JP7178073B2 - Financing possible amount evaluation device, corporate value evaluating device, financing possible amount evaluation method, and financing possible amount evaluation program - Google Patents

Description

The present disclosure relates to a fund procureable amount evaluation device that evaluates the fund procureable amount, a fund procureable amount evaluation method, a fund procureable amount evaluation program, and a corporate value evaluation device that calculates a corporate value from the fund procureable amount.

When investors, investment funds, etc. invest, or when banks, etc. provide loans, the amount of funds that can be procured, which is a factor in the growth potential of the investee company and its forecast, or the assumption for its estimation Enterprise value needs to be evaluated. Many attempts have been made to evaluate corporate value based on future earnings forecasts. For example, Patent Literature 1 and Patent Literature 2 disclose techniques for evaluating corporate value by predicting the future earnings of an evaluation target company based on the financial information of the evaluation target company.

JP 2007-257436 A JP 2009-087219 A

However, in Patent Document 1 and Patent Document 2, since the corporate value is evaluated based on the financial information of the company to be evaluated, the past financial information of the company to be evaluated is required in the evaluation calculation process. Therefore, it may be difficult for newly started companies or companies with a short business history to raise the amount of funds available or assess the corporate value.

There is also a method of evaluating the corporate value of a target company based on the corporate value of similar companies. Possible amount or enterprise valuation becomes difficult.

Therefore, in the present disclosure, based on the amount of financing of a company similar to the company to be evaluated, there is no need to use the past financial information of the company to be evaluated, which was conventionally required. The purpose is to provide a method for evaluating the amount of funds available and a program for evaluating the amount of funds available.

A fund procurement possible amount evaluation device according to one aspect of the present disclosure includes a corporate information database that stores past examples of corporate information and fund raising amounts that are used as a basis for calculating the fund procurement amount, and the evaluation target company's Target company information input part for entering company name, business details, etc. Target company business category estimation part for estimating the business category from the business details of the target company, and records matching the business category are selected from the company information database a business content similarity calculation unit that calculates the business content similarity based on the business content; a similar company selection unit that selects multiple records from those with the highest business content similarity; A business progress similarity calculation unit that calculates business progress similarity based on the number of employees and the number of days since establishment, an area stage coefficient calculation unit that calculates a normalization coefficient based on the area where the company is located and the investment stage, and funding available and a funding available amount calculation unit for calculating the amount.

A method for evaluating the amount of funds that can be raised in one aspect of the present disclosure includes a corporate information database that stores past examples of corporate information and the amount of funds that are used as the basis for calculating the amount of funds that can be raised, and the evaluation target company. A target company business category estimation step of estimating the business category from the business contents of the evaluation target company based on the input of the company name, business details, etc., and the same business category company of selecting records matching the business category from the company information database. A selection step, a business content similarity calculation step of calculating the business content similarity based on the business content, a similar company selection step of selecting a plurality of records from those with a high business content similarity, the number of employees and the progress after establishment A business progress similarity calculation step of calculating the business progress similarity based on the number of days, an area stage coefficient calculation step of calculating a normalization coefficient based on the area where the company is located and the investment stage, and a fund to calculate the fund available amount and a procurable amount calculation step.

The fund raising amount evaluation program in one aspect of the present disclosure includes a corporate information database that stores past examples of corporate information and funding amounts that are the basis for calculating the fund raising amount, and the evaluation target company A target company business category estimation step of estimating the business category from the business contents of the evaluation target company based on the input of the company name, business details, etc., and the same business category company of selecting records matching the business category from the company information database. A selection step, a business content similarity calculation step of calculating the business content similarity based on the business content, a similar company selection step of selecting a plurality of records from those with a high business content similarity, the number of employees and the progress after establishment A business progress similarity calculation step of calculating the business progress similarity based on the number of days, an area stage coefficient calculation step of calculating a normalization coefficient based on the area where the company is located and the investment stage, and a fund to calculate the fund available amount It is a program for causing a computer to execute a procurable amount calculation step.

According to this disclosure, it is possible to calculate the amount of funds that can be raised even for a newly started company or a company with a short business history. In addition, by using the amount of funds that can be raised as a factor in predicting the growth potential of a company, it is possible to help investors make investment decisions.

1 is a diagram illustrating an example of a configuration of a fund raising possible amount evaluation device according to an embodiment of the present disclosure; FIG. It is a figure which shows an example of a structure of a company information database. FIG. 4 is a table showing weights of words for each business and the content of the business in TF-IDF. FIG. 11 is a table showing degrees of topics presumed to be each project in LDA; It is the figure which made the item which is necessary in the fund procurement possible amount calculation section the table. It is a flow chart which shows an example of the flow of the capital-ravable amount evaluation process in a capital-ravable amount evaluation apparatus. 1 is a schematic block diagram showing the configuration of a computer according to an embodiment; FIG.

<Configuration>
Hereinafter, a fund procureable amount evaluation device, a corporate value evaluation device, a fund procureable amount evaluation method, and a fund procureable amount evaluation program according to an embodiment of the present disclosure will be described with reference to the drawings. In addition, in all the drawings for explaining the embodiments, common components are denoted by the same reference numerals, and repeated explanations are omitted.

FIG. 1 shows the overall configuration of a fund procurement possible amount evaluation device 1 according to an embodiment of the present disclosure. This capital raiseable amount evaluation device includes a company information database 101, a target company information input unit 102, a target company business category estimation unit 103, a same business category company selection unit 104, a business content similarity calculation unit 105, It has a similar company selection unit 106 , a business progress similarity calculation unit 107 , an area stage coefficient calculation unit 108 , and a fund procurement possible amount calculation unit 109 . The fund raiseable amount evaluation device is realized by installing a program in a computer. It should be noted that each unit that constitutes the apparatus for evaluating the amount of funds that can be raised may be implemented by being distributed among a plurality of computers.

The company information database 101 stores past examples of company information and the amount of funds that can be raised as a basis for the evaluation device to calculate the amount of funds that can be raised. FIG. 2 shows an example of the configuration of precedent company information held by the company information database 101. As shown in FIG. The company information database 101 stores precedent company information such as company name, business description, number of employees, number of days since establishment, investment stage, location, business category, and fund raising amount. The company name is the name of the company, and the business content indicates the business content described in natural sentences. The number of employees and the number of days elapsed since establishment indicate the number of employees at the time of funding for the record and the number of days elapsed since the establishment of the company. The investment stage is shown as Series A, Series B, M&A, IPO, etc., depending on the number and status of fund raising. The locality indicates the country where the company is located, and the business category defines a plurality of categories in advance and assigns the closest ones. The amount of funds raised indicates the actual amount of funds raised by the company at that point in time.

Note that since one record is created for each fundraising, multiple records may exist when one company raises funds multiple times.

The target company information input unit 102 inputs the company name, business description, number of employees, number of days since establishment, location, and investment stage of the company to be evaluated. Business descriptions are obtained in the form of natural sentences from the articles of incorporation or websites of the target companies.

The target company business category estimation unit 103 estimates the business category from the business content of the evaluation target company. A plurality of business categories are set in advance according to the description prepared in the enterprise information database 101 .

The target business category is obtained, for example, as follows. A set of a sufficient number of business contents and corresponding business categories is stored in advance in the target company business category estimation unit 103 as learning data, or can be acquired from the outside. Next, the business content is cut out for each word. Then, the extracted word string is vectorized by TF-IDF (Term Frequency-Inverse Document Frequency). Assuming that the number of types of all words appearing in the training data is L, this vector has L dimensions, and the elements of each dimension of the vector are calculated by equation (1). where w _b is the weight of word w in business b, f _w,b is the number of times word w appears in business b, |B| is the total number of businesses, f _w,B is word w in all businesses indicates the number of times it appears in .

Since the number of vectors obtained by Equation 1 is enormous (about tens of thousands of dimensions), dimensionality reduction is performed by uniformly treating synonyms that appear in the same context and are correlated by PCA (Principal Component Analysis). Furthermore, using One-vs-rest multilabel SVM (Support Vector Machine), the business category of the company to be evaluated is estimated by identifying whether it belongs to each business category set in advance. By estimating the business category as described above, it is possible to appropriately set a characteristic business category for selecting similar companies from the business content acquired in natural sentences.

The same business category company selection unit 104 selects from the company information database 101 records of companies having business categories that match one or more business categories estimated by the target company business category estimation unit 103 . When a plurality of business categories are estimated, a company matching even one of them may be selected, or a company matching two or more of them may be selected. Here, by once selecting a company based on the business category, it is possible to lighten the load of calculating the degree of business content similarity in the latter stage.

The business content similarity calculation unit 105 calculates the business content similarity between the evaluation target company and the companies selected by the same business category company selection unit 104 .

The business content similarity is calculated, for example, as follows. First, the company to be evaluated and the business contents of the records registered in the company information database 101 to be calculated are cut out for each word. Next, create a document for each business, removing general words with little meaning. Then, the topic vector of each word group is estimated using LDA (Latent Dirichlet Allocation) from the document of the business content of the record registered in the company information database 101 to be evaluated and the calculation target. Specifically, the calculation is performed using Equation 2. Equation 2 is a general formula for LDA. α is the vector that indicates the parameters that determine the distribution of each topic, β is the matrix that indicates the parameters that determine the distribution of words for each topic, _{θd is the distribution of topics that belong to each business content, and zdn is} the topic for each word. , D is each document set, and w _dn is each word.

Using the cosine similarity from the topic vector of each document calculated by Equation 2, the business content similarity between the company to be evaluated and the record registered in the company information database 101 to be calculated is calculated. Specifically, the calculation is performed using Equation 3. However, cos θ _i indicates the degree of similarity between the company to be evaluated and record i, M is the topic vector of the company to be evaluated, and M _i is the topic vector of record i.

The similar company selection unit 106 selects N companies with the highest degree of similarity from the business content similarities calculated by the business content similarity calculation unit 105 . The number of N is set appropriately.

The business progress similarity calculator 107 calculates the business progress similarity between the company to be evaluated and the record registered in the company information database 101 to be calculated. This is calculated by Equation 4, for example, using the number of employees, the number of days since establishment, and the rate of increase in the number of employees. Here, the rate of increase in the number of employees is obtained by dividing the number of employees by the number of days elapsed since establishment. However, t is the company to be evaluated, i is the record registered in the company information database 101 to be calculated, and k _employees, k _{work history} , and k _increase are k _employees + k _{work history} + k _increase = 1. Set heuristically to satisfy In addition, the business progress similarity is not only the number of employees and / or the number of days since establishment and / or the speed of increase in the number of employees, but also sales, profit, capital, total amount of funds raised in the past, number of sales offices / branches, etc. can be calculated as an element.

The area stage coefficient calculation unit 108 calculates an area stage coefficient between the company to be evaluated and the record registered in the company information database 101 to be calculated. For example, it is calculated by Equation 5 using the location area and the investment stage.

The possible fund procurement amount calculation unit 109 calculates the possible fund procurement amount from the N records selected by the similar company selection unit 106 . For example, the business content similarity between the evaluation target company and record i calculated by the business content similarity calculation unit 105 is c _i =cos θ _i , and the business progress similarity calculation unit 107 calculates the business content similarity between the evaluation target company and record i. The business progress similarity degree p _i and the financing amount of the record i are set to F _i , and are calculated by Equation 6. Note that this calculation formula calculates the weight of each record based on the business content similarity and the business progress similarity, and takes a weighted average of the funding amounts for the N records selected by the similar company selection unit 106 . This allows us to calculate the amount of funding available based on the exact amount of funding by placing weight on those that are similar to the company being evaluated.

Further, using the area stage coefficient calculated by the area stage coefficient calculation unit 108, the possible fund procurement amount calculation unit 109 calculates the possible fund procurement amount using Equation (7). At this time, _Ei is the area stage coefficient calculated by Equation 5, and by multiplying the area stage coefficient that takes into account the location area and the investment stage, it is possible to calculate a more accurate amount of funds that can be raised. Become.

Note that the target company business category estimation unit 103 and the same business category company selection unit 104 can be omitted. If the target company business category estimation unit 103 and the same business category company selection unit 104 are omitted, the business category item is unnecessary in the company information database 101 . At this time, the business content similarity calculation unit 105 calculates the business content similarity with all the records in the enterprise information database 101 . Next, when omitting the area stage coefficient calculation unit 108 and calculating the amount of funds that can be raised by Equation 6, the location area and investment stage items are unnecessary in the company information database 101, and in the target company information input unit 102 Also, the location area and investment stage items are unnecessary.

Based on the above basic structure, we calculate the amount of funds that the target company can raise. The business content is written in a specific language, but it is also possible to support multiple languages by connecting an automatic translator before this.

In addition, in the fund procurement amount calculation unit 109, the enterprise value can be calculated by dividing the fund procurement amount by the investor shareholding ratio acquired by the investor at the time of the fund procurement. That is, in Equation 6 or Equation 7, the corporate value F _i ' of record i, the investor shareholding ratio invested at the time of fund raising of record i is s _i , and F _i ' = F _i /s _i By replacing It is possible to configure a corporate value evaluation device that calculates corporate value.

<Specific example>
A specific calculation image of the amount of funds that can be raised by the company to be evaluated in this disclosure will be explained. The business content of the evaluation target company input in the target company information input unit 102 can be input in the form of natural sentences such as "artificial intelligence software development and platform for artificial intelligence development". Such business contents are input to the target company business category estimation unit 103 .

In the target company business category estimation unit 103, for example, the business category of the company to be evaluated is “artificial intelligence software development and artificial intelligence development platform”, and the business information database 101 includes “software contract development” as business 1, business As 2, the case of a record with a business content of "software development platform" will be described. At this time, each business content is extracted word by word by morphological analysis. The business content of the evaluated company is "artificial intelligence, software, development, and artificial intelligence, development, platform", the business content of business 1 is "software, contract development", the business content of business 2 is " The platform for software development.

Next, TF-IDF calculates a vector having the weight of each word as an element, in which a word with a lower frequency of appearance is weighted more heavily. Here, a specific example in which the correspondence table between each business and words is broken down into words will be described with reference to FIG. As shown at the bottom of the table in FIG. 3(a), the total number of occurrences of each word of the business to be evaluated and the business in the business information database 101 is calculated. Here, "software" 3 times, "no" 3 times, "contract" 1 time, "development" 4 times, "platform" 2 times, and "artificial intelligence" 2 times. Then each word is multiplied by the reciprocal of the total number of occurrences. Then, as shown in the table in Fig. 3(b), the weight of each word in the evaluated company is "software" 1/3, "no" 1/3, "contract" 0, "development" 2/4, " “Platform” is 1/2, and “artificial intelligence” is 2/2. Similarly, in business 1, “software” is 1/3, “no” is 1/3, “contract” is 1, “development” is 1/4, “platform” is 0, and “artificial intelligence” is 0. is a word with a large amount of information. In business 2, “software” is 1/3, “no” is 1/3, “contract” is 0, “development” is 1/4, “platform” is 1/2, and “artificial intelligence” is 0, and the weight of “platform” is It is a word with a large amount of information. In fact, log(|B|/f _w,B ) is used as shown in Equation 1 instead of the reciprocal of the number of appearances of the word in all businesses. The dimensions of the vectors are then compressed by PCA for computational and learning efficiency. In PCA, dimensionality is compressed by synthesizing a group of correlated variables, that is, by uniformly handling synonyms that appear in the same context and have a correlation. One-vs-rest multilabel SVM determines whether the business content of the company to be evaluated belongs to a certain business category. Specifically, learning is performed for each business category using the business content and business category of the records in the enterprise information database 101 as learning data. With the business content of the business to be evaluated as an input, the One-vs-rest multilabel SVM determines whether or not the business belongs to each business category. Here, if there are “artificial intelligence,” “application development,” and “telecommunications” as business categories, the business subject to evaluation belongs to the “artificial intelligence” and “application development” businesses, but not to “telecommunications.” The conclusion is that there is no

A specific example in the business content similarity calculation unit 105 is shown. For example, a description will be given of a case where the business content of the company to be evaluated is "to provide dog hotels by matching dog owners with nearby dog lovers." This business content is a record of the company information database 101, and the case where there are businesses 1 to 4 selected by the same business category company selection unit 104 will be described. Project 1 is a “camera app for dog photography that plans events with dogs on a site that matches the transfer of puppies”, and project 2 is a “pet collar system to prevent lost children and an app that embeds GPS in the collar”. "Know the current location of the pet", the business content of business 3 is "providing a mattress with excellent sleep function for dogs", and the business content of business 4 is "a walk service service that provides the route and dog behavior to the owner with a GPS application. We will notify you and send you a photo.” At this time, each business content is extracted word by word, meaningless words (stop words) are deleted, and a document is created for each business content. Specifically, the stop words "no, to, wo, and shi" are removed from the evaluated companies and become "dog, owner, nearby, dog lover, matching, dog, hotel, provide." Similarly, business 1 is "puppy, transfer, matching, site, dog companion, event, planning, dog, photography, use, camera, application", business 2 is "pet, lost child, prevention, collar, system, collar, GPS , embed, app, pet, current location, know”, business 3 is “sleep, function, excellent, mattress, dog, use, provide”, business 4 is “walk, agency, service, route, dog, behavior, GPS, App, Owner, Notification, Photo, Send".

Next, the LDA shown in Equation 2 estimates the topic and the topic distribution of each document. Specifically, Topic 1 = (Dog: 0.6, Matching: 0.3, Hotel: 0.1), Topic 2 = (Dog: 0.4, GPS: 0.4, App: 0.1, Photo: 0.1), Topic 3 = (Sleep: 0.5 , dog: 0.3, mattress: 0.2). Also, as shown in Figure 4, the topic distribution of each document is as follows. Project 3 = (0.15, 0.15, 0.7), Project 4 = (0.15, 0.75, 0.1).

After calculating the topic distribution of each document by the LDA shown in Equation 2, using the cosine similarity shown in Equation 3 from the topic distribution of each document, the business content similarity between the company to be evaluated and each business is calculated. calculate. Specifically, considering each topic as a dimension of a vector, the evaluated company is a vector of M = (0.8, 0.15, 0.05), and the business 1 is a vector of M ₁ = (0.4, 0.4, 0.2). And the cosine similarity is, as shown in Equation 8, the denominator is the sum of the squares of the elements of the company to be evaluated and the sum of the squares of the elements of the business 1, and the square root of the product, and the numerator is the evaluation target Multiplying and adding the elements of the same dimension for the company and business 1 yields 0.8. Since the cosine similarity is used for the business content similarity, the business content similarity between the evaluation target business and the business 1 in this example is 0.80.

A specific example in the business progress similarity calculation unit 107 is shown. For example, when the number of employees of the company to be evaluated is 10, the number of days elapsed since its establishment is 600 days, and the number of employees of business 1, which is a record in the enterprise information database 101, is 30, and the number of days elapsed since its establishment is 500 days. , the rate of increase in the number of employees of the company to be evaluated is 1/60, the rate of increase in the number of employees in business 1 is 3/50, and the business progress similarity is calculated by Equation 4. A small value of 10/30 = 0.33 was adopted for the similarity ratio of the employees of the company to be evaluated and Business 1, and a small value of 500 was used for the similarity ratio of the number of days elapsed since the establishment of the company to be evaluated and Business 1. /600 = 0.83 is adopted, and a small value (1/60)/(3/50) = 0.28 is adopted for the similarity ratio of the rate of increase in the number of employees between the evaluated company and business 1, so Equation 4 is 0.33k _employees + 0.83k _{work history} + 0.28k _increase . In practice, since constant values are set for k number of _employees , k _{career history} , and k _increase , specific figures are calculated from this.

A specific example in the fund procurement possible amount calculation unit 109 is shown. For example, as shown in FIG. 5, for business 1 whose business content similarity, business progress similarity, funding amount, and area stage coefficient are records in the enterprise information database 101 (0.8, 0.5, 10,000,000, 0.8) , (0.6, 0.5, 30,000,000, 0.6) for business 2 and (0.5, 0.6, 5,000,000, 1.5) for business 3. The fundable amount is then calculated by Equation 6. Specifically calculated, the sum of the product of the similarity of business content and the similarity of business progress in business 1, 2 and 3 is 0.8 x 0.5 + 0.6 x 0.5 + 0.5 x 0.6 = 1.0 , the denominator of Equation 6 is 1. Therefore, the fundable amount of the company to be evaluated is 0.8×0.5×10,000,000×0.8÷1＋0.6×0.5×30,000,000×0.6÷1＋0.5×0.6×5,000,000×1.5÷1＝3,200,000＋5,400,000＋2,250,000＝ Calculated as 10,850,000. As a result, it is possible to calculate the business value of the company to be evaluated as the amount that can be raised.

<Process flow>
Using FIG. 6, the flow of processing of the fund raising possible amount evaluation device according to the embodiment of the present disclosure will be described.

First, the business category is estimated using the business content of the evaluation target company input from the target company information input unit 102 . For this, for example, the business content is vectorized by TF-IDF (Equation 1), the dimensions are compressed by PCA, and then the business category is estimated by using One-vs-rest multilabel SVM (step S1).

Subsequently, records of companies having business categories that match the one or more business categories estimated by the target company business category estimation unit 103 are selected from the company information database 101 (step S2).

Next, the company to be evaluated and the business contents of the records registered in the company information database 101 to be calculated are cut out for each word. Here, a document is created for each business content from which general words with little meaning are deleted. Then, the topic vector of each document is estimated using LDA (Latent Dirichlet Allocation) from documents of business contents of records registered in the company information database 101 to be evaluated and to be calculated. Specifically, the calculation is performed using Equation 2. From the topic vector obtained by Equation 2, the business content similarity between the company to be evaluated and the record registered in the company information database 101 is calculated by Equation 3 (step S3).

Subsequently, from the business content similarities calculated by the business content similarity calculation unit 105, N business content similarities having the highest similarities are selected (step S4).

Next, the business progress similarity between the company to be evaluated and the record registered in the company information database 101 to be calculated is calculated by Equation 4 (step S5).

Further, the area stage coefficient between the company to be evaluated and the record registered in the company information database 101 to be calculated is calculated by Equation 5 (step S6).

Finally, using the value calculated by the business content similarity calculation unit 105, the value calculated by the business progress similarity calculation unit 107, and the value calculated by the area stage coefficient calculation unit 108, according to Equation 7, fundraising Calculate the possible amount (step S7).

From the above, it is possible to calculate the amount of funds that the target company can raise.

(program)
FIG. 7 is a schematic block diagram showing the configuration of the computer 801. As shown in FIG. A computer 801 includes a CPU 802 , a main memory device 803 , an auxiliary memory device 804 and an interface 805 .

Here, the details of the program for realizing each function constituting the capital raiseable amount evaluation device 1 according to the embodiment will be described.

The fund raiseable amount evaluation device 1 is implemented in the computer 801 . The operation of each component of the capital raising possibility evaluation device 1 is stored in the auxiliary storage device 804 in the form of a program. The CPU 802 reads out the program from the auxiliary storage device 804, develops it in the main storage device 803, and executes the above processing according to the program. Also, the CPU 802 secures a memory area corresponding to the above-described memory unit in the main memory device 803 according to the program.

Specifically, the computer 801 has a company information database containing the amount of funds that can be raised, and calculates the current amount of funds that can be raised based on the company information of the target company to be evaluated. Calculate the business similarity between the target company and multiple companies in the company information database based on the business details of the target company and multiple companies in the company information database in the fund raising possibility evaluation program to be evaluated A business content similarity calculation step, a similar company selection step of selecting a plurality of company information from those having a high business content similarity with the target company calculated by the business content similarity calculation unit, and a target company and similar company selection unit a business progress similarity calculation step for calculating respective business progress similarities between the multiple companies selected by the similar company selection unit; and a fund procurement amount evaluation program for causing a computer to execute a fund procurement amount calculation step of calculating the fund procurement amount of the company to be evaluated based on.

Auxiliary storage device 804 is an example of a non-transitory tangible medium. Other examples of non-transitory tangible media include magnetic disks, magneto-optical disks, CD-ROMs, DVD-ROMs, semiconductor memories, etc. that are connected via interface 805 . Further, when this program is delivered to the computer 801 via a network, the computer 801 receiving the delivery may develop the program in the main storage device 803 and execute the above process.

Also, the program may be for realizing part of the functions described above. Furthermore, the program realizes the above-described functions in combination with other programs already stored in the auxiliary storage device 804, a so-called difference file (difference program).
may be

The embodiments of the present disclosure have been described above. It is intended to be included within the scope of the invention corresponding to the examples.

REFERENCE SIGNS LIST 1... Fundable amount evaluation device, 101... Company information database, 102 Target company information input unit, 103... Target company business category estimation unit, 104... Same business category company selection unit, 105... Business content similarity calculation unit, 106 ...Similar company selection unit 107...Business progress similarity calculation unit 108...Area stage coefficient calculation unit 109...Fundable amount calculation unit 801...Computer 802...CPU 803...Main storage device 804...Auxiliary storage device, 805 ... interface

Claims (10)

Information on the business of each company, and
about each companyFunding amountAt least selected from the group consisting of information on the performance of a piece of informationIn a fund procurement possible amount evaluation device that has a company information database including the
business content similarity for calculating respective business content similarities between the evaluation target company and the plurality of companies in the business information database based on the business content of the evaluation target company and the plurality of companies in the business information database a degree calculator;
a similar company selection unit that selects a plurality of company information from those having a high business content similarity with the evaluation target company calculated by the business content similarity calculation unit;
The evaluation target company and the multiple companies selected by the similar company selection unit, the number of employees, the number of days since establishment, the speed of increase in the number of employees, sales, profits, capital, the amount of funds raised in the past, and the number of sales offices and branches. to determine the relationship between the evaluation target company and the selected companies usingEach project progress similarity betweencalculatea business progress similarity calculation unit;
The business content similarity and the business progress similarity of the plurality of companies selected by the similar company selection unitand information on the results of fund raising amounts of the plurality of selected companies;A fund procurement amount calculation unit that calculates the fund procurement amount of the evaluated company based on
A fund procurement possible amount evaluation device characterized by comprising: In the business content similarity calculation unit,
a target company business category estimation unit for estimating the business category of the evaluation target company;
Based on the business category estimated by the target company business category estimation unit, a same business category company selection unit that selects company information including the same business category as the company to be evaluated from the company information database;
The capital raiseable amount evaluation device according to claim 1, further comprising: Each area between the evaluation target company and the plurality of companies selected by the similar company selection unit based on the location area of the evaluation target company and the multiple companies selected by the similar company selection unit and the investment stage 3. The capital raising possibility amount evaluation device according to claim 2, further comprising an area stage factor calculating part for calculating a stage factor, wherein the capital raising possibility amount calculating part further uses the area stage factor. The target company business category estimation unit cuts out the business content for each word, vectorizes it by TF-IDF, compresses the dimension of the vector by PCA, and estimates the business category using One-vs-rest multilabel SVM. 4. The equipment for assessing the amount of funds that can be raised according to claim 2 or 3, characterized in that: The business content similarity calculation unit estimates a topic from the word string of the business content by LDA, and calculates the business content similarity using the cosine similarity of each topic. 5. The apparatus for assessing the amount of funds that can be raised according to any one of 4. 6. The business progress similarity calculation unit according to any one of claims 1 to 5, wherein the calculation is performed using the number of employees and/or the number of days since establishment and/or the increase speed of the number of employees. Financing availability evaluation device. The possible fund procurement amount evaluation according to any one of claims 1 to 6, wherein the fund procurement possible amount calculation unit calculates the fund procurement possible amount of the evaluation target company by Equation 1. Device.

However, in the above formula 1, V indicates the amount of funds that can be raised by the evaluation target company, ci indicates the similarity of business content, pi indicates the similarity of business progress, Fi indicates the amount of financing, ci, pi, and Fi each indicate one of the N companies selected by the similar company selection unit. 8. The apparatus for evaluating the amount of funds that can be raised according to any one of claims 1 to 7, wherein the company information database further includes an investor shareholding ratio,
An apparatus for evaluating the amount of funds that can be raised, wherein the amount of funds that can be raised is calculated in consideration of an investor's shareholding ratio in the amount of funds that can be raised. A method for assessing the amount of funds available,
Information on the business of each company, and
Based on the information on the amount of funds raised for each company, the number of employees, the number of days since establishment, the rate of increase in the number of employees, sales, profits, capital, the amount of funds raised in the past, and the number of sales offices and branches. an information processing device having a corporate information database containing at least one information selected from the group of calculating the amount of funds that can be raised based on the corporate information of the company to be evaluated;
The information processing device
business content similarity for calculating respective business content similarities between the evaluation target company and the plurality of companies in the business information database based on the business content of the evaluation target company and the plurality of companies in the business information database a degree calculation step;
a similar company selection step of selecting a plurality of company information from those having a high degree of business content similarity with the evaluation target company calculated in the business content similarity calculation step;
The number of employees, the number of days since establishment, the rate of increase in the number of employees, sales, profits, capital, and the amount of past funding of each of the evaluation target company and the multiple companies selected by the similar company selection step using at least one piece of information selected from the group consisting of the actual amount and the number of business offices/branch offices, calculating the business progress similarity between the evaluation target company and the plurality of selected companies. a business progress similarity calculation step;
The evaluation target company based on the business content similarity and the business progress similarity of the plurality of companies selected by the similar company selection step, and the information on the fund raising amount of the selected companies. A fund procurement amount calculation step that calculates the fund procurement amount of
A funding availability valuation method that performs A funding availability assessment program comprising:
Information about the business of each company, and
Based on the information on the amount of funds raised for each company, the number of employees, the number of days since establishment, the rate of increase in the number of employees, sales, profits, capital, the amount of funds raised in the past, and the number of sales offices and branches. a processor of a computer having a company information database containing at least one information selected from the group consisting of: to the processor;
business content similarity for calculating respective business content similarities between the evaluation target company and the plurality of companies in the business information database based on the business content of the evaluation target company and the plurality of companies in the business information database a degree calculation step;
a similar company selection step of selecting a plurality of company information from those having a high degree of business content similarity with the evaluation target company calculated in the business content similarity calculation step;
The number of employees, the number of days since establishment, the rate of increase in the number of employees, sales, profits, capital, and the amount of past funding of each of the evaluation target company and the multiple companies selected by the similar company selection step using at least one piece of information selected from the group consisting of the actual amount and the number of business offices/branch offices, calculating the business progress similarity between the evaluation target company and the plurality of selected companies. a business progress similarity calculation step;
The evaluation target company based on the business content similarity and the business progress similarity of the plurality of companies selected by the similar company selection step, and the information on the fund raising amount of the selected companies. A fund procurement amount calculation step for calculating the fund procurement amount of
A funding availability assessment program that runs

Images