24 is trained to output a first score by inputting information on the number of appearances of the first specific keyword 144a included in the first management information 143A. The second trained model 146B of FIG. 25 is trained to output a second score by inputting information on the number of appearances of the second specific keyword 144b included in the second management information 143B.

The first specific keyword 144a is each specific keyword included in each of the keyword groups in the trend prediction keyword 144A, as in the second embodiment. The appearance count information of the first specific keyword 144a is, for example, information on the appearance count of each specific keyword associated with each of the keyword groups in the trend prediction keyword 144A. Specifically, the appearance frequency information of the first specific keyword 144a is the word hit information HGA1 to HGAn of each keyword group in the trend prediction keyword 144A shown in FIG. 12, and these are separately input to the first trained model 146A. NS. However, the appearance count information of the first specific keyword 144a is not limited to the word hit information. For example, the appearance count information of each specific keyword included in each keyword group in the trend prediction keyword 144A shown in FIG. 12 HA11, HA12, ..., HA21. , HA22, ... May be input to the first trained model 146A separately.

The second specific keyword 144b is each specific keyword included in each of the keyword groups in the keyman keyword 144B and the equipment status keyword 144C, as in the second embodiment. The second specific keyword 144b may be a specific keyword included in only one of the keyman keyword 144B and the equipment status keyword 144C. The appearance count information of the second specific keyword 144b is, for example, information on the appearance count of each specific keyword associated with each of the keyword groups in the keyman keyword 144B and the equipment status keyword 144C.

Specifically, the appearance count information of the second specific keyword 144b includes word hit information HGB1 to HGBn of each keyword group in the keyman keyword 144B shown in FIG. 13 and word hit information HGC1 to HGCn of each keyword group in the equipment status keyword 144C. And these are separately input to the second trained model 146B. However, the appearance count information of the second specific keyword 144b is not limited to the word hit information. For example, the appearance count information of each specific keyword included in each keyword group of the keyman keyword 144B shown in FIG. 13 HB11, HB12, ..., HB21, The second trained model 146B separately includes HB22, ..., And the number of occurrences of each specific keyword included in each keyword group of the equipment status keyword 144C, HC11, HC12, ..., HC21, HC22, ... It may be input to.

The first trained model 146A and the second trained model 146B are each composed of a neural network trained by machine learning. The first trained model 146A and the second trained model 146B each include an input layer, an intermediate layer, and an output layer. Each layer of the input layer, the intermediate layer, and the output layer is composed of a plurality of nodes.

For example, in the trained model 146A in which n word hit information of the trend prediction keyword 144A shown in FIG. 12 is input as the appearance count information of the first specific keyword 144a and the first score is output, the number of nodes in the input layer is n. The number of nodes in the output layer is 1. Further, as the appearance count information of the second specific keyword 144b, n word hit information of the keyman keyword 144B and n word hit information of the equipment status keyword 144C shown in FIG. 13 are input, and the second score is output. In model 146B, the number of nodes in the input layer is n + n, and the number of nodes in the output layer is 1. The number of nodes included in each layer and the number of layers in the intermediate layer are not limited to those shown in the figure. The first trained model 146A and the second trained model 146B may be composed of CNN or RNN.

The first trained model 146A and the second trained model 146B are deep learning models constructed by supervised learning using, for example, a deep learning algorithm. Specifically, the first trained model 146A uses teacher data in which the number of occurrences of the first specific keyword 144a included in the document data of the first management information 143A and the first score (correct answer label) are set as one set. learn. The first trained model 146A sets a correct answer label in which the output of the output layer corresponding to the first score is included in the teacher data when the appearance count information of the first specific keyword 144a included in the teacher data is input to the input layer. The learning process is performed so as to approach the first score shown.

The second trained model 146B learns using the teacher data in which the appearance number information of the second specific keyword 144b included in the document data of the second management information 143B and the second score (correct answer label) are set as one set. The second trained model 146B sets a correct answer label in which the output of the output layer corresponding to the second score is included in the teacher data when the appearance count information of the second specific keyword 144b included in the teacher data is input to the input layer. The learning process is performed so as to approach the second score shown.

The first score and the second score, which indicate the correct answer labels included in the teacher data, can be made into scores obtained from the knowledge and experience of the skilled sales staff so that the knowledge and experience of the skilled sales staff can be learned. As a result, the first trained model 146A and the second trained model 146B can be constructed so as to output the optimum score as calculated by the sales staff.

As described above, in the fourth embodiment, the first specific is provided by providing two trained models 146A and trained models 146A for separately analyzing different types of first management information 143A and second management information 143B. Since the output of the first score by inputting the keyword appearance count information and the output of the second score by the second specific keyword appearance count information can be obtained separately, the output of the first score and the second score can be obtained with one trained model. It is possible to improve the analysis accuracy of the first management information and the second management information as compared with the case of obtaining the output of.

By using the trained model 146A as shown in FIG. 24 for the first specific keyword determination process and inputting the first specific keyword appearance count information of the first management information 143A, the weighting coefficient is preliminarily set as in the second embodiment. Even if it is not set, the first score whose weight is optimized by the first trained model 146A can be obtained. As a result, the analysis result of the first specific keyword determination of the first management information 143A can be effectively linked to the selection of a promising business partner company candidate.

By using the trained model 146B as shown in FIG. 25 for the second specific keyword determination process and inputting the appearance count information of the second specific keyword 144b of the second management information 143B, the weighting coefficient is as in the second embodiment. The second score whose weights are optimized by the second trained model 146B can be obtained without setting in advance. As a result, the analysis result of the second specific keyword determination of the second management information 143B can be effectively linked to the selection of a promising business partner company candidate.

FIG. 26 is a diagram showing another configuration example of the first trained model 146A. The first trained model 146A is composed of, for example, a neural network learned by machine learning, and is trained to output a first score by inputting the first management information 143A and the first specific keyword information. There is. Specifically, the first trained model 146A of FIG. 26 is trained using teacher data in which the first management information 143A, the first specific keyword information, and the first score (correct answer label) are set as one set.

In the first trained model 146A of FIG. 26, when the first management information 143A and the first specific keyword information included in the teacher data are input to the input layer, the output of the output layer corresponding to the first score is the teacher data. The learning process is performed so as to approach the first score indicating the correct answer label included in.

The first management information 143A to be input to the first trained model 146A is, for example, numerical values (for example, vectorization) of keywords (words and the like) included in document data such as a business plan by a known method. The first specific keyword information is, for example, numerical values (for example, vectorization) of specific keywords associated with each keyword group in the trend prediction keyword 144A of FIG. 12 by a known method. As the management information and the specific keyword information, for example, digitized data that is normalized or standardized can be used as input data or teacher data.

FIG. 27 is a diagram showing another configuration example of the second trained model 146B. The second trained model 146B is composed of, for example, a neural network learned by machine learning, and is trained to output a second score by inputting the second management information 143B and the second specific keyword information. There is. Specifically, the second trained model 146B of FIG. 27 is trained using the teacher data in which the second management information 143B, the second specific keyword information, and the second score (correct answer label) are set as one set. In the second trained model 146B of FIG. 27, when the second management information 143B and the second specific keyword information included in the teacher data are input to the input layer, the output of the output layer corresponding to the second score is the teacher data. The learning process is performed so as to approach the second score indicating the correct answer label included in.

The second management information 143B to be input to the second trained model 146B is, for example, a numerical value (for example, vectorization) of keywords (words, etc.) included in document data such as a securities report by a known method. The second specific keyword information is, for example, numerical values (for example, vectorization) of specific keywords associated with each keyword group of the keyman keyword 144B and the equipment status keyword 144C in FIG. 13 by a known method. As the second management information 143B and the second specific keyword information, for example, digitized data that is normalized or standardized can be used as input data or teacher data.

In this way, by using the first learned model 146A as shown in FIG. 26 for the first specific keyword determination process of the first management information 143A, the first management information 143A and the first specific keyword information can be input. 1 The first score can be obtained from the output of the trained model 146A. This makes it possible to simplify the first specific keyword determination process. Further, by using the second learned model 146B as shown in FIG. 27 for the second specific keyword determination process of the second management information 143B, the second learning has been completed by inputting the second management information 143B and the second specific keyword information. A second score can be obtained from the output of model 146B. This makes it possible to simplify the second specific keyword determination process.

In the fourth embodiment, two first trained models 146A and two trained models 146B that acquire scores from two different types of first management information 143A and second management information 143B have been exemplified. The trained model is not limited to two. When calculating the score of three or more types of management information, three or more learned models may be provided for each type of management information.

Further, in the fourth embodiment, the case where different scores are acquired by different trained models for each management information is illustrated, but the present invention is not limited to this, and for example, as shown in FIG. 28, a plurality of trained models are used. The total score may be obtained for each company from the management information of.

FIG. 28 is a diagram showing another configuration example of the trained model. FIG. 28 illustrates a case where one third learned model 146C acquires a total score as shown in FIG. 19 from the first management information 143A and the second management information 143B for each company. FIG. 19 illustrates a case where the score acquisition unit 123 calculates and acquires the total score, but FIG. 28 illustrates a case where the score acquisition unit 123 acquires the total score from the output of the third trained model 146C. ..

The third learned model 146C includes information on the number of occurrences of the first specific keyword 144a associated with the first management information 143A of one company and the second specific keyword 144b associated with the second management information 143B of the company. It is learned to output the total score of the company by inputting the appearance count information. The total score is associated with each company and stored as total score information in the score database 145 of the storage unit 14.

The appearance count information of the first specific keyword 144a associated with the first management information 143A is information on the appearance count of the first specific keyword 144a in the first management information 143A. For example, it is information on the number of occurrences of each specific keyword associated with each of the keyword groups in the trend prediction keyword 144A. Specifically, it is the word hit information HGA1 to HGAn of each keyword group in the trend prediction keyword 144A shown in FIG. However, the present invention is not limited to this, and the appearance count information of the first specific keyword 144a may be, for example, the appearance count of each specific keyword included in each keyword group in the trend prediction keyword 144A shown in FIG.

The appearance count information of the second specific keyword 144b associated with the second management information 143B is information on the appearance count of the second specific keyword 144b in the second management information 143B. For example, it is information on the number of occurrences of each specific keyword associated with each of the keyword groups in the keyman keyword 144B and the equipment status keyword 144C. Specifically, the word hit information HGB1 to HGBn of each keyword group in the keyman keyword 144B shown in FIG. 13 and the word hit information HGC1 to HGCn of each keyword group in the equipment status keyword 144C. However, not limited to this, the appearance count information of the second specific keyword 144b is included in, for example, the appearance count of each specific keyword included in each keyword group of the keyman keyword 144B shown in FIG. 13 and each keyword group of the equipment status keyword 144C. It may be one or both of the number of occurrences of each specific keyword.

The third trained model 146C is composed of a neural network trained by machine learning. The third trained model 146C includes an input layer, an intermediate layer, and an output layer. Each layer of the input layer, the intermediate layer, and the output layer is composed of a plurality of nodes. The number of nodes included in each layer and the number of layers in the intermediate layer are not limited to those shown in the figure. The third trained model 146C may be composed of CNN or RNN.

The third trained model 146C is a deep learning model constructed by supervised learning using, for example, a deep learning algorithm. Specifically, in the third learned model 146C, the appearance frequency information of the first specific keyword 144a (first specific KW) associated with the first management information 143A and the second specific keyword associated with the second management information 143B Learning is performed using the teacher data in which the appearance number information of 144b (second specific KW) and the total score (correct answer label) are set as one set. The third learned model 146C uses the appearance count information of the first specific keyword 144a associated with the first management information 143A and the appearance count information of the second specific keyword 144b associated with the second management information 143B as an input layer. When input, the learning process is performed so that the output of the output layer corresponding to the total score approaches the total score indicating the correct answer label included in the teacher data.

In the storage unit 14 of FIG. 23, the third trained model 146C may be additionally stored in the first trained model 146A and the second trained model 146B. In this case, as shown in FIG. 19, the first score from the first trained model 146A is displayed in the first approach information, and the second score from the second trained model 146B is displayed in the second approach information. 3 The total score from the trained model 146C is displayed in the comprehensive approach information. The storage unit 14 of FIG. 23 may store the third trained model 146C instead of the first trained model 146A and the second trained model 146B.

According to the third trained model 146C of FIG. 28, the information on the number of occurrences of the specific keyword associated with each of the plurality of different types of management information is input separately, so that the number of appearances of the specific keyword in the plurality of management information is summarized. It is possible to improve the accuracy of the total score as compared with the case of inputting.

<Fifth Embodiment>
A fifth embodiment of the present invention will be described. The fifth embodiment exemplifies a case where a change in the management policy is estimated from the transition of the score of the management information accumulated for each company and a company having the change in the policy is selected. A company whose management policy has changed is likely to be the one with the best timing for opening a business. Since the content of management information changes when the management policy changes, there is a high possibility that the specific keywords used will also change. Therefore, for example, if the number of specific keywords increases and the score of management information increases, it can be seen that the possibility of real estate transaction demand has increased.

Therefore, in the present embodiment, a company having a change in the management information is found and extracted from the transition of the score of the specific keyword of the management information. As a result, companies whose management policy has changed can be extracted, and it becomes possible to prioritize the approach to the company with the best timing for opening sales.

In the fifth embodiment, the score of the management information of each period (each next) is accumulated for each company in the score database 145 (score DB) of FIG. In the process of selecting the business partner company candidate of the fifth embodiment performed by the company selection unit 124 of FIG. 2, the score of the management information acquired by the score acquisition unit 123 is equal to or higher than a predetermined threshold value and immediately before the score database 145 is acquired. Select a company with a score higher than the above as a candidate for a business partner.

The process of selecting a business partner company candidate according to the fifth embodiment is executed, for example, by the procedure shown in FIG. 29. FIG. 29 is a flowchart showing a specific example of the business partner company candidate selection process of the fifth embodiment, and corresponds to a modified example of step S130 of FIG.

In the fifth embodiment, in step S130 of FIG. 6, the control unit 12 selects a company whose latest management information score is higher than the score of the immediately preceding management information as a business partner company candidate. Specifically, in step S131 of FIG. 29, the company selection unit 124 determines whether or not the score (latest score) of the latest management information stored in the score database 145 is equal to or higher than a predetermined threshold value. When the company selection unit 124 determines in step S131 that the score of the latest management information is equal to or higher than a predetermined threshold value, the score database 145 in step S132 indicates the score of the management information immediately before the company (score immediately before). ) To get.

Next, in step S133, the company selection unit 124 determines whether or not the score of the latest management information is higher than the score of the immediately preceding management information. When the company selection unit 124 determines in step S133 that the score of the latest management information is higher than the score of the immediately preceding management information, the company selection unit 124 selects the company as the management company in step S134.

When the company selection unit 124 determines that the score of the latest management information is not equal to or higher than a predetermined threshold value (NO in step S131), or when it determines that the score is not higher than the score of the immediately preceding management information (NO in step S133). , Do not select the company as a management company.

The process of step S130 of the fifth embodiment is performed for the scores of the latest management information for all the companies stored in the score database 145. When the selection process for all the companies is completed, the control unit 12 causes the terminal device 20 to display the companies selected in step S130 in step S140 of FIG. Specifically, the output unit 126 generates display data and transmits it to the terminal device 20, so that the company list selected by the company selection unit 124 is displayed on the terminal device 20 as approach information. For example, the output unit 126 generates display data so that the selected companies are displayed at the top in descending order of score.

As described above, according to the fifth embodiment, even the change in the score from the immediately preceding management information, which cannot be understood only from the latest management information score, can be known. In other words, it can be inferred whether there has been a change in management policy that was not possible with the latest management information score alone. According to the fifth embodiment, a company whose latest management information score is equal to or higher than a predetermined threshold value and higher than the previous management information score is selected as a business partner candidate, so that the management policy tends to increase the score. Companies that have changed can be preferentially extracted and displayed.

As a result, it is possible to predict the optimal company for the timing of sales from the transition of the score of the management information without relying on the experience and intuition of a skilled sales person, and it is possible to display such a company list. Therefore, the time and effort required to select a candidate for a business partner can be significantly reduced, and sales activities can be made more efficient than when a company is selected based only on the latest management information score.

(First modification of the fifth embodiment)
Next, as a first modification of the fifth embodiment, a case where the fifth embodiment of extracting a company from the first score of the first management information 143A (medium-term management plan, etc.) according to the second embodiment is applied is exemplified. do. Here, even if the first score of the latest first management information 143A is the same, the case where the first score has not changed and the case where the first score has changed from the immediately preceding first management information will be compared and described.

FIG. 30 is a diagram showing a comparative example of the transition of the first score of the first management information 143A, and FIG. 31 is a diagram showing a specific example of the transition of the first score of the first management information. 30 and 31 are graphs showing the transition of the first score of the first management information 143A from the first period to the third period. The horizontal axis is the time and the vertical axis is the first score. The third period is the first score of the latest first management information 143A, and the second period is the first score of the immediately preceding first management information.

The first score of the latest first management information 143A (third period) is the same in FIGS. 30 and 31, but the first score of the immediately preceding first management information (second period) is different. In FIG. 30, although it is relatively high as the first score "80" of the first management information 143A (second period) immediately before, it is the same as the first score "80" of the latest first management information 143A (third period). be. That is, in FIG. 30, the first score of the latest first management information 143A (third period) has not changed from the first score of the immediately preceding first management information (second period).

On the other hand, in FIG. 31, the latest first management information 143A (third period) has the same relatively high first score "80" as in FIG. 30, but the first score of the immediately preceding first management information (second period). It has changed four times larger than "20".

Therefore, in the latest first management information 143A, it is highly possible that there is no change in the management policy in FIG. 30, whereas in FIG. 31, it can be inferred that the management policy has changed significantly. In FIG. 31, for example, as the trend forecast keyword 144A, the first specific keyword such as "asset efficiency improvement" is not included in the immediately preceding first management information 143A, but is included in the latest first management information. This is the case.

It can be inferred that such companies have changed their management policies to aim for "asset efficiency" such as selling and organizing real estate. This is the timing when a company that had a low possibility of reorganization demand for real estate equipment became more likely to have a reorganization demand for real estate equipment. If a company with a change in the first score of the first management information 143A is opened at that timing as shown in FIG. 31, it will be compared with a company having no change in the first score of the first management information as shown in FIG. It can be said that the possibility of concluding a real estate sales contract is much higher.

According to FIGS. 30 and 31, it can be understood that whether or not there has been a change in the management policy of the company cannot be determined by looking only at the first score of the latest first management information 143A. By comparing the first score of the latest first management information 143A with the first score of the immediately preceding first management information, it becomes possible to infer that the management policy has changed.

The process of selecting the first business partner company according to the first modification of the fifth embodiment is executed by, for example, the procedure shown in FIG. 32. FIG. 32 is a flowchart showing a specific example of the first business partner company candidate selection process according to the first modification of the fifth embodiment, and corresponds to the modification of step S230 of FIG.

In the fifth embodiment, in step S230 of FIG. 14, the control unit 12 determines the latest first management information among the companies whose first score of the latest first management information 143A is equal to or higher than a predetermined threshold (first threshold). the first score is to select a larger enterprise than the first score of the first management information immediately before as the first business partners candidate. Specifically, in step S231 of FIG. 32, the company selection unit 124 first sets the first score (latest first score) of the latest first management information 143A stored in the first score information 145A as a predetermined threshold value. Judge whether it is above or not. When the company selection unit 124 determines in step S231 that the first score of the latest first management information 143A is equal to or higher than a predetermined threshold value, in step S232, the first management immediately before the company is found in the score database 145. The first score of the information 143A (the first score immediately before) is acquired.

Next, in step S233, the company selection unit 124 determines whether or not the first score of the latest first management information 143A is larger than the first score of the immediately preceding first management information. When the company selection unit 124 determines in step S233 that the first score of the latest first management information 143A is larger than the first score of the immediately preceding first management information, the company is selected as the first in step S234. Select as a management company.

When the company selection unit 124 determines that the first score of the latest first management information 143A is not equal to or higher than a predetermined threshold value (NO in step S231), it determines that the score is not greater than the first score of the immediately preceding first management information. If this is the case (NO in step S233), the company is not selected as the first management company.

The process of step S230 of FIG. 14 (steps S231 to S234 of FIG. 32) of the first modification is performed for all the companies stored in the score database 145. When the selection process for all the companies is completed, the control unit 12 causes the terminal device 20 to display the company selected in step S230 as the first business destination company supplementary selection in step S240 of FIG. Specifically, the output unit 126 generates display data and transmits it to the terminal device 20, so that the company list selected by the company selection unit 124 is displayed on the terminal device 20 as the first approach information.

For example, the output unit 126 generates display data so that the selected companies are displayed at the top in descending order of the first score. As a result, the first business partner company candidates selected by the sales support device 10 are listed as, for example, the first approach information of the display screen SC1 of FIG. The display screen SC1 may display the amount of change in the score from the immediately preceding first score. For example, in the case of FIG. 31, the score change amount is “60”, which is obtained by subtracting the immediately preceding first score “20” from the latest first score “80”.

For example, when the company A on the display screen SC1 of FIG. 18 is clicked, the transition of the first score from the past to the latest may be displayed as shown on the display screen SC11 of FIG. 33. The display screen SC11 of FIG. 33 is a case where the latest first score is “70”. By displaying the transition of the first score in this way, the change of the first score from the immediately preceding first management information 143A can be confirmed at a glance. Furthermore, by displaying a plurality of past first scores, it is possible to check how the first scores have changed in the past and visualize changes in the management policy of the company.

As described above, according to the first modification of the fifth embodiment, even the change in the first score from the immediately preceding first management information, which cannot be understood only from the latest first management information 143A, can be known. That is, it is possible to infer a change in management policy that was not possible only with the first score of the latest first management information 143A. According to the first modification, a company whose first score of the latest first management information 143A is equal to or higher than a predetermined threshold value and larger than the first score of the immediately preceding first management information is selected as a business partner company candidate. , Companies whose management policy has changed in the direction of increasing the first score can be preferentially extracted and displayed.

As a result, it is possible to predict with a high probability the optimal company for the timing of opening a business from the transition of the first score of the first management information 143A, and it is possible to display a list of such companies. Therefore, it is possible to significantly reduce the time and effort required for selecting a business partner company candidate, and it is possible to further improve the efficiency of sales activities as compared with the case of selecting a company only from the first score of the latest first management information 143A. In addition, by displaying the transition of the first score, not only can you find out the leading business partner candidates without relying on the experience and intuition of a skilled sales person, but you can also see at a glance the optimal timing to open a business to that company. I understand.

In the first modification, the case of applying the fifth embodiment of extracting a company from the first score of the first management information 143A (medium-term management plan, etc.) according to the second embodiment is illustrated, but this is limited to this. It is not something that can be done. The fifth embodiment may be applied when a company is extracted from the second score or the total score of the second management information 143B (securities report, etc.) according to the second embodiment.

(Second modification of the fifth embodiment)
Next, as a second modification of the fifth embodiment, a case where a company is extracted from the number of times the first specific keyword appears in the first management information 143A (medium-term management plan, etc.) accumulated for each company will be illustrated. In the above first modification, by comparing the latest score with the immediately preceding score, it can be inferred whether or not there has been a change in the management policy. On the other hand, in the second modification, by comparing the latest number of appearances of the first specific keyword with the number of appearances of the first specific keyword immediately before, it is not only whether or not the management policy has changed, but also the direction of the management policy. It can be inferred whether it has changed to (the direction of change in management policy).

FIG. 34 is a diagram showing a comparative example of the transition of the number of occurrences of the first specific keyword in the first management information 143A, and FIG. 35 is a diagram showing a specific example of the transition of the number of appearances of the first specific keyword in the first management information. Is. 34 and 35 are graphs showing the transition of the number of appearances of the first specific keyword for each keyword group in the first management information 143A of the first to third periods. The third period is the number of times the first specific keyword appears in the latest first management information 143A, and the second period is the number of times the first specific keyword appears in the immediately preceding first management information.

In FIGS. 34 and 35, the two keyword groups “efficiency keyword” and “financial keyword” in the trend prediction keyword 144A are displayed separately at the top and bottom. The horizontal axis is the time, and the vertical axis is the number of occurrences of the first specific keyword. The number of occurrences of the first specific keyword here is, for example, the number of appearances of each first specific keyword associated with each of the keyword groups in the trend prediction keyword 144A. Specifically, the number of occurrences of the first specific keyword is the word hit information of the "efficiency keyword" and the "financial keyword" among the word hit information HGA1 to HGAn in the trend prediction keyword 144A shown in FIG.

The word hit information HGA1 to HGAn here is "0" if a plurality of first specific keywords included in each keyword group do not appear in the first management information 143A, and a plurality of first specific keywords included in the keyword group. 1 If even one specific keyword appears in the first management information, it is "1". The word hit information HGA1 to HGAn is not limited to this, and may be, for example, the total value of the number of times a plurality of first specific keywords included in the keyword group appear in the first management information 143A.

In FIG. 34, there is no change in the number of appearances of the “efficiency-type keyword” and the “financial-type keyword” in the first management information 143A from the first period to the third period. Therefore, it can be inferred that the company in FIG. 34 has no change in management policy at least for "efficiency system" and "finance system" in the latest first management information 143A.

On the other hand, in FIG. 35, the "efficiency-based keyword" that did not appear until the first management information 143A (second period) immediately before appears in the latest first management information 143A (third period). Moreover, the "financial keywords" that appeared until the first management information 143A (second period) immediately before are no longer appearing in the latest first management information 143A (third period). Therefore, it can be inferred that the company in FIG. 35 changed the management policy from "financial system" to "efficiency system" in the latest first management information 143A.

The keyword group of "financial keywords" includes the first specific keywords such as "interest-bearing debt reduction", and the keyword group of "efficiency keywords" includes the first specific keywords such as "asset efficiency improvement". Therefore, the company in Fig. 35 changed its policy from promoting financial improvement by "reducing interest-bearing liabilities" to "asset efficiency" by more aggressively selling real estate assets. I can guess.

Then, it is highly likely that the real estate transaction will succeed if the business is opened at this timing. Moreover, it is expected that the possibility of successful real estate transactions will further increase if an approach with asset efficiency as the keyword is taken. In this way, in the second modification, it is possible to infer how the management policy has changed, so it becomes easier to formulate an appropriate sales strategy according to the direction of the management policy. Moreover, since the change in the management policy can be inferred from the first management information 143A including the management plan in which the trend of the real estate transaction is easily described, the trend of the real estate transaction can be predicted. In this way, it becomes possible to operate with the optimal approach for each company without relying on the experience and intuition of skilled sales staff.

In the second modification of the fifth embodiment, the first score of the first management information 143A of each period is accumulated in the score database 145 (score DB) of FIG. 8 for each company, and the first management of each period. The number of times the first specific keyword appears in the information 143A is accumulated for each keyword group. In the process of selecting the business partner company candidate of the second modification performed by the company selection unit 124 of FIG. 8, the first score of the latest first management information 143A acquired by the score acquisition unit 123 of FIG. 8 is equal to or higher than a predetermined threshold value. Therefore, there is a keyword group in which the number of appearances of the first specific keyword of the first management information 143A acquired by the score acquisition unit 123 is larger than the number of appearances of the first specific keyword of the first management information 143A acquired from the score database 145. Select a company as a candidate for the first business partner.

The process of selecting the first business partner company according to the second modification of the fifth embodiment is executed by, for example, the procedure shown in FIG. 36. FIG. 36 is a flowchart showing a specific example of the first business partner company candidate selection process according to the second modification of the fifth embodiment, and corresponds to another modification of step S230 of FIG.

In the fifth embodiment, in step S230 of FIG. 14, the control unit 12 determines the first specific keyword of the latest first management information 143A among the companies whose first score of the latest first management information 143A is equal to or higher than a predetermined threshold value. A company whose number of appearances is larger than the number of times of appearance of the first specific keyword in the first management information 143A immediately before is selected as a candidate for the first business destination company. Specifically, in step S235 of FIG. 36, the company selection unit 124 first determines whether the first score (latest first score) of the latest first management information 143A stored in the score database 145 is equal to or higher than a predetermined threshold value. Judge whether or not.

When the company selection unit 124 determines in step S235 that the first score of the latest first management information 143A is equal to or higher than a predetermined threshold value, in step S236, the latest first management of the company is performed from the score database 145. The word hit information of each word group of the information 143A is acquired as the first specific keyword appearance count (latest KW appearance count). Further, in step S236, the company selection unit 124 acquires the word hit information of each word group of the first management information 143A immediately before the company from the score database 145 as the number of times the first specific keyword appears (the number of times the KW appears immediately before). ..

Next, in step S237, the company selection unit 124 determines whether or not there is a keyword group in which the number of appearances of the first specific keyword in the latest first management information 143A is larger than the number of appearances of the first specific keyword in the immediately preceding first management information. do. Specifically, the word hit information of the latest first management information 143A is compared with the word hit information of the immediately preceding first management information for each word group, and the latest word hit information of the first management information is immediately preceding. It is determined whether or not there is a keyword group larger than the word hit information of the first management information.

For example, in FIG. 34, the latest number of appearances of the first specific keyword (third period) is "0" for both "financial keywords" and "efficiency keywords", and the number of appearances of the first specific keyword immediately before (second period) is also It is "0". Therefore, assuming that the keyword groups are only "financial keywords" and "efficiency keywords", in the case of FIG. 34, the latest number of occurrences of the first specific keyword in step S237 is greater than the number of occurrences of the immediately preceding first specific keyword. It is judged that there is no large keyword group.

On the other hand, in FIG. 35, the latest number of appearances of the first specific keyword (third period) is "1" and the number of appearances of the immediately preceding first specific keyword (second period) is "0" for the "financial keyword". Therefore, assuming that the keyword groups are only "financial keywords" and "efficiency keywords", in the case of FIG. 35, the latest number of occurrences of the first specific keyword in step S237 is greater than the number of occurrences of the immediately preceding first specific keyword. It is judged that there is a large keyword group.

When the company selection unit 124 determines in step S237 that the number of appearances of the first specific keyword in the latest first management information 143A is larger than the number of appearances of the first specific keyword in the immediately preceding first management information, the company selection unit 124 determines that there is a keyword group. In step S238, the company is selected as the first management company.

When the company selection unit 124 determines that the first score of the latest first management information 143A is not equal to or higher than a predetermined threshold value (NO in step S235), the number of appearances of the first specific keyword in the latest first management information is immediately preceding. If it is determined that there is no keyword group larger than the number of occurrences of the first specific keyword in the first management information (NO in step S237), that company is not selected as the first management company.

The process of step S230 of FIG. 14 (step S235 to step S238 of FIG. 36) of the second modification is performed for all the companies stored in the score database 145. When the selection process for all the companies is completed, in step S240 of FIG. 14, the control unit 12 causes the terminal device 20 to display the company selected in step S230 as the first business destination company supplementary selection. Specifically, the output unit 126 generates display data and transmits it to the terminal device 20, so that the company list selected by the company selection unit 124 is displayed on the terminal device 20 as the first approach information.

For example, the output unit 126 generates display data so that the selected companies are displayed at the top in descending order of the first score. As a result, the first business partner company candidates selected by the sales support device 10 are listed as, for example, the first approach information of the display screen SC1 of FIG. On the display screen SC1, a keyword group larger than the number of occurrences of the first specific keyword immediately before may be displayed. For example, in the case of FIG. 35, the characters of "efficiency-type keywords" may be displayed as a keyword group larger than the number of occurrences of the first specific keyword immediately before.

For example, when the company A of the first approach information is clicked on the display screen SC1 of FIG. 18, the transition of the latest number of appearances of the first specific keyword from the past can be displayed as shown in the display screen SC12 of FIG. 37. good. By displaying the transition of the number of appearances of the first specific keyword in this way, the change in the number of appearances of the first specific keyword from the immediately preceding first management information 143A can be confirmed at a glance. Furthermore, by displaying a plurality of past occurrences of the first specific keyword, it is possible to confirm the past transition of the number of appearances of the first specific keyword and visualize changes in the management policy of the company.

As described above, according to the second modification of the fifth embodiment, since the change from the immediately preceding first specific keyword appearance count, which cannot be known only from the latest first specific keyword appearance count, can be known, the first specific keyword appearance count. It is possible to preferentially extract and display companies whose management policy has changed in the direction of increasing. Furthermore, by dividing the first specific keyword into keyword groups for each characteristic of corporate management related to real estate transactions, it is possible to know the transition of the number of times the first specific keyword appears for each keyword group. According to this, it is possible to predict the best company to open a business and display a list of such companies.

As a result, the time and effort required to select a candidate for a business partner can be greatly reduced, and sales activities are made more efficient than when a company is selected based on the number of appearances of the first specific keyword in the latest first management information 143A. can. In addition, by displaying the transition of the number of appearances of the first specific keyword, not only can you know the potential sales position company without relying on the experience and intuition of a skilled sales person, but also until the optimal timing to open a business to that company. You can tell at a glance.

(Third variant of the fifth embodiment)
Next, as a third modification of the fifth embodiment, a case where a company is extracted from the number of appearances of the second specific keyword in the second management information 143B (securities report, etc.) according to the second embodiment will be illustrated. In the third modification, the officer information and equipment information are changed by comparing the latest second specific keyword appearance count of the second management information 143B including officer information and equipment information with the immediately preceding second specific keyword appearance count. You can guess if it was there.

FIG. 38 is a diagram showing a comparative example of the transition of the number of appearances of the second specific keyword in the second management information 143B, and FIG. 39 is a diagram showing a specific example of the transition of the number of appearances of the second specific keyword in the second management information. Is. 38 and 39 are graphs showing the transition of the number of appearances of the second specific keyword for each keyword group in the second management information 143B of the first to third periods. The third period is the number of times the second specific keyword appears in the latest second management information 143B, and the second period is the number of times the second specific keyword appears in the immediately preceding second management information.

In FIGS. 38 and 39, two keyword groups, the keyman keyword 144B and the equipment status keyword 144C, are displayed separately at the top and bottom. The horizontal axis is the time, and the vertical axis is the number of occurrences of the second specific keyword. The number of occurrences of the second specific keyword here is, for example, the number of appearances of each second specific keyword associated with each of the keyword group in the keyman keyword 144B and the keyword group in the equipment status keyword 144C.

Specifically, the number of occurrences of the second specific keyword is the word hit information of one of the word hit information HGB1 to HGBn in the keyman keyword 144B shown in FIG. 13 and the word hit information HGC1 of each keyword group in the equipment status keyword 144C. It is word hit information of one of ~ HGCn.

The word hit information HGB1 to HGBn and HGC1 to HGCn here are "0" if a plurality of second specific keywords included in each keyword group do not appear in the second management information 143B, respectively, and are in the keyword group. If even one of the plurality of included second specific keywords appears in the second management information, it is "1". The word hit information HGB1 to HGBn and HGC1 to HGCn are not limited to this, and may be, for example, the total value of the number of times a plurality of second specific keywords included in the keyword group appear in the second management information 143B.

In FIG. 38, there is no change in the number of appearances of the “keyman keyword” and the “equipment status keyword” in the second management information 143B from the first period to the third period. Therefore, it can be inferred that the company in FIG. 34 has no change in management policy at least for "key man" and "equipment status" in the latest second management information 143B.

On the other hand, in FIG. 39, the "keyman keyword" that did not appear until the second management information 143B (second period) immediately before appears in the latest second management information 143B (third period). Therefore, it can be inferred that the company in FIG. 39 has changed in "key man" in the latest second management information 143B. For example, it is conceivable that a person from a bank who had never existed before became an officer due to a change in management policy. Bank graduates have experience and knowledge of real estate transactions such as real estate lending and real estate collateral, so they are likely to be key people who are easier to accept real estate transactions than other officers. Therefore, the timing when a bank graduate becomes an officer is the best time to start a real estate transaction business.

Further, in FIG. 39, the "equipment status keyword" that did not appear until the second management information 143B (second period) immediately before appears in the latest second management information 143B (third period). Therefore, it can be inferred that the company in FIG. 39 has changed in the "equipment status" in the latest second management information 143B. For example, it is conceivable that equipment such as real estate has been acquired due to changes in management policy.

Then, it is highly likely that the real estate transaction will succeed if the business is opened at this timing. For example, if you approach when a bank graduate becomes an officer as a key man, the possibility of successful real estate transactions increases, and if you approach when you acquire a new real estate as equipment status, it includes a proposal to use that real estate. It is thought that the possibility of successful transactions will increase. In this way, in the third modification, changes in the key man and equipment status can be inferred from the second management information 143B, so that it becomes easy to formulate an appropriate sales strategy according to the direction of such a management policy. In this way, it becomes possible to operate with the optimal approach for each company without relying on the experience and intuition of skilled sales staff.

In the third modification of the fifth embodiment, the second score of the second management information 143B of each period is accumulated in the score database 145 (score DB) of FIG. 8 for each company, and the second management of each period. The number of times the second specific keyword appears in the information 143B is accumulated for each keyword group. In the process of selecting the business partner company candidate of the third modification performed by the company selection unit 124 of FIG. 8, the second score of the latest second management information 143B acquired by the score acquisition unit 123 of FIG. 8 is a predetermined threshold value (third threshold value). 2 thresholds) or more, and the number of appearances of the second specific keyword of the second management information 143B acquired by the score acquisition unit 123 is larger than the number of appearances of the second specific keyword of the second management information 143B acquired from the score database 145. Select a company with a large keyword group as a candidate for a second business partner.

The process of selecting the second business partner company according to the third modification of the fifth embodiment is executed by, for example, the procedure shown in FIG. 40. FIG. 40 is a flowchart showing a specific example of the second business partner candidate selection process according to the third modification of the fifth embodiment, and corresponds to the modification of step S330 of FIG.

In the fifth embodiment, in step S330 of FIG. 16, the control unit 12 appears the second specific keyword of the latest second management information among the companies whose second score of the latest second management information 143B is equal to or higher than a predetermined threshold value. number of times to select the larger companies than the second specific keyword number of occurrences of the second management information immediately before as the two business partners candidate. Specifically, in step S335 of FIG. 40, the company selection unit 124 first determines whether the second score (latest second score) of the latest second management information 143B stored in the score database 145 is equal to or higher than a predetermined threshold value. Judge whether or not.

When the company selection unit 124 determines in step S335 that the second score of the latest second management information 143B is equal to or higher than a predetermined threshold value, the company selection unit 124 finds the latest second management of the company from the score database 145 in step S336. The word hit information of each word group of the information 143B is acquired as the second specific keyword appearance count (latest KW appearance count). Further, in step S336, the company selection unit 124 acquires the word hit information of each word group of the second management information 143B immediately before the company from the score database 145 as the number of times the second specific keyword appears (the number of times the KW appears immediately before). ..

Next, in step S337, the company selection unit 124 determines whether or not there is a keyword group in which the number of appearances of the second specific keyword in the latest second management information 143B is larger than the number of appearances of the second specific keyword in the immediately preceding second management information. do. Specifically, the word hit information of the latest second management information 143B is compared with the word hit information of the immediately preceding second management information for each word group, and the latest word hit information of the second management information is immediately preceding. It is determined whether or not there is a keyword group larger than the word hit information of the second management information.

For example, in FIG. 39, the latest number of appearances of the second specific keyword (third period) is "0" for both the "keyman keyword" and the "equipment status keyword", and the number of appearances of the second specific keyword immediately before (second period) is also ". It is "0". In this case, since the second score is "0", if the predetermined threshold value in step S335 is 1 or more, it is considered that the second score of the latest second management information 143B is not equal to or higher than the predetermined threshold value. Even if it is determined in step S335 that the second score is equal to or higher than a predetermined threshold value, it is determined in step S337 that there is no keyword group in which the latest second specific keyword appearance count is larger than the immediately preceding second specific keyword appearance count. NS.

On the other hand, in FIG. 40, the latest number of appearances of the second specific keyword (third period) is "1" and the number of appearances of the immediately preceding second specific keyword (second period) is "0" for the "keyman keyword" and the "equipment status keyword". ". Therefore, in the case of FIG. 39, it is determined in step S337 that there is a keyword group in which the latest second specific keyword appearance count is larger than the immediately preceding second specific keyword appearance count.

When the company selection unit 124 determines in step S337 that the number of appearances of the second specific keyword in the latest second management information 143B is larger than the number of appearances of the second specific keyword in the immediately preceding second management information, the company selection unit 124 determines that there is a keyword group. In step S338, the company is selected as the second management company.

When the company selection unit 124 determines that the second score of the latest second management information 143B is not equal to or higher than a predetermined threshold value (NO in step S335), the number of appearances of the second specific keyword in the latest second management information is immediately preceding. If it is determined that there is no keyword group larger than the number of appearances of the second specific keyword in the second management information (NO in step S337), that company is not selected as the second management company.

The process of step S330 of FIG. 16 (step S335 to step S338 of FIG. 40) of the third modification is performed for all the companies stored in the score database 145. When the selection process for all the companies is completed, in step S340 of FIG. 16, the control unit 12 causes the terminal device 20 to display the company selected in step S330 as the supplementary selection of the second business partner company. Specifically, the output unit 126 generates display data and transmits it to the terminal device 20, so that the company list selected by the company selection unit 124 is displayed on the terminal device 20 as the second approach information.

For example, the output unit 126 generates display data so that the selected companies are displayed at the top in descending order of the second score. As a result, the second business partner company candidates selected by the sales support device 10 are listed as, for example, the second approach information of the display screen SC1 of FIG. On the display screen SC1, a keyword group larger than the number of occurrences of the second specific keyword immediately before may be displayed. For example, in the case of FIG. 35, the characters “keyman keyword” and “equipment status keyword” may be displayed as a keyword group larger than the number of occurrences of the second specific keyword immediately before.

For example, when the company A of the first approach information is clicked on the display screen SC1 of FIG. 18, the transition of the latest number of appearances of the second specific keyword from the past can be displayed as shown in the display screen SC13 of FIG. good. By displaying the transition of the number of appearances of the second specific keyword in this way, the change in the number of appearances of the second specific keyword from the immediately preceding second management information 143B can be confirmed at a glance. Furthermore, by displaying a plurality of past occurrences of the second specific keyword, it is possible to confirm the past transition of the number of appearances of the second specific keyword and visualize changes in the management policy of the company.

As described above, according to the third modification of the fifth embodiment, the change from the number of appearances of the second specific keyword immediately before can be known only from the latest number of appearances of the second specific keyword, so that the number of appearances of the second specific keyword can be found. It is possible to preferentially extract and display companies whose management policy has changed in the direction of increasing. Furthermore, by dividing the second specific keyword related to real estate transactions into a keyword group of key man and equipment status, the transition of the number of appearances of the second specific keyword for each keyword group can be known. According to this, it is possible to predict the optimal company for the timing of opening a business by the second approach different from the first approach, and it is possible to display a list of such companies.

As a result, the time and effort required to select a candidate for a business partner can be greatly reduced, and sales activities are made more efficient than when a company is selected based on the number of appearances of the second specific keyword in the latest second management information 143B. can. In addition, by displaying the transition of the number of times the second specific keyword appears, not only can you find out the leading business partner candidates without relying on the experience and intuition of a skilled sales person, but also until the optimal timing to open a business to that company. You can tell at a glance.

<Modification example>
The present invention is not limited to the above-described embodiments, and various applications and modifications described below are possible, for example. Moreover, it is also possible to appropriately combine one or more arbitrarily selected modes of these modifications and each of the above-described embodiments. Further, it is clear that a person skilled in the art can come up with various modifications or modifications within the scope of the claims, which naturally belong to the technical scope of the present invention. It is understood that.

(1) In the storage unit 24 in the first to fourth embodiments, evaluation information is evaluated for each list, for example, whether or not the list of business partner company candidates displayed in FIGS. 5 and 19 is useful. May be memorized. For example, an item of "evaluation" may be added to the display items of FIGS. 5 and 19, and the evaluation value entered in the "evaluation" item of each list may be used as evaluation information. The evaluation information is associated with the management information of the evaluated list, and is stored in, for example, the score database 145.

In the "evaluation" item, as described above, it is possible to enter a plurality of grades of evaluation for each list, such as whether or not it was useful for sales, such as 3 grades or 5 grades. When evaluating in three stages, for example, enter 1 if it is not useful, 3 if it is useful, and 2 if neither can be said as an evaluation value in the "evaluation" item. Also, when evaluating on a 5-point scale, for example, 1 if it is not useful, 2 if you can make an appointment, 3 if you can make a business, 4 if you are interested, if you reach a contract Enter 5 as an evaluation value in the "evaluation" item. The evaluation value is not limited to these.

Further, in the storage unit 24 in the first to fourth embodiments, for example, sales information indicating whether or not the real estate owned by the business partner candidate displayed in FIGS. 5 and 19 has been sold is evaluated as evaluation information. It may be memorized as. Not only does the sale of real estate lead to high praise, but if such companies can be excluded from the business position, operating efficiency can be improved. For example, an item of "buy and sell" may be added to the display items of FIGS. 5 and 19, and a sell button that can be turned on and off may be provided in the "buy and sell" item of each list. The sell button may have any configuration such as a slider button or a radio button. By default, the sell button is turned off so that you can turn on the sell button if the property is sold. According to this, since at a glance whether the real estate owned by the company displayed as operating partners candidate is sale, to enhance the operating efficiency by multiplying the sales companies that are not yet real estate sale can.

(2) The specific keywords in the first to second embodiments can be added or deleted. For example, keywords extracted by machine learning of management information may be added to the keyword database 144. Specifically, a keyword (co-occurrence word) that co-occurs with the specific keyword in the above embodiment may be extracted by machine learning and added to the keyword database 144 as a specific keyword.

Further, the specific keyword may be added or deleted according to the relevance to the evaluation information described above. For example, the correlation of a specific keyword with respect to the above evaluation information may be obtained, and the specific keyword whose correlation is equal to or less than a predetermined value may be deleted. Further, the regression equation may be obtained by multiple regression analysis or the like with the above evaluation information as the objective variable and the number of occurrences of each specific keyword as the explanatory variable, and the specific keyword whose coefficient of each explanatory variable is equal to or less than a predetermined value may be deleted. As a result, the specific keyword can be determined by the specific keyword having a high correlation with the evaluation information, so that the accuracy of selecting the highly evaluated business partner company candidate can be improved.

(3) In the third embodiment or the fourth embodiment, each of the trained models of FIGS. 21, 22, 24 to 27 illustrates the case where the trained model is composed of a neural network trained by machine learning. The neural network is not limited to the above, and any kind such as SVM (Support Vector Machine) may be used. Each of the above-mentioned trained models exemplifies the case where learning is performed by supervised learning, but any machine learning such as semi-supervised learning and unsupervised learning may be used.

(4) In the input of each learned model in the third embodiment or the fourth embodiment, the number of appearances of the specific keyword may be further input for each management information having a different "year" or "period". For example, in the learned model 146 of FIG. 21, the number of appearances of the specific keyword of the current management information such as this year and this term and the number of appearances of the specific keyword of the past management information such as the previous year and the previous term can be input. As a result, the change over time in the number of appearances of the specific keyword included in the management information can be reflected in the output, so that a noise factor due to the change over time can be eliminated. Moreover, changes in real estate transaction trends due to changes in the current management information compared to the past can be reflected in the output, so it is possible to improve the accuracy of selecting promising business partner candidates from the output score. can.

100 ... Sales support system, 10 ... Sales support device, 11 ... Communication unit, 12 ... Control unit, 121 ... Information acquisition unit, 122 ... Keyword judgment unit, 123 ... Score acquisition unit, 124 ... Company selection unit, 126 ... Output unit , 14 ... Storage unit, 141 ... Program storage unit, 142 ... Corporate information database, 143 ... Management information database, 143A ... First management information, 143B ... Second management information, 144 ... Keyword database, 144a ... First specific keyword, 144b ... 2nd specific keyword, 144A ... Trend prediction keyword, 144B ... Keyman keyword, 144C ... Equipment status keyword, 145 ... Score database, 145A ... 1st score information, 145B ... 2nd score information, 146 ... Trained model, 146A ... 1st trained model, 146B ... 2nd trained model, 146C ... 3rd trained model, 20 ... terminal device, 21 ... communication unit, 22 ... control unit, 24 ... storage unit, 25 ... input unit, 26 ... Display, 30 ... Management information providing server, GA1, GA2 ... Keyword group, GB1, GB2 ... Keyword group, GC1, GC2 ... Keyword group, N ... Network, Sa ... Score, Sa1 ... 1st score, Sa2 ... 2nd score ..

Claims (9)

A storage unit that stores specific keywords related to real estate transactions,
Information acquisition department that acquires management information of multiple companies,
A keyword determination unit that performs a specific keyword determination that determines whether or not the management information acquired by the information acquisition unit includes a specific keyword stored in the storage unit.
A company selection department that selects a company that is determined by the keyword judgment unit to include the specific keyword in the management information as a business partner company candidate, and a company selection department.
It is equipped with an output unit that outputs display data for displaying the business partner candidate.
The management information includes a first management information including a management plan information and a second management information which is a type of management information different from the first management information and includes an officer information.
The specific keyword includes a first specific keyword for predicting the trend of real estate transactions and a second specific keyword for searching for a company having an officer who is a key person in real estate transactions.
The keyword determination unit determines the specific keyword based on the first specific keyword for the first management information, and determines the specific keyword based on the second specific keyword for the second management information. .. The display data, operating according to claim 1 including a business partners candidates selected by the first management information, and operating partners candidates selected by the second management information, the data for the displayed separately Support device. In the storage unit, a plurality of keyword groups in which the specific keywords are grouped and a weighting coefficient associated with these are stored in advance according to the characteristics of corporate management related to the real estate transaction.
The keyword determination unit includes a score acquisition unit that calculates and acquires a score of the management information based on the number of appearances of the specific keyword included in the management information and a weighting coefficient of the keyword group including the specific keyword. ,
The sales support device according to claim 1 or 2 , wherein the display data includes data for displaying the business partner company candidates in descending order of the score of the management information. A trained model trained to output a score by inputting information on the number of occurrences of the specific keyword is stored in the storage unit.
The keyword determination unit includes a score acquisition unit that acquires a score output from the learned model as a score of the management information by inputting information on the number of appearances of the specific keyword included in the management information.
The sales support device according to any one of claims 1 to 3 , wherein the display data includes data for displaying the business partner company candidates in descending order of the score of the management information. In the storage unit, a trained model trained to output a score by inputting the management information and the information of the specific keyword is stored.
The keyword determination unit includes a score acquisition unit that acquires a score output from the learned model as a score of the management information by inputting the management information and the information of the specific keyword.
The sales support device according to any one of claims 1 to 4 , wherein the display data includes data for displaying the business partner company candidates in descending order of the score of the management information. The storage unit includes a score database that accumulates scores of the management information for each company.
The company selection department claims to select a company whose score of the management information acquired by the score acquisition unit is equal to or higher than a predetermined threshold value and larger than the score immediately before the score acquired from the score database as the business partner company candidate. Item 3. The sales support device according to item 3. The storage unit includes a score database that accumulates the score of the management information for each company and stores the number of appearances of a specific keyword of the management information for each keyword group.
In the company selection unit, the score of the management information acquired by the score acquisition unit is equal to or higher than a predetermined threshold value, and the number of appearances of the specific keyword of the management information acquired by the score acquisition unit is acquired from the score database. The sales support device according to claim 3 , wherein a company having the keyword group larger than the number of appearances of the specific keyword in the management information immediately before is selected as the business partner company candidate. The management information includes a first management information including a management plan information and a second management information which is a type of management information different from the first management information and includes an officer information.
The business partner company candidates include the first business partner company candidate and the second business partner company candidate.
In the company selection unit, the score of the first management information acquired by the score acquisition unit is equal to or higher than a predetermined threshold value, and the number of appearances of the first specific keyword of the first management information acquired by the score acquisition unit is A company having the keyword group larger than the number of appearances of the first specific keyword in the first management information immediately before being acquired from the score database is selected as the first business partner company candidate, and is also selected.
The score of the second management information acquired by the score acquisition unit is equal to or higher than a predetermined threshold value, and the number of occurrences of the second specific keyword of the second management information acquired by the score acquisition unit is acquired from the score database. The sales support device according to claim 7 , wherein a company having the keyword group larger than the number of appearances of the second specific keyword in the second management information immediately before is selected as the second business destination company candidate. Acquire management information of multiple companies,
A specific keyword determination is performed to determine whether or not the acquired management information includes a specific keyword related to the real estate transaction stored in the storage unit.
A company determined to include the specific keyword in the management information is selected as a candidate company for business.
It is a process of outputting display data for displaying the business partner company candidate.
The management information includes a first management information including a management plan information and a second management information which is a type of management information different from the first management information and includes an officer information.
The specific keyword includes a first specific keyword for predicting the trend of real estate transactions and a second specific keyword for searching for a company having an officer who is a key person in real estate transactions.
A sales support program that causes a computer to execute a process in which the first management information is determined by the first specific keyword and the second management information is determined by the second specific keyword. ..

Images