It asset inventory system

Abstract

To take inventory of IT assets, such as a web site to be managed, and answer the grounds for determination regarding a result thereof in a natural language.SOLUTION: An IT asset inventory system 1 which identifies IT assets to be managed in a management organization includes: an IT asset search unit 12 which collects information on IT assets which are candidates to be managed, based on the content of target company information 15 designated by a user, as information related to the management organization; and a determination unit 13 which obtains an answer by inputting a prompt to a generative AI 3, the prompt including an instruction statement to give an answer about the content of the target IT assets, a designation of the information related to the management organization, whether the target IT assets relate to the management organization or not, and the grounds for the determination, for each of the IT assets of the asset candidates 17 collected by the IT asset search unit 12.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a technology for corporate security measures, and in particular to a technology that is effective when applied to an IT asset inventory system that keeps track of IT assets such as websites managed by a company. [Background technology]

[0002] Security management of publicly accessible websites is crucial for corporate information security. There have been many cases of unauthorized access by third parties, resulting in the leakage of customer information and other data. In particular, large companies that have, in addition to their official corporate website, specific websites for each product or service they handle, or websites dedicated to limited-time campaigns, must also properly manage the security of these websites. In some cases, it may be necessary to manage the entire website, including websites owned by group companies.

[0003] However, in companies with multiple websites, there are cases where the organization that manages them is not even aware that it is the organization that manages them. For example, in companies with overseas bases, the headquarters in Japan may not thoroughly instruct or manage the overseas bases, leaving them to their own devices, resulting in the launch of websites that are not even known to the headquarters in Japan. Furthermore, situations often arise where overseas bases are forced to handle multiple tasks with a small number of staff, resulting in insufficient management of information processing systems, including websites, and websites with security issues may be operating without the headquarters in Japan being aware of them.

[0004] Such websites can occur not only at overseas bases. For example, in large companies with multiple business divisions and group companies, there may be websites that the overall website administrator does not have full control over, such as websites set up by each business division for each product or service, or old websites owned by acquired companies. In addition, there may be websites that are easily overlooked, such as websites that are left unmanaged due to the previous administrator's transfer or retirement, or websites that are dedicated to limited-time campaigns and remain active even after the campaign ends.

[0005] Furthermore, these websites often lack the latest and most appropriate security measures and are therefore vulnerable, making them at high risk of being used as a gateway for unauthorized intrusion by third parties.

[0006] However, in order to identify these websites that should be managed, a company must take an inventory of the websites it manages. To do this, it first searches the countless websites on the Internet for those that may be relevant to the company. For example, if a mechanical search is performed using a known method, a huge number of candidate websites, including unrelated websites, may be identified, depending on the size of the websites. Then, the company must determine whether or not each of the identified candidate websites should be managed by the company. While some easy-to-identify websites can be identified mechanically, currently many decisions must be made manually, making it virtually impossible to determine if there are a large number of candidate websites.

[0007] In this regard, for example, Japanese Patent No. 6785087 (Patent Document 1) describes a website inventory system that, in the inventory process of identifying and extracting websites that a target company or other entity should manage from the many websites publicly available on the Internet, uses machine learning to mechanically make the judgment of whether or not a website should be managed, which is difficult to do manually, thereby significantly reducing the workload and processing time and improving accuracy. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Patent No. 6785087 Summary of the Invention [Problem to be solved by the invention]

[0009] According to the conventional technology described in Patent Document 1, it is possible to efficiently inventory websites (hidden IT assets) that should be managed by the target company or the like but are not yet identified.

[0010] However, the results of the inventory pose a challenge in that it is difficult to provide a logical answer to the target company, etc., regarding the basis for determining why a website should (or does not need to) be subject to management, since this is a black box judgment based on a judgment model generated by machine learning.

[0011] Therefore, the object of the present invention is to provide an IT asset inventory system that can inventory IT assets such as websites to be managed and provide a response in natural language explaining the basis for the judgment on the results.

[0012] The above and other objects and novel features of the present invention will become apparent from the description of this specification and the accompanying drawings. [Means for solving the problem]

[0013] Among the inventions disclosed in this application, the outline of representative inventions will be briefly explained as follows.

[0014] An IT asset inventory system that is a representative embodiment of the present invention is an IT asset inventory system that identifies IT assets that should be managed in a management organization, and has an IT asset search unit that collects information on IT assets that are candidates for management based on information specified by a user as information related to the management organization, and a judgment unit that, for each IT asset collected by the IT asset search unit, inputs a prompt to a generation AI that includes a command statement that specifies the content of the target IT asset and information related to the management organization, as well as whether the target IT asset is related to the management organization and the basis for that determination, and obtains an answer. [Effects of the Invention]

[0015] The effects obtained by the representative inventions disclosed in this application can be briefly explained as follows.

[0016] That is, according to the representative embodiment of the present invention, it is possible to take an inventory of IT assets such as websites to be managed, and to provide a response in natural language on the basis of the results of the inventory. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a diagram showing an overview of an example of the configuration of an IT asset inventory system according to an embodiment of the present invention. [Figure 2] FIG. 10 is a diagram outlining an example of a primary determination by a generation AI in one embodiment of the present invention as to whether a target website is related to a target company. DETAILED DESCRIPTION OF THE INVENTION

[0018] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In all drawings used to explain the embodiments, the same parts are generally designated by the same reference numerals, and repeated explanations will be omitted. However, parts that have been designated and explained in one drawing may be referred to by the same reference numerals in the explanation of other drawings, although they will not be shown again.

[0019] <Summary> As mentioned above, in order for a management organization such as a company to identify the IT assets such as websites that should be managed, it is necessary to take an inventory of the IT assets related to the company.To do this, it is necessary to first search the countless IT assets such as websites on the Internet for those that may be related to the company, and then make a judgment on whether or not the huge number of candidates should be managed, which is difficult to do manually.

[0020] The IT asset inventory system, which is one embodiment of the present invention, uses a trained generative AI such as GPT (Generative Pretrained Transformer, registered trademark) to make a primary judgment on whether or not an asset should be included in the management list. This significantly reduces the burden of manually making the final secondary judgment, and enables logical answers to be given in natural language on the basis of the judgment.

[0021] In this embodiment, we will focus on websites (web servers) as IT assets to be managed by management organizations such as companies (i.e., the contents of the IT assets become the response content of the web server), but IT assets can also broadly include various servers that can be reached from the Internet, such as database servers and mail servers (i.e., the contents of these IT assets become the response data of each server).

[0022] <System configuration> 1 is a diagram showing an overview of an example of the configuration of an IT asset inventory system according to one embodiment of the present invention. The IT asset inventory system 1 is configured, for example, with server equipment or a virtual server built on a cloud computing service, and uses a central processing unit (CPU) (not shown) to execute middleware such as an operating system (OS), a database management system (DBMS), and a web server program, which are deployed to memory from a storage device such as a hard disk drive (HDD) or solid state drive (SSD), as well as software running on those systems, thereby realizing the function of searching the Internet 2 and taking an inventory of IT assets such as websites that a target company should manage.

[0023] The IT asset inventory system 1 has various units implemented as software, such as an inventory management unit 11, an IT asset search unit 12, a determination unit 13, and a monitoring and verification unit 14. It also has various data stores implemented as databases, file tables, etc., such as target company information 15, target host information 16, asset candidate 17, and asset list 19.

[0024] The inventory management unit 11 has a user interface function such as accepting input of inventory-related settings and instructions from the user via a user terminal 4, which is an information processing terminal such as a PC (Personal Computer), tablet terminal, or smartphone connected via a network (not shown), and presenting the inventory results to the user, as well as a function related to controlling the entire inventory process. It may also be configured to accept input of the contents of a secondary judgment by the user regarding the inventory results primarily judged by the generation AI.

[0025] The IT asset search unit 12 has a function of searching the Internet 2 for websites that may be related to the target company, etc., according to predetermined criteria and conditions, etc., based on target company information 15 input by the user, and outputting a list of websites resulting from the search as asset candidates 17. The target company information 15 may include information related to the target company, etc., such as the company name, industry, company logo image, corporate color, owned brands, the URL (Uniform Resource Locator) and domain name of the main website, the organization name, etc.

[0026] Further useful information may be target host information 16 that is set in advance or obtained by the IT asset search unit 12. The target host information 16 may include, for example, Whois information (domain name registrant information) obtained using an external search service or the like for the actual domain name of the target site.

[0027] Methods for searching the Internet 2 include, for example, a method of obtaining information by following links from a main website such as a corporate site, a method of obtaining information by keyword search or image search based on the contents of the target company information 15 and target host information 16, a method of changing the host name or domain name in the FQDN (Fully Qualified Domain Name) of the corporate site according to a pattern (for example, a list of host name configuration patterns that are generally expected to be frequently used in companies, etc. is set in the IT asset search unit 12), and copyright investigation, and various other methods known as OSINT (Open Source Intelligence) can be appropriately combined.

[0028] The judgment unit 13 has the function of creating a prompt from a prompt template 18 for each website of the candidate assets 17 searched by the IT asset search unit 12 based on the contents of the target company information 15, target host information 16, etc., and inputting this into the generation AI 3 to obtain the results of a primary judgment as to whether the target website is related to the target company, etc. (whether it should be managed), and outputting the websites that are determined to be related as an asset list 19.

[0029] The asset list 19 may be presented to the user terminal 4 via the inventory management unit 11, and the user may ultimately make a secondary judgment as to whether each website is truly related to the target company, etc. By performing a primary judgment using the generation AI 3 in the judgment unit 13, the number of websites that are subject to manual secondary judgment by the user can be significantly reduced, thereby significantly reducing the workload and processing time, and improving the accuracy of the judgment. The contents of the judgment (primary judgment) processing by the judgment unit 13 will also be described later.

[0030] The monitoring and verification unit 14 has a function of monitoring whether the result of the primary determination by the determination unit 13 using the generated AI 3 is significantly incorrect, and also has a function of verifying the performance of the prompts input to the generated AI 3.

[0031] In the above monitoring function, since it is not practical to individually monitor the accuracy of all the judgment results by the generation AI 3, it is also possible to monitor the validity of the judgment results in general. For example, to monitor changes in a predetermined statistical value related to the judgment, it is possible to monitor whether the statistical value (e.g., the number of assets in the asset list 19 / the number of assets in the asset candidate 17) when a website inventory was conducted for the target company in the past (e.g., the previous time) and the statistical value when an inventory was conducted for the target company this time differ by more than a predetermined value.

[0032] Furthermore, the above-mentioned verification function, for example, samples a small number of websites from the asset candidates 17, creates actual prompts from the prompt template 18, and inputs them into the generation AI 3, and compares the obtained output with the correct answer input by the user. This allows for prior verification of the performance of the prompt, such as the comprehensiveness of the information related to the management organization used to create the prompt and the validity of the instructions to the generation AI 3 in the prompt template 18. These monitoring and verifications enable the quality of the inventory to be maintained. Note that the processing by the monitoring and verification unit 14 may be configured to be performed asynchronously with the inventory processing based on instructions from the user, or predetermined monitoring processing may be configured to be performed automatically in conjunction with the processing by the determination unit 13.

[0033] 1, the IT asset inventory system 1 is configured as a single server system, but this is merely a logical configuration, and one or more of the above-mentioned components may be physically configured as separate server systems. Also, in the example of Fig. 1, the IT asset inventory system 1 is configured to be accessed from a user terminal 4 to use the service, but the IT asset inventory system 1 may also be configured to be built on the user terminal 4.

[0034] <Primary judgment by generative AI> 2 is a diagram outlining an example of a primary determination by generation AI 3 in one embodiment of the present invention as to whether a target website is related to a target company. In this embodiment, the target website is first accessed to acquire content such as HTML (HyperText Markup Language), and determination unit 13 then creates a prompt by inputting the acquired content, the contents of target host information 16 (such as domain registration information listed in Whois information), and the contents of target company information 15 (such as the name, URL, and organization name of the target company) into prompt template 18.

[0035] The command portion of the prompt includes a command statement in natural language that indicates whether the target website is related to the target company and the basis for that determination. The command statement specifies how to determine whether the target website is related to the target company. For example, the command statement may determine whether the target website is related to the target company based on whether the domain registrant information or web content of the target website contains keywords or synonyms of the company name, organization name, brand name, etc. obtained from the target company information 15. Alternatively, the command statement may include a numerical value (credit score) as a response, or, if the determination is based on image data such as the target company's logo or favicon, a command statement may include highlighting the relevant image. The content of the command statement can be edited as inventory-related settings by the user via the user terminal 4.

[0036] By inputting such a prompt into the generation AI3, an answer is obtained from the generation AI3. In the example shown in the figure, it is determined that the target website is relevant to the target company (the generation AI response is "Customer Relevance: Yes"), and the answer given is a natural language sentence that states, "Customer-related keywords exist in the content of the target website..." as the basis for this. Note that it is also possible to use multiple types of generation AI3 and input the same prompt, and evaluate and judge trustworthiness based on the commonality of the answers obtained from each.

[0037] This process is repeated recursively for multiple websites to be assessed, and the FQDNs of websites determined to be relevant to the target company are listed, resulting in an asset list 19 as the result of the first assessment by the generation AI 3. The list may be sorted and output by the credit score, etc., returned by the generation AI 3. This asset list 19 is a list of candidate websites that the target company should manage, such as by collecting security information and implementing security measures. This list may be used as the managed websites directly, or the user may refer to the content to make a final second assessment of whether the websites are truly relevant to the target company. In this case, the user can refer to the natural language assessment basis generated by the generation AI 3, thereby reducing the assessment burden and improving the accuracy of the assessment.

[0038] Furthermore, the user may analyze the trends of the manual secondary judgment results (the results of the primary judgment by the generation AI 3) based on the content of the erroneous judgment, and modify the prompt template 18 (or the prompt), for example, by describing conditions that are likely to result in an erroneous judgment, and adding instructions to the generation AI 3 to prevent an erroneous judgment. This can further improve the accuracy of the judgment.

[0039] The asset list 19, which is a list (or candidates) of websites that the target company should manage, can be continuously inventoried and updated to the latest status. In this case, for example, of the asset candidates 17 found as a result of the search by the IT asset search unit 12, only the asset candidates 17 that have been added since the previous inventory may be subject to judgment by the judgment unit 13. At this time, the user may be notified of the added asset candidates 17, and the asset candidates 17 specified by the user may be subject to judgment. Alternatively, the results of the previous inventory may be reset, and all asset candidates 17 may be subject to judgment again. Furthermore, judgment may be performed again only when there is a change in the prompts entered into the generation AI 3 due to changes in corporate information, such as due to M&A at the target company.

[0040] As described above, according to the IT asset inventory system 1, which is one embodiment of the present invention, in an inventory to identify websites that a company should manage, a primary determination of whether a huge number of candidate websites extracted from the Internet 2 should be managed is made using the generation AI 3. This significantly reduces the burden on the user who would otherwise manually perform the final secondary determination, and can improve the accuracy of the determination. Furthermore, by providing a logical response in natural language explaining the basis for the determination, it is possible to promote user understanding.

[0041] In addition, when creating keywords to be used when searching the Internet2, by collecting information related to the target company, such as URLs, domain names, organization names, and brand names, from multiple data sources, it is possible to comprehensively extract keywords that are closely related to the target company.By combining this with the use of generation AI3 to extract high-quality keywords, it is possible to prevent false negatives due to a lack of keywords (overlooking websites that should be managed).

[0042] The invention made by the inventor has been specifically described above based on the embodiments, but it goes without saying that the present invention is not limited to the above embodiments and can be modified in various ways without departing from the spirit of the invention. Furthermore, the above embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those having all of the described configurations. Furthermore, it is possible to add, delete, or replace part of the configuration of the above embodiments with other configurations.

[0043] Furthermore, the above-described configurations, functions, processing units, processing means, etc. may be partially or entirely implemented in hardware, for example, by designing them as integrated circuits. The above-described configurations, functions, etc. may also be implemented in software, with a processor interpreting and executing a program that implements each function. Information such as the programs, tables, and files that implement each function can be stored in a storage device such as a memory, hard disk, or SSD, or in a storage medium such as an IC card, SD card, or DVD.

[0044] In addition, in the above figures, the control lines and information lines shown are those that are considered necessary for explanation, and do not necessarily show all the control lines and information lines that are actually implemented. In reality, it can be assumed that almost all components are interconnected. [Industrial Applicability]

[0045] The present invention can be used in an IT asset inventory system that keeps track of IT assets such as websites managed by a company. [Explanation of symbols]

[0046] 1...IT asset inventory system, 2...Internet, 3...generative AI, 4...user terminal, 11...Inventory management section, 12...IT asset search section, 13...Determination section, 14...Monitoring and verification section, 15...Target company information, 16...Target host information, 17...Asset candidate, 18...Prompt template, 19...Asset list

Claims

1. An IT asset inventory system for identifying IT assets to be managed in a management organization, comprising: an IT asset search unit that collects information on IT assets that are servers reachable from the Internet and that are candidates for the management target based on the content designated by a user as information related to the management organization; An IT asset inventory system having a determination unit that creates, for each IT asset collected by the IT asset search unit, a prompt including the server's response data, which is the content of the target IT asset, and information related to the management organization, as well as a command statement for determining whether the target IT asset is related to the management organization, and inputs the created prompt into a generation AI to obtain an answer.

2. 2. The IT asset inventory system according to claim 1, further comprising: An IT asset inventory system having a verification unit that compares, for a portion of each IT asset collected by the IT asset search unit, the judgment result by the generation AI in the judgment unit for the target IT asset with correct information regarding whether the IT asset specified by the user is related to the management organization.

3. 2. The IT asset inventory system according to claim 1, The judgment unit inputs the same prompt to multiple generated AIs and evaluates the commonality of the content of the answers obtained from each generated AI.

4. 2. The IT asset inventory system according to claim 1, further comprising: An IT asset inventory system having a monitoring unit that determines whether the ratio of the number of IT assets determined to be related to the management organization in the determination results by the generation AI in the determination unit to the number of each IT asset collected by the IT asset search unit deviates by more than a predetermined value from the same ratio when IT assets that should be managed in the management organization were identified in the past.

Images