COLLECTION DEVICE, COLLECTION METHOD, AND COLLECTION PROGRAM

The collection device efficiently detects malicious sites by analyzing user-generated content across services, using machine learning to identify patterns and generate search queries, thereby enhancing detection accuracy and speed.

JP7589745B2Active Publication Date: 2024-11-26NIPPON TELEGRAPH & TELEPHONE CORP
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2022556745
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2020-10-14
Publication Date
2024-11-26
Estimated Expiration
2040-10-14

AI Technical Summary

Technical Problem

Conventional techniques for detecting malicious sites on the Web are insufficient in terms of detection accuracy, speed, and range, particularly due to the need for accessing malicious sites and slow detection speeds.

Method used

A collection device that acquires user-generated content, generates search queries using key phrases, and collects content across multiple services to identify potentially malicious content, utilizing machine learning to determine malignancy based on feature differences between legitimate and attacker-generated content.

Benefits of technology

Enables rapid and accurate detection of malicious sites over a wide range by leveraging machine learning to identify patterns in user-generated content, allowing real-time threat detection and reporting.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007589745000002
    Figure 0007589745000002
  • Figure 0007589745000003
    Figure 0007589745000003
  • Figure 0007589745000004
    Figure 0007589745000004
Patent Text Reader

Abstract

An acquisition unit (15a) acquires user-generated content generated in each service in a predetermined period. A generation unit (15b) generates a search query by using a word that appears in user-generated content for each service. A collection unit (15c) collects user-generated content generated in a plurality of the services by using the generated search query.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a collection device, a collection method, and a collection program. [Background technology]

[0002] Social engineering (SE) attacks that exploit vulnerabilities in user psychology are becoming mainstream as a threat on the Web. Furthermore, user-generated content such as videos, blogs, and bulletin board posts that are generated by attackers using online services and posted on the Web are increasing as a route to redirect users to malicious websites.

[0003] On the other hand, user-generated content generated by attackers is generated in large quantities in real time, targeting specific events such as concerts and sports, and is spread across many services disguised as legitimate users. Therefore, rapid, accurate and wide-area detection technology is required.

[0004] For example, conventionally, a search engine is used to detect malicious sites and recursively create queries for searching for the malicious sites (see Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] Luca Invernizzi, Paolo Milani Comparetti, “EVILSEED: A Guided Approach to Finding Malicious Web Pages”, [online], [Retrieved July 27, 2020], Internet<URL:https: / / sites.cs.ucsb.edu / ~vigna / publications / 2012_SP_Evilseed.pdf> Summary of the Invention [Problem to be solved by the invention]

[0006] However, the conventional techniques are insufficient in terms of detection accuracy, detection speed, and detection range. For example, the technique described in Non-Patent Document 1 requires access to a malicious site, and has a problem of slow detection speed.

[0007] The present invention has been made in view of the above, and has an object to quickly and accurately detect malicious sites over a wide area. [Means for solving the problem]

[0008] In order to solve the above-mentioned problems and achieve the objective, the collection device of the present invention is characterized by having an acquisition unit that acquires user-generated content generated in each service over a specified period of time, a generation unit that generates a search query using words that appear in the user-generated content for each service, and a collection unit that collects user-generated content generated in multiple services using the generated search query. Effect of the Invention

[0009] According to the present invention, it is possible to quickly and accurately detect malicious sites over a wide area. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram for explaining an overview of the detection device of the present embodiment. [Diagram 2] FIG. 2 is a schematic diagram illustrating a schematic configuration of the detection device of the present embodiment. [Diagram 3] FIG. 3 is a diagram for explaining the processing of the collection function unit. [Figure 4] FIG. 4 is a diagram for explaining the process of the generation unit. [Diagram 5] FIG. 5 is a diagram for explaining the process of the determination function unit. [Figure 6] FIG. 6 is a diagram for explaining the process of the calculation unit. [Figure 7]FIG. 7 is a diagram for explaining the process of the calculation unit. [Figure 8] FIG. 8 is a diagram for explaining the process of the calculation unit. [Figure 9] FIG. 9 is a diagram for explaining the process of the calculation unit. [Figure 10] FIG. 10 is a diagram for explaining the process of the extraction function unit. [Figure 11] FIG. 11 is a diagram for explaining threat information. [Figure 12] FIG. 12 is a diagram for explaining threat information. [Figure 13] FIG. 13 is a flowchart showing a processing procedure of the collection function unit. [Figure 14] FIG. 14 is a flowchart showing a processing procedure of the determination function unit. [Figure 15] FIG. 15 is a flowchart showing a processing procedure of the determination function unit. [Figure 16] FIG. 16 is a flowchart illustrating a processing procedure of the extraction function unit. [Figure 17] FIG. 17 is a flowchart illustrating a processing procedure of the extraction function unit. [Figure 18] FIG. 18 is a diagram illustrating an example of a computer that executes a detection program. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to this embodiment. In addition, in the description of the drawings, the same parts are indicated by the same reference numerals.

[0012] [Detection device overview] 1 is a diagram for explaining an overview of a detection device. The detection device 1 of this embodiment collects and analyzes user-generated content, such as videos, blogs, and bulletin board posts, that are generated and posted on the Web by users in online services such as Facebook (registered trademark) and Twitter (registered trademark).

[0013] Specifically, we focus on the fact that attackers generate and spread large amounts of user-generated content in a concentrated manner in response to events that attract users' attention, and that they generate user-generated content in similar contexts that make users want to access malicious sites.

[0014] Therefore, the detection device 1 efficiently collects user-generated content that is likely to be malicious by an attacker, using the characteristic that user-generated content by an attacker is spread in a similar context at a specific timing, and analyzes whether it is malicious or not. Furthermore, when the detection device 1 determines that the user-generated content is malicious as a result of the analysis, it extracts threat information, which is a characteristic that may be a threat, from the malicious user-generated content, and outputs a threat report.

[0015] For example, the detection device 1 extracts similar contexts of user-generated content to generate search queries, and efficiently collects user-generated content that is likely to be malicious using the search queries. In addition, the detection device 1 specializes in a specific service, and by learning the characteristic differences between user-generated content generated by attackers and user-generated content generated by legitimate users, determines whether a large amount of user-generated content generated at the same time for a specific service is malicious.

[0016] Furthermore, in a given service, detection device 1 learns the characteristic differences between user-generated content generated by an attacker and user-generated content generated by a legitimate user, and the Web content obtained by accessing a URL described in the user-generated content. Then, detection device 1 uses the learned characteristic differences to determine whether user-generated content generated in large quantities in the same period in a given service is malicious.

[0017] Furthermore, when the detection device 1 determines that the user-generated content is malicious, the detection device 1 extracts threat information, which is a characteristic that may be a threat, from the malicious user-generated content and outputs a threat report. In this way, the detection device 1 detects attacks that may be a threat in real time.

[0018] [Detection device configuration] Fig. 2 is a schematic diagram illustrating a schematic configuration of the detection device of this embodiment. As illustrated in Fig. 2, the detection device 1 of this embodiment is configured to include a collection function unit 15A, a determination function unit 15B, and an extraction function unit 15C. These function units may be implemented in hardware different from the detection device 1. In other words, the detection device 1 may be implemented as a detection system having a collection device, a determination device, and an extraction device.

[0019] The detection device 1 is realized by a general-purpose computer such as a personal computer, and includes an input unit 11, an output unit 12, a communication control unit 13, a storage unit 14, and a control unit 15.

[0020] The input unit 11 is realized by using input devices such as a keyboard and a mouse, and inputs various instruction information such as a command to start processing to the control unit 15 in response to an input operation by an operator. The output unit 12 is realized by a display device such as a liquid crystal display, a printing device such as a printer, etc. For example, the output unit 12 displays the results of a detection process described below.

[0021] The communication control unit 13 is realized by a NIC (Network Interface Card) or the like, and controls communication between the control unit 15 and an external device via a telecommunication line such as a LAN (Local Area Network) or the Internet. For example, the communication control unit 13 controls communication between the control unit 15 and a server or the like that manages user-generated content of each service.

[0022] The storage unit 14 is realized by a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 14 stores in advance a processing program for operating the detection device 1 and data used during execution of the processing program, or stores the data temporarily each time processing is performed. The storage unit 14 may be configured to communicate with the control unit 15 via the communication control unit 13.

[0023] In this embodiment, storage unit 14 stores threat information and the like obtained as a result of a detection process described later. Furthermore, storage unit 14 may store user-generated content acquired by acquisition unit 15a described later from a server of each service or the like prior to the detection process.

[0024] Returning to the explanation of Fig. 2, the control unit 15 is realized using a CPU (Central Processing Unit) or the like, and executes a processing program stored in a memory. As a result, the control unit 15 functions as a collection function unit 15A, a determination function unit 15B, and an extraction function unit 15C, as exemplified in Fig. 2.

[0025] The collection function unit 15A includes an acquisition unit 15a, a generation unit 15b, and a collection unit 15c. The determination function unit 15B includes a calculation unit 15d, a learning unit 15e, and a determination unit 15f. The extraction function unit 15C includes an extraction unit 15g, a learning unit 15e, and a determination unit 15f.

[0026] Each or some of these functional units may be implemented in different hardware. For example, as described above, the collection function unit 15A, the determination function unit 15B, and the extraction function unit 15C may be implemented in different hardware as a collection device, a determination device, and an extraction device, respectively. The control unit 15 may also include other functional units.

[0027] [Collection Function Division] 3 is a diagram for explaining the processing of the collection function unit. As shown in FIG. 3, collection function unit 15A extracts similar contexts as key phrases from a group of user-generated contents generated at the same time in a certain service, and generates a search query. In addition, collection function unit 15A efficiently collects user-generated contents from any service that are likely to be malicious, using the generated search query of the key phrase that is likely to be malicious.

[0028] Returning to the explanation of Fig. 2, acquisition unit 15a acquires user-generated content generated in each service during a predetermined period. Specifically, acquisition unit 15a acquires the user-generated content from the server of each service via input unit 11 or communication control unit 13.

[0029] For example, acquisition unit 15a acquires user-generated content that includes a URL for a predetermined service. In this case, acquisition unit 15a may acquire user-generated content regularly at predetermined times, or by specifying the time of posting using "since" or "until." Furthermore, acquisition unit 15a may acquire only user-generated content that includes a URL by using "filters." This allows acquisition unit 15a to acquire user-generated content that includes the URL of an external site in real time.

[0030] Acquiring unit 15a may store the acquired user-generated content in storage unit 14, for example, prior to processing by generating unit 15b, which will be described later.

[0031] The generator 15b generates a search query using words that appear in the user-generated content for each service. For example, the generator 15b generates a search query using a combination of the appearing words.

[0032] Specifically, generation unit 15b converts the acquired user-generated content into a feature vector with a predetermined number of dimensions. For example, generation unit 15b sets a vector of embedded representations of words that represent a combination of words that appear in each user-generated content in a vector space that represents the vocabulary that appears in the user-generated content, i.e., the totality of words that appear, as the feature vector of the user-generated content. Then, generation unit 15b learns a model of embedded representations of words in advance and applies a text summarization technique. In other words, the text summarization technique is used to extract, as a key phrase, a word combination of embedded representations that is similar to the embedded representations of the entire target text (text).

[0033] As a result, generating unit 15b extracts key phrases that represent the context of each piece of user-generated content, and generates a search query for searching for user-generated content that includes the extracted key phrases.

[0034] Specifically, generation unit 15b calculates the similarity between the entire text of the user-generated content and the key phrase candidates according to the following formula (1): where doc is the entire target sentence, C is the key phrase candidate, and K is a set of extracted word combinations (phrases).

[0035]

number

[0036] In the above formula (1), by changing λ, it is possible to extract a variety of key phrases.

[0037] For example, generating unit 15b extracts a combination of words using an n-gram technique that extracts n consecutive words from a text. Then, generating unit 15b calculates the cosine similarity between the entire text of the user-generated content and each of the extracted n-gram phrases using formula (1) above, and extracts the largest of the phrases whose calculated similarity value is higher than a predetermined threshold as a key phrase.

[0038] Here, Fig. 4 is a diagram for explaining the processing of the generating unit 15b. In the example shown in Fig. 4, the generating unit 15b extracts a combination of words by 3-grams. In addition, the generating unit 15b extracts key phrases by calculating the cosine similarity between the entire text of the user-generated content "Japan vs United States Free live streaming click here" and each of the 3-gram phrases "japan vs united", "vs united states", "united states free", ....

[0039] Alternatively, the generation unit 15b generates a search query using the frequency of occurrence of each word. For example, the generation unit 15b counts the frequency of occurrence of 2-gram phrases and 3-gram phrases in text of user-generated content acquired during a predetermined period. Then, the generation unit 15b extracts a phrase whose frequency of occurrence is equal to or greater than a predetermined threshold as a key phrase, and generates a search query for searching for user-generated content containing the key phrase.

[0040] For example, generation unit 15b extracts 3-gram phrases from the text of all user-generated content posted every hour for 24 hours on March 1, and calculates the frequency of appearance of each phrase. Next, generation unit 15b extracts, as a key phrase, one that is a statistically abnormal value (outlier) from among the 3-gram phrases that appeared in user-generated content for one hour from 0:00 to 1:00 on the following day, March 2. In other words, when a large amount of user-generated content including a phrase that does not usually appear is posted at a specific timing, generation unit 15b sets this phrase as a key phrase.

[0041] For example, the generation unit 15b calculates the positive outliers using the z-score. In the example shown in Fig. 4, for the phrase "japan vs united," the occurrence counts for each hour during the 24 hours on March 1 are 0, 0, 0, 2, 4, 10, 2, 5, 10, 2, 4, 5, 6, 2, 2, 5, 12, 20, 15, 20, 10, 20, 25, and 30, respectively. In this case, the average is 8.792 times, and the standard deviation is 8.602.

[0042] Also, assume that this phrase appeared 50 times in the one hour period from 0:00 to 1:00 on March 2. In this case, the z-score is calculated as Z=(50-8.792) / 8.602=4.790. If the outlier threshold is 1.96, which corresponds to a significant appearance frequency of 5%, generation unit 15b uses the phrase "japan vs united" as a key phrase and generates a search query to search for user-generated content containing this key phrase.

[0043] Furthermore, the generating unit 15b selects a potentially malicious search query for each service. For example, the generating unit 15b calculates the maliciousness of the generated search query based on a search query used to search for user-generated content that was determined to be malicious most recently for each service. The generating unit 15b then selects a search query whose maliciousness is equal to or exceeds a predetermined threshold as a search query for the service.

[0044] Here, the generation unit 15b calculates the ratio of the number of user-generated contents determined to be malicious, using the number of user-generated contents that were searched for using the search query in the past 24 hours and determined to be malicious or benign, as the maliciousness of the search query. In addition, the generation unit 15b calculates the average value of the maliciousness of each word in the key phrase, regarding it as the maliciousness of the detection query.

[0045] For example, suppose that in the past 24 hours, on a certain service, the number of malicious user-generated contents found by the search query for the key phrase "rugby world cup streaming" was 20, and the number of benign user-generated contents found by the search query for the key phrase "free live streaming" was 100, and the number of benign user-generated contents found by the search query for the key phrase "rugby japan vs korea" was 10, and the number of malicious user-generated contents found by the search query for the key phrase "rugby japan vs korea" was 10, and the number of benign user-generated contents found by the search query for the key phrase "rugby japan vs korea" was 10.

[0046] In this case, the malignancy of the word "japan" is α=10 / (10+100), the malignancy of the word "rugby" is β={20 / (20+50)+10 / (10+100)} / 2, and the malignancy of the word "streaming" is γ={20 / (20+50)+100 / (100+100)} / 2.

[0047] Therefore, the maliciousness score of the search query for the key phrase “japan rugby streaming” is calculated as (α+β+γ) / 3=0.225.

[0048] In this manner, the generating unit 15b calculates the maliciousness of the search query for each service, and selects a search query whose calculated maliciousness is equal to or greater than a threshold as a search query for user-generated content that may be malicious for that service.

[0049] The collection unit 15c uses the generated search query to collect user-generated content generated in multiple services. For example, the collection unit 15c uses a search query generated in user-generated content of a certain service to collect user-generated content of another service. In addition, the collection unit 15c also uses the same search query to collect multiple types of user-generated content in each service together with the date and time of generation.

[0050] For example, for service a where user-generated content such as text posting, video posting, and event announcement is generated, collection unit 15c applies the same search query to three types of collection URLs to collect the three types of user-generated content together with the posted (generated) date and time. Also, for service b where user-generated content such as video posting and video distribution is generated, collection unit 15c applies the same search query to a common collection URL to collect two types of user-generated content together with the posted date and time.

[0051] This allows collection unit 15c to efficiently collect user-generated content that is spread in a context similar to a specific timing. In particular, collection unit 15c can easily and quickly collect user-generated content that is likely to be malicious for each service by using the potentially malicious search queries selected by generation unit 15b.

[0052] Note that collection unit 15c collects user-generated content with an upper limit on the amount of collection, such as 100 queries per hour, which can reduce the load on the server of each service that is the collection destination.

[0053] [Judgment function section] Fig. 5 is a diagram for explaining the processing of the determination function unit. As shown in Fig. 5, the determination function unit 15B acquires a machine learning model representing each feature amount by learning using the difference in features between user-generated content generated by an attacker and user-generated content generated by a legitimate user for a specific service. The determination function unit 15B learns the machine learning model using, as the feature amount, a text feature amount representing the co-occurrence of phrases in the user-generated content and a group feature amount representing the similarity of words appearing in each user-generated content.

[0054] This allows the determination function unit 15B to use the learned machine learning model to determine whether user-generated content of the service subsequently generated is malicious or not. For example, the determination function unit 15B can perform in real time the maliciousness determination of a large amount of user-generated content of a specific service generated at the same time.

[0055] Returning to the description of Fig. 2, calculation unit 15d calculates features of user-generated content generated by a user in a predetermined period of time in a predetermined service. In this embodiment, the features of the user-generated content are text features representing features of combinations of words that co-occur in a plurality of pieces of user-generated content, and group features representing features related to similarities of words among a plurality of pieces of user-generated content generated in a predetermined period of time.

[0056] 6 to 9 are diagrams for explaining the processing of the calculation unit. First, calculation unit 15d calculates text feature quantities representing features of combinations of words co-occurring in a plurality of pieces of user-generated content. Specifically, calculation unit 15d calculates text feature quantities of a set of user-generated content using an optimized word distributed representation model for each phrase co-occurring in a set of collected user-generated content.

[0057] More specifically, as shown in Fig. 6, calculation unit 15d optimizes a model that outputs a feature vector of a distributed representation of phrases co-occurring in each user-generated content in a set of user-generated contents in advance. In the example shown in Fig. 6, calculation unit 15d uses, as input weights, a matrix (see 1.) in which each row represents a word (1-gram phrase) and 2-gram phrase that appear in a set of malicious user-generated contents, and each column represents each user-generated content (document). In addition, calculation unit 15d calculates the average of each row corresponding to each phrase (see 2.).

[0058] In addition, the calculation unit 15d calculates the inner product (see 3.) using a matrix with each document as a row and each word as a column as output weights, and optimizes the model that outputs the feature vector of the distributed representation of each phrase (see 4.).

[0059] Then, as shown in FIG. 7, calculation unit 15d first extracts words present in a dictionary from the character strings of URLs in the collected user-generated contents U, and replaces the words with the character strings of the URLs (Word Segmentation).

[0060] Calculation unit 15d also optimizes the model of the distributed representation for words (1-gram phrases) and 2-gram phrases that appear in the set U of user-generated content, as shown in Fig. 6. Then, calculation unit 15d uses the optimized model of the distributed representation to calculate the feature vector VEC u Then, calculation unit 15d generates a feature vector VEC of each user-generated content u as a text feature of the set of user-generated content u. u Calculate the average.

[0061] Here, malicious user-generated content tends to have many similar words even in events with different timing. Therefore, for a set U of malicious user-generated content, the feature vector VECu The average of can be a feature that reflects the characteristics of the set U of user-generated content.

[0062] Furthermore, calculation unit 15d calculates a group feature that indicates a feature related to the similarity of words between a plurality of user-generated contents generated during a predetermined period. Specifically, as shown in FIG. 8, calculation unit 15d applies the Minhash-LSH algorithm to words (1-gram phrases) that appear in a set U of user-generated contents collected during the same period, and calculates the similarity between each of the user-generated contents. Here, "the same period" means that the time difference between the dates and times of generation is within a predetermined time threshold σ. Then, when the calculated similarity exceeds a predetermined similarity threshold τ, calculation unit 15d regards this set of user-generated contents as a set of similar user-generated contents.

[0063] Calculation unit 15d identifies a group feature for a set of similar user-generated content, the group feature being the size of the set, the number of users in the set, the number of unique URLs listed in the set, the average number of URLs listed in the user-generated content in the set, or the average posting time interval in the set.

[0064] For example, as illustrated in FIG. 9, calculation unit 15d determines whether each collected user-generated content set is a similar user-generated content set, and if it is a similar user-generated content set, identifies a group feature amount.

[0065] 9 illustrates, for example, that user-generated content 1 was generated by user1, and that the words that appear are "Free live streaming URL1 URL1." It also illustrates that user-generated contents 1 to 3 are a set of the same similar user-generated content. It also illustrates that the group features of this set of similar user-generated content are the average posting time interval, the set size is 3, the number of unique users in the set is 2 (user1, user2), the number of unique URLs in the set is 2 (URL1, URL2), and the average number of URLs per content is 1.67.

[0066] Also, it is illustrated that user-generated contents 4 and 5 are the same similar user-generated content set, and it is illustrated that user-generated contents 6 and 7 are not the similar user-generated content set.

[0067] Malicious user-generated contents tend to be distributed in similar contexts at the same time. Therefore, it is possible to identify group features for a collection of malicious user-generated contents as described above. In other words, if group features can be identified in this way, it means that there is a high possibility that this collection of user-generated contents is malicious.

[0068] Returning to the description of Fig. 2, learning unit 15e learns using the calculated feature amounts of user-generated content generated by legitimate users and feature amounts of content generated by malicious users. Furthermore, determination unit 15f determines whether or not the user-generated content has been generated by a malicious user, based on the learned model.

[0069] Specifically, learning unit 15e performs supervised learning of a machine learning model using text features that represent the co-occurrence of phrases in user-generated content and group features that represent the similarity of words that appear in each user-generated content. Furthermore, determination unit 15f uses the learned machine learning model to determine whether user-generated content of the service acquired thereafter is malicious or not.

[0070] In this way, judgment function unit 15B learns the characteristics of user-generated content that is likely to be malicious and is generated at specific times such as events, and is able to use the learning results to make a judgment on the maliciousness of user-generated content collected in real time.

[0071] [Extraction function section] Fig. 10 is a diagram for explaining the processing of the extraction function unit. As shown in Fig. 10, the extraction function unit 15C extracts the feature amount of the Web content obtained by accessing the URL included in the user-generated content in any service. For example, the extraction function unit 15C identifies the IP address of the FQDN (fully qualified domain name) finally reached.

[0072] In addition, extraction function unit 15C learns about user-generated content generated by an attacker and user-generated content generated by a legitimate user by using the feature amount. Extraction function unit 15C then uses the learned feature amount to determine whether user-generated content generated in large quantities in any service during the same period is malicious.

[0073] In addition, when the extraction function unit 15C determines that the user-generated content is malicious, the extraction function unit 15C extracts threat information, which is a characteristic that may be a threat, from the malicious user-generated content and outputs a threat report. In this way, the extraction function unit 15C can detect attacks that may be a threat in real time.

[0074] Returning to the description of Fig. 2, extraction unit 15g accesses entrance URLs described in user-generated content generated by users in a plurality of services during a predetermined period, and extracts features of the user-generated content. The features extracted here include features related to the web content of the destination website and features related to the plurality of user-generated content generated during the predetermined period.

[0075] Specifically, the extraction unit 15g first uses the URL described in the collected user-generated content as the entry URL, accesses the entry URL, and identifies the URL of the site finally reached, that is, the arrival URL. Note that if the entry URL is one that uses a URL shortening service, it is treated as the entry URL as it is.

[0076] Many of the URLs written in user-generated content are created using URL shortening services such as bit[.]ly and tinyuri[.]com. A URL shortening service converts long URLs into shorter, simpler URLs and issues them. Many URL shortening services associate long URLs on other sites with short URLs issued by their own service, and redirect the access to the short URL to the original long URL.

[0077] Therefore, the extraction unit 15g creates a Web crawler by combining, for example, a scraping framework Scrapy and a headless browser Splash capable of Javascript (registered trademark) rendering. As a result, the extraction unit 15g accesses a URL described in the user-generated content and records communication information.

[0078] For example, the extraction unit 15g records the Web content of the final website and the number of redirections. In the case of a communication pattern in which the entry URL is "http: / / bit.ly / aaa" → "http: / / redirect.com / " → the destination URL is "http: / / malicious.com", the number of redirections (2) and the Web content of the final website "malicious.com" are recorded.

[0079] Then, the extraction unit 15g extracts features of the Web content, such as the number of HTML tags of the destination site, distributed representations of character strings displayed on the destination site, the number of redirects, the number of FQDNs (fully qualified domain names) transitioning from the entrance URL to the destination URL, etc. Here, the tags to be counted in HTML are, for example, the top 30 tags that frequently appear on malicious sites, which enables the extraction unit 15g to extract features of malicious user-generated content.

[0080] Furthermore, the extraction unit 15g identifies the IP address of the FQDN that is ultimately reached. Furthermore, when the same IP address is reached from a plurality of services at the same time, the extraction unit 15g regards a set of these user-generated contents as a set of similar user-generated contents.

[0081] Then, extraction unit 15g extracts features of the user-generated contents, such as the number of user-generated contents, the number of services, the number of entrance URLs, the number of users, and distributed representations of text, from the set of similar user-generated contents.

[0082] Learning unit 15e learns using the extracted feature quantities of user-generated content generated by legitimate users and feature quantities of content generated by malicious users. Furthermore, determination unit 15f determines whether or not the user-generated content is generated by a malicious user based on the learned model.

[0083] Specifically, the learning unit 15e performs supervised learning of a machine learning model using the extracted feature amount of the Web content of the final destination Web site and the feature amount of the user-generated content generated at the same time. Furthermore, the determination unit 15f uses the learned machine learning model to determine whether the user-generated content of the service acquired thereafter is malicious or not.

[0084] In this way, learning unit 15e learns the characteristics of a set of user-generated contents that are generated in a similar context at a specific timing such as an event and have a URL that reaches the same IP address and is highly likely to be malicious. Therefore, determination unit 15f can use the learning result to determine whether user-generated contents collected in real time are malicious.

[0085] Moreover, when it is determined that the user-generated content is generated by a malicious user, the extraction unit 15g outputs the characteristics of the attack of the user-generated content as threat information. Here, FIG. 11 and FIG. 12 are diagrams for explaining the threat information. As shown in FIG. 11, the threat information includes, for example, a key phrase included in the user-generated content, and an entrance URL, a destination URL, and the like described in the user-generated content of each service. In the example shown in FIG. 11, the user-generated content of service a and service b including the key phrase "rugby world cup", the entrance URL described in each, and the destination URL common to services a and b are shown. The extraction unit 15g outputs these threat information to a predetermined destination via the output unit 12 or the communication control unit 13.

[0086] Specifically, as shown in Fig. 12, threat information includes alerts such as notifications to recipients, blacklists, etc. In the example shown in Fig. 12, for example, user-generated content with a context including words such as "regularly held (once a week), free, live broadcast, J-League" is alerted. In particular, an attacker's account using this context and the abused service are reported. In addition, a blacklist is presented that includes the entrance URL written in the user-generated content, the relay URL to which the entry URL is transferred, and the destination URL to which the relay URL is ultimately reached.

[0087] In addition, in the example shown in Figure 12, it is presented that the destination URL of the malicious user-generated content in the above context and the malicious user-generated content in the context containing the words "regularly held (once every four years), free, live broadcast, Tokyo Olympics", etc., is a common malicious site.

[0088] In this way, the extraction function unit 15C performs a maliciousness determination on user-generated content that is likely to be malicious and was generated in large quantities on any service during the same period, using a feature amount obtained by accessing an entrance URL. Furthermore, when the user-generated content is determined to be malicious, the extraction function unit 15C extracts threat information from the malicious user-generated content and outputs a threat report. This makes it possible for the extraction function unit 15C to detect in real time attacks that could be threats among user-generated content that is likely to be malicious and was generated in large quantities on any service during the same period, and to output attack information.

[0089] In addition, when the above-mentioned determination function unit 15B determines that the user-generated content is malicious, extraction unit 15g may output attack characteristics such as character strings and URLs contained in the guidance context of the user-generated content as threat information.

[0090] [Detection process] Next, the detection process by the detection device 1 according to this embodiment will be described with reference to Fig. 13 to Fig. 17. First, Fig. 13 is a flowchart showing the collection process procedure of the collection function unit. The flowchart in Fig. 13 starts, for example, when the user performs an operation input to instruct the start.

[0091] First, acquisition unit 15a acquires user-generated content generated in each service during a predetermined period (step S1). Specifically, acquisition unit 15a acquires the user-generated content from the server of each service via input unit 11 or communication control unit 13.

[0092] Next, the generating unit 15b generates a search query using words that appear in the user-generated content for each service. For example, the generating unit 15b generates a search query using a combination of the appearing words (step S2).

[0093] Furthermore, the generating unit 15b calculates the maliciousness of the search query for each service, and selects a search query whose calculated maliciousness is equal to or greater than a threshold as a search query for user-generated content that may be malicious for the service.

[0094] The collection unit 15c uses the selected search query to collect user-generated content generated in a predetermined service (step S3), thereby completing a series of collection processes.

[0095] Next, Fig. 14 and Fig. 15 are flowcharts showing the processing procedure of the determination function unit. First, the flowchart in Fig. 14 shows the learning process in the determination function unit 15B, which is started, for example, when the user performs an operation input to instruct the start of the learning process.

[0096] Calculation unit 15d calculates features of user-generated content of a predetermined service collected by collection function unit 15A during a predetermined period (step S4). Specifically, calculation unit 15d calculates text features representing features of combinations of words co-occurring in a plurality of pieces of user-generated content, and group features representing features related to similarities of words among a plurality of pieces of user-generated content generated during a predetermined period.

[0097] Furthermore, the learning unit 15e performs learning using the calculated feature amounts of the user-generated content generated by the authorized user and the feature amounts of the content generated by the malicious user (step S5), thereby completing a series of learning processes.

[0098] Next, the flowchart in FIG. 15 shows the determination process in the determination function unit 15B, which is started, for example, at the timing when the user performs an operation input to instruct the start of the process.

[0099] Calculation unit 15d calculates the feature amount of the user-generated content of a predetermined service collected by collection function unit 15A during a predetermined period (step S4).

[0100] Next, the determination unit 15f determines whether or not the user-generated content has been generated by a malicious user, using the learned model (step S6), thereby completing a series of determination processes.

[0101] 16 and 17 are flowcharts showing the processing procedure of the extraction function unit. First, the flowchart in Fig. 16 shows the learning process in the extraction function unit 15C, which is started, for example, when the user performs an operation input to instruct the start of the learning process.

[0102] First, extraction unit 15g accesses the entrance URLs described in the user-generated content of the multiple services collected by collection function unit 15A during a predetermined period, and extracts the feature amounts of the user-generated content (step S14). Specifically, extraction unit 15g extracts the feature amounts of the web content of the destination website and the feature amounts of the multiple user-generated content generated during the predetermined period.

[0103] Furthermore, the learning unit 15e performs learning using the extracted feature amounts of the user-generated content generated by the authorized users and the feature amounts of the content generated by the malicious users (step S5), thereby completing a series of learning processes.

[0104] Next, the flowchart in FIG. 17 shows the determination process in the extraction function unit 15C, which is started, for example, at the timing when the user performs an operation input to instruct the start of the process.

[0105] First, extraction unit 15g accesses the entrance URLs described in the user-generated content of a plurality of services collected by collection function unit 15A during a predetermined period, and extracts the feature amount of the user-generated content (step S14).

[0106] Furthermore, the determination unit 15f determines, based on the learned model, whether or not the user-generated content has been generated by a malicious user (step S6).

[0107] Then, when the determination unit 15f determines that the user-generated content has been generated by a malicious user, the extraction unit 15g outputs the characteristics of the attack of this user-generated content as threat information (step S7). This ends a series of determination processes.

[0108] Note that, similarly to the process of Fig. 17, the process of step S7 may be performed after the process of step S6 shown in Fig. 15. That is, when determination function unit 15B determines that the user-generated content has been generated by a malicious user, extraction unit 15g may output the attack characteristics of this user-generated content as threat information.

[0109] As described above, in the collection function unit 15A of this embodiment, the acquisition unit 15a acquires user-generated content generated in each service during a predetermined period. In addition, the generation unit 15b generates a search query using words that appear in the user-generated content for each service. In addition, the collection unit 15c collects user-generated content generated in multiple services using the generated search query.

[0110] This allows collection function unit 15A to efficiently collect user-generated content that is likely to be malicious and is distributed in a context similar to a specific timing, and as a result, detection device 1 can quickly and accurately detect malicious sites over a wide area.

[0111] Furthermore, the generator 15b selects a search query that is likely to be malicious for each service, which enables the collection function unit 15A to easily and quickly collect user-generated content that is likely to be malicious for each service.

[0112] In addition, in the determination function unit 15B, the calculation unit 15d calculates the feature amount of the user-generated content generated by the user during a predetermined period. Furthermore, the learning unit 15e learns using the calculated feature amount of the user-generated content generated by the legitimate user and the feature amount of the content generated by the malicious user. Furthermore, the determination unit 15f determines whether the user-generated content is generated by the malicious user or not, based on the learned model.

[0113] This allows determination function unit 15B to learn the characteristics of user-generated content generated at specific times such as events, and to use the learning results to determine whether user-generated content collected in real time is malicious. In this way, determination function unit 15B can detect malicious sites quickly and accurately.

[0114] In addition, the features of the user-generated content calculated by calculation unit 15d include text features representing the characteristics of combinations of words that co-occur in multiple user-generated contents, and group features representing the characteristics related to the similarities of words among multiple user-generated contents generated during a specified period.

[0115] This enables determination function unit 15B to learn using the characteristics of user-generated content that is highly likely to be malicious, and to use the learning results to determine whether user-generated content collected in real time is malicious.

[0116] In addition, in extraction function unit 15C, extraction unit 15g accesses entrance URLs described in user-generated content generated by users in multiple services during a predetermined period and extracts features of the user-generated content. Furthermore, learning unit 15e learns using the features of the extracted user-generated content generated by legitimate users and the features of the content generated by malicious users. Furthermore, determination unit 15f determines whether the user-generated content is generated by a malicious user using the learned model.

[0117] This allows extraction function unit 15C to determine whether user-generated content collected in real time is malicious, using the characteristics of user-generated content of various services that is generated at specific times such as events, etc. In this way, extraction function unit 15C can quickly and accurately detect malicious sites over a wide range.

[0118] The feature quantity extracted by extraction unit 15g includes feature quantities related to the web content of the destination web site and feature quantities related to a plurality of user-generated contents generated during a predetermined period of time, which enables extraction function unit 15C to extract valid threat information of malicious sites.

[0119] In addition, when it is determined that the user-generated content is generated by a malicious user, extraction unit 15g outputs the characteristics of the attack of the user-generated content as threat information, which enables extraction function unit 15C to present valid threat information of malicious sites to a predetermined destination.

[0120] In addition, in the detection device 1 of this embodiment, the acquisition unit 15a acquires user-generated content generated in each service during a predetermined period. The generation unit 15b generates a search query using words that appear in the user-generated content for each service. The collection unit 15c collects user-generated content generated in a plurality of services using the generated search query. The calculation unit 15d calculates the feature amount of the collected user-generated content of the predetermined service. The learning unit 15e learns using the feature amount of the user-generated content generated by the legitimate user and the feature amount of the content generated by the malicious user. The determination unit 15f determines whether the user-generated content is generated by the malicious user using the learned model. In addition, when the extraction unit 15g determines that the user-generated content is generated by the malicious user, it accesses the entrance URL described in the user-generated content and outputs the attack feature of the user-generated content as threat information.

[0121] This allows the detection device 1 to quickly detect malicious user-generated content by using the characteristics of user-generated content generated at specific times such as events, and to present valid threat information of malicious sites to a predetermined recipient. In this way, the detection device 1 can quickly detect malicious sites over a wide area.

[0122] In addition, the generator 15b selects search queries that are likely to be malicious for each service, which enables the detection device 1 to easily collect user-generated content that is likely to be malicious and more quickly detect malicious user-generated content.

[0123] The features of the user-generated content calculated by calculation unit 15d include text features representing features of combinations of words co-occurring in multiple user-generated content, and group features representing features related to similarities of words among multiple user-generated content generated during a predetermined period. This enables detection device 1 to process user-generated content that is likely to be malicious, and detect malicious user-generated content more quickly.

[0124] Furthermore, learning unit 15e learns using the feature amounts of the user-generated content of the multiple services extracted by extraction unit 15g, and determination unit 15f determines whether the user-generated content of the multiple services is generated by a malicious user or not, based on the learned model. This makes it possible to more quickly detect malicious user-generated content by using the features of the user-generated content of any service.

[0125] The feature quantity extracted by the extraction unit 15g includes a feature quantity related to the web content of the destination website and a feature quantity related to a plurality of user-generated contents generated during a predetermined period of time, which enables the detection device 1 to present valid threat information of malicious sites to a predetermined destination.

[0126] [program] A program in which the process executed by the detection device 1 according to the above embodiment is written in a language executable by a computer can also be created. As an embodiment, the detection device 1 can be implemented by installing a detection program that executes the above detection process as package software or online software on a desired computer. For example, the above detection program can be executed by an information processing device, so that the information processing device can function as the detection device 1. The information processing device referred to here includes desktop or notebook personal computers. In addition, the information processing device also includes mobile communication terminals such as smartphones, mobile phones, and PHS (Personal Handyphone System), and further slate terminals such as PDA (Personal Digital Assistant). The function of the detection device 1 may be implemented in a cloud server.

[0127] 18 is a diagram showing an example of a computer that executes a detection program. The computer 1000 includes, for example, a memory 1010, a CPU 1020, a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.

[0128] The memory 1010 includes a ROM (Read Only Memory) 1011 and a RAM 1012. The ROM 1011 stores a boot program such as a BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to a hard disk drive 1031. The disk drive interface 1040 is connected to a disk drive 1041. A removable storage medium such as a magnetic disk or an optical disk is inserted into the disk drive 1041. The serial port interface 1050 is connected to a mouse 1051 and a keyboard 1052, for example. The video adapter 1060 is connected to a display 1061, for example.

[0129] Here, the hard disk drive 1031 stores, for example, an OS 1091, an application program 1092, a program module 1093, and program data 1094. Each piece of information described in the above embodiments is stored in the hard disk drive 1031 or memory 1010, for example.

[0130] The detection program is stored in the hard disk drive 1031 as, for example, a program module 1093 in which instructions to be executed by the computer 1000 are written. Specifically, the hard disk drive 1031 stores the program module 1093 in which each process executed by the detection device 1 described in the above embodiment is written.

[0131] Furthermore, data used for information processing by the detection program is stored as program data 1094, for example, in the hard disk drive 1031. Then, the CPU 1020 reads out the program module 1093 and the program data 1094 stored in the hard disk drive 1031 into the RAM 1012 as necessary, and executes each of the above-mentioned procedures.

[0132] Note that the program module 1093 and program data 1094 related to the detection program are not limited to being stored in the hard disk drive 1031, and may be stored in, for example, a removable storage medium and read by the CPU 1020 via the disk drive 1041 or the like. Alternatively, the program module 1093 and program data 1094 related to the detection program may be stored in another computer connected via a network such as a LAN or a WAN (Wide Area Network), and read by the CPU 1020 via the network interface 1070.

[0133] Although the embodiment to which the invention made by the present inventor is applied has been described above, the present invention is not limited by the description and drawings which form a part of the disclosure of the present invention according to the present embodiment. In other words, other embodiments, examples, operation techniques, etc. made by those skilled in the art based on the present embodiment are all included in the scope of the present invention. [Explanation of symbols]

[0134] 1. Detection device 11 Input section 12 Output section 13 Communication control section 14 Storage section 15 Control section 15A Collection function unit 15B Judgment function section 15C Extraction function section 15a Acquisition part 15b Generator 15c Collection Department 15d Calculation section 15e Learning Department 15f Judgment section 15g extraction part

Claims

1. an acquisition unit that acquires user-generated content generated in each service during a predetermined period of time; a generation unit that extracts a key phrase expressing a context of the user-generated content by using words appearing in the user-generated content for each service, generates a search query including the key phrase, calculates a maliciousness of the search query based on the search query used to search for user-generated content determined to be malicious within a predetermined period of the past for each service, and selects the search query whose maliciousness is equal to or exceeds a predetermined threshold as a search query that may be malicious for the service; a collection unit that collects user-generated content generated in a plurality of services using the selected search query; A collection device comprising:

2. The collection device according to claim 1 , wherein the generation unit generates a search query using a combination of appearing words.

3. The collection device according to claim 1 , wherein the generation unit generates a search query using a frequency of occurrence of each word.

4. A collection method performed by a collection device, comprising: acquiring user-generated content generated in each service during a predetermined period of time; a generation step of extracting a key phrase expressing the context of the user-generated content by using words appearing in the user-generated content for each service, generating a search query including the key phrase, calculating the maliciousness of the search query based on the search query used to search for user-generated content determined to be malicious within a predetermined period of time for each service, and selecting the search query whose maliciousness is equal to or exceeds a predetermined threshold as a search query that may be malicious for the service; a collection step of collecting user-generated content generated in a plurality of services using the selected search query; A collection method comprising:

5. acquiring user-generated content generated in each service during a predetermined period of time; a generation step of extracting a key phrase expressing a context of the user-generated content by using words appearing in the user-generated content for each service, generating a search query including the key phrase, calculating a maliciousness of the search query based on the search query used to search for user-generated content determined to be malicious within a predetermined period of time for each service, and selecting the search query whose maliciousness is equal to or exceeds a predetermined threshold as a search query that may be malicious for the service; a collection step of collecting user-generated content generated in a plurality of services using the selected search query; A collection program for causing a computer to execute the above.

Citation Information

Patent Citations

  • Illegal content search device, and illegal content search method and program

    JP2019174926A

  • Search spam analysis and detection

    US20110087648A1

  • Method and system for generating phrase blacklist to prevent certain content from appearing in a search result in response to search queries

    US20180218079A1