Information collection device, information collection method, and program

Abstract

This information collection device has: a first acquisition unit that is configured to identify a trusted node that a given node trusts on the basis of a condition that is set for each of a plurality of nodes constituting a network, the condition being that one of the nodes trusts another of the nodes, and to acquire the condition of each of the trusted nodes; and a second acquisition unit that is configured to acquire information relating to an object of interest from the identified node on the basis of the condition acquired by the first acquisition unit. Due to the information collection device having the first and second acquisition unit, the possibility that highly reliable information can be collected is raised.

Description

[Technical Field]

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

[0002] When individuals buy or sell information or goods over the Internet, they may refer to reviews to assess whether the other party is trustworthy. Similarly, when making decisions, they may refer to information such as reviews written by others.

[0003] Countermeasures against fake information are an issue when it comes to information such as reviews and articles written by third parties. When information is managed on a centralized platform with a centralized administrator, it is the administrator's responsibility to prevent the falsification of transactions and accounts, thereby minimizing the amount of fake information. However, it is difficult for providers to take measures against tampering, and at the same time, the responsibility for taking measures is heavy, which, combined with the maintenance of the system, increases service costs.

[0004] As a possible solution to these problems, there are decentralized infrastructure technologies such as blockchain and DAG that can provide security of stored data at low cost without the service provider having to take responsibility. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] Sepandar D. Kamvar, Mario T. Schlosser, Hector GarciaMolina, "The EigenTrust Algorithm for Reputation Management in P2P Networks", [online], Internet<URL:https: / / dl.acm.org / doi / pdf / 10.1145 / 775152.775242> [Non-patent document 2] Jakob Schaerer, Severin Zumbrunn, Torsten Braun, "Veritaa: A distributed public key infrastructure with signature store", [online], Internet<URL:https: / / onlinelibrary.wiley.com / doi / epdf / 10.1002 / nem.2183> Summary of the Invention [Problem to be solved by the invention]

[0006] However, conventional technologies have difficulty dealing with Sybil attacks on reviews and other information. As a result, it is difficult to collect reliable information, and there is a risk that fake information will be used as reference.

[0007] The present invention has been made in view of the above points, and has an object to increase the possibility of collecting highly reliable information. [Means for solving the problem]

[0008] In order to solve the above problem, the information collection device is configured for each of the plurality of nodes that make up the network. A trust relationship indicating a trusted node and a range of nodes that the node references in a graph based on the trust relationship. Based on conditions , among the plurality of nodes A node The set of nodes included in the range indicated by the condition set in Identify the Belonging to the set In front of each Chronicles The system has a first acquisition unit configured to acquire the conditions of a node, and a second acquisition unit configured to acquire information about a certain target from a node identified based on the conditions acquired by the first acquisition unit. [Effects of the Invention]

[0009] This can increase the likelihood of collecting reliable information. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a diagram illustrating an example of a hardware configuration of a terminal 10 used by a node in an embodiment of the present invention. [Figure 2] 2 is a diagram illustrating an example of a functional configuration of a terminal 10 according to an embodiment of the present invention. [Figure 3] 10 is a flowchart illustrating an example of a processing procedure executed by the terminal 10 when collecting reviews on a certain object. [Figure 4] FIG. 10 is a diagram for explaining a confidence range. [Figure 5] FIG. 10 is a diagram illustrating an example of provisional f. [Figure 6] FIG. 10 is a diagram illustrating an example of node information. DETAILED DESCRIPTION OF THE INVENTION

[0011] In this embodiment, a node refers to a party that evaluates trustworthiness or a party whose trustworthiness is evaluated. Each node has a terminal connected to a network.

[0012] In this embodiment, a framework is proposed in which each of a plurality of nodes can select reviews from nodes that the node can trust from among the reviews (which cannot be tampered with) shared on a decentralized platform.

[0013] In this framework, each person shares their method for narrowing down the range of trustworthy nodes and their method for evaluating reputation on a distributed ledger, which enables semi-automatic selection of methods and parameters by referencing the shared information, and also ensures user satisfaction by allowing the shared information to be changed by individual nodes.

[0014] More specifically, in this embodiment, in a set of nodes that are evaluated as trustworthy by the "node trust evaluation method," not only the reviews themselves but also the set of the "node trust evaluation method" and the "node information aggregation method" used by each node, which is equal to the trust evaluation function f, are shared. Then, the following (a) to (c) are realized.

[0015] (a) A user can define their own f by referring to the f of a node they trust (a combination of "node trust evaluation method" and "node information aggregation method"). If the majority of people within their trusted range (direct trust + indirect trust) share the correct f, the commonly used f is considered to be a safe evaluation method that takes into account each person's attack countermeasures.

[0016] To achieve this, each node publishes its own "trust relationship" and "trust evaluation function f" as a set in a tamper-proof form. This makes it possible to share f, confirm the trust relationship with the provider of f (closeness in the trust chain, number of paths), and confirm the number of uses of f.

[0017] In addition, publishing incorrect information can damage trust in one's own node. In such cases, disadvantages arise, such as indirect trusters excluding one's own node from the range of trust they have, so the system can suppress incorrect information.

[0018] (b) When an attack such as a fake review occurs, knowledgeable people can deal with it by modifying their own f, and those who do not have knowledge can deal with it by using f that has been fixed by other users, thereby increasing the reliability of reviews.

[0019] Rather than simply obtaining f from other nodes and using it as is, each node can define and use f by verifying and modifying the conditions, which can be verified, and be "satisfied" with the surrounding attack reviews.

[0020] For this purpose, f is a function whose processing conditions (such as the range of trustworthy indirect trust and the method of majority voting statistics taking into account the closeness of the trust relationship) can be confirmed, and each of its variables, conditions, and calculations can be reviewed and changed by the user as needed.

[0021] (c) Even if there are few trustworthy nodes, sufficient reliability and information volume can be ensured by aggregating the “node trust evaluation method” and “node information aggregation method” from random seed nodes and evaluating them against other nodes.

[0022] In other words, by continuing to publish and propagate their own f (publish → use → update → publish...), even if a new attack is launched, appropriate countermeasures can be taken as collective intelligence. By filtering and rating reviews using a commonly used f, it is possible to evaluate the reliability of a person by inheriting the best practices of people within one's trusted circle.

[0023] An embodiment of the present invention will be described below with reference to the drawings. Fig. 1 is a diagram showing an example of the hardware configuration of a terminal 10 used by a node in an embodiment of the present invention. The terminal 10 in Fig. 1 has a drive device 100, an auxiliary storage device 102, a memory device 103, a CPU 104, an interface device 105, a display device 106, and an input device 107, which are all interconnected by a bus B.

[0024] A program for realizing processing on the terminal 10 is provided by a recording medium 101 such as a CD-ROM. When the recording medium 101 storing the program is set in the drive device 100, the program is installed from the recording medium 101 to the auxiliary storage device 102 via the drive device 100. However, the program does not necessarily have to be installed from the recording medium 101, but may be downloaded from another computer via a network. The auxiliary storage device 102 stores the installed program as well as necessary files, data, etc.

[0025] The memory device 103 reads and stores the program from the auxiliary storage device 102 when an instruction to start the program is received. The CPU 104 realizes functions related to the terminal 10 in accordance with the program stored in the memory device 103. The interface device 105 is used as an interface for connecting to a network. The display device 106 displays a GUI (Graphical User Interface) or the like according to the program. The input device 107 is composed of a keyboard, a mouse, etc., and is used to input various operation instructions.

[0026] Fig. 2 is a diagram showing an example of the functional configuration of the terminal 10 according to the embodiment of the present invention. In Fig. 2, the terminal 10 includes a node information acquisition unit 11, an f aggregation unit 12, a review aggregation unit 13, a publishing unit 14, and a correction unit 15. Each of these units is realized by a process executed by the CPU 104 of one or more programs installed in the terminal 10.

[0027] In this embodiment, each node (terminal 10) publishes f ("node trust evaluation method" and "node information aggregation method") used to evaluate trustworthiness in a distributed ledger or the like in a form that cannot be tampered with.

[0028] The "node trust evaluation method" is a technique for identifying a trusted node (hereinafter referred to as a "trusted node"), and includes parameters indicating the conditions (hereinafter referred to as "trust conditions") for being a trusted node (defining the trust range) and filters for the trust conditions.

[0029] The trust condition may be, for example, one or more of the following (1) to (5). (1) Indirect trust object degree (2) Number of users (3) Whether or not the ID is certified by a third party (4) Random reference (5) Designated Direct Trustee The indirect trust degree in (1) is a condition for identifying the trust range by the number of hops recursively tracing the trust relationship.

[0030] FIG. 4 is a diagram for explaining a trust range. In FIG. 4, the trust relationship is expressed in a graph format. In the graph shown in FIG. 4, black circles represent nodes. The leftmost node is the node that is the starting point of the trust range (hereinafter referred to as the "starting node"). The lines connecting the nodes represent the trust relationship. The direct trust range and the indirect trust range are each indicated by a dashed line. As shown in the figure, the indirect trust range also includes the direct trust range.

[0031] The group of nodes included in the indirect trust range, which has n hops from the origin node, is called the nth-order indirect trustee. Since the group of nodes included in the direct trust range has 1 hop, this group of nodes corresponds to the first-order indirect trustee. The indirect trustee degree is this value of n.

[0032] The number of usage records in (2) is a condition for identifying a trusted node based on the number of usage records of f that the node publishes.

[0033] (3) The presence or absence of third-party certification of the ID is a condition for determining a node as a trusted node if its ID (node ​​ID) is certified (for example, signed) by a third party.

[0034] The random reference in (4) is a condition in which m nodes randomly selected from the indirect trust range are trusted nodes. Therefore, in the case of random reference, the value of m is specified as a parameter.

[0035] The designated direct trustee in (5) is a condition that a designated node among direct trustees (primary indirect trustees) is designated as a trusted node.

[0036] Two or more values ​​may be specified from (1) to (5). In this case, the logical product of the conditions indicated by the selected values ​​becomes the valid condition.

[0037] On the other hand, a filter is a parameter for excluding some nodes from a group of nodes (trust range) that meet the trust conditions. For example, a filter can exclude f, which is related to a node with an absolutely low evaluation, from a group of nodes that meet the trust conditions based on the EigenTrust value.

[0038] Examples of filters include the following (1) to (3). (1) Sybil Limit (2) EigenTrust (3) Full Trust (1) SybilLimit is a filtering method that focuses on the lack of social connections between honest users and attackers (https: / / ieeexplore.ieee.org / document / 4531141).

[0039] (2) EigenTrust is a filtering method that focuses on trust evaluation through the propagation and convergence of trust values ​​(http: / / ilpubs.stanford.edu:8090 / 562 / 1 / 2002-56.pdf).

[0040] The full trust in (3) means no exclusion, i.e., in this case, no node is excluded from the group of trusted nodes that meets the trust conditions.

[0041] The "node information aggregation method" is a method for aggregating node information published by each trusted node identified by the "node trust evaluation method" into a single piece of node information. In this embodiment, the "node information aggregation method" is composed of an extraction method, a processing method, a minimum number of reviews, etc.

[0042] The extraction method is a method of extracting one piece of node information that aggregates the node information of each trusted node. For example, the following values ​​can be specified as the extraction method. (1) Average extraction (2) Most frequent extraction (1) Average extraction is to extract the average of node information. (2) Mode extraction is to extract the node information that appears most frequently (most frequently).

[0043] The processing method indicates the calculation method for the parameters of the node information extracted by the extraction method. For example, for (2) most frequent extraction, examples of processing methods include rounding off the average value of the parameters of the node information with the highest frequency of appearance (most common), or weighting according to the indirect trust degree.

[0044] The minimum number of reviews is a threshold for determining whether or not to select f of the node information aggregated by the extraction method and processing method as a candidate for the node's own f. Specifically, if the number of reviews obtained from each trusted node is equal to or greater than the minimum number of reviews, f of the aggregated node information is selected as a candidate for the node's own f.

[0045] The processing procedure executed by the terminal 10 will be described below. Fig. 3 is a flowchart for explaining an example of the processing procedure executed by the terminal 10 when collecting reviews for a certain object. The object for which reviews are to be collected (hereinafter referred to as "object of interest") is input by the user. For example, the object of interest is specified by the name of a specific object, such as the name of a restaurant.

[0046] In step S101, the node information acquisition unit 11 acquires provisional information (hereinafter referred to as "provisional f") of its own node f (a set of "node trust evaluation method" and "node information aggregation method"). The provisional f may be stored in advance in the auxiliary storage device 102 or the like, or may be automatically generated by the node information acquisition unit 11.

[0047] Fig. 5 is a diagram showing an example of tentative f. Fig. 5 shows a "node trust evaluation method" and a "node information aggregation method" that constitute tentative f.

[0048] In the "Node Trust Evaluation Method" in Figure 5, "Random Reference, param=100" is a setting related to the trust condition, and "Filter=SybilLimit, FilterParam=10" is a setting related to the filter. In other words, in this example, the trust condition is to randomly select 100 nodes from the indirect trust range as trusted nodes. In addition, the SybilLimit method is set as the filter, and 10 is set as the parameter for that method.

[0049] The "node information aggregation method" in Figure 5 shows that the f with the highest appearance frequency is extracted, the average value is used as the parameter of the extracted f, and the minimum number of reviews is 5.

[0050] Following step S101, the node information acquisition unit 11 acquires a list of trusted nodes based on the "node trust evaluation method" of the node f (S102). When step S102 is executed for the first time, the node f is a provisional f. In this case, 100 nodes are randomly selected from the indirect trust range.

[0051] The indirect range of trust can be specified based on the "trust relationship" included in the node information (set for each node) that each node makes public.

[0052] FIG. 6 is a diagram showing an example of node information. As shown in FIG. 6, the node information includes "node trust evaluation method" and "node information aggregation method" that constitute f, as well as "trust relationship" and "review." Of these, "trust relationship" is information indicating nodes trusted by the node related to the node information. The example in FIG. 6 indicates that node B trusts node A. That is, FIG. 6 shows an example of node information for node B. Note that each node may have performed reviews for multiple targets. Therefore, each node information may include multiple "reviews."

[0053] This node information is made public for each node in a way that cannot be tampered with. For example, each node information is recorded in a distributed ledger.

[0054] Next, the node information acquisition unit 11 acquires f ("node trust evaluation method" and "node information aggregation method") and "reviews" from the node information of each trusted node included in the list of trusted nodes (S103). Note that with regard to "reviews," it is sufficient to acquire only "reviews" related to the target of interest.

[0055] Next, the f aggregating unit 12 aggregates the f ("node trust evaluation method" and "node information aggregation method") of each trusted node based on the "node information aggregation method" of the local node (S104). When step S104 is executed for the first time, the "node information aggregation method" of the local node is the "node information aggregation method" of the provisional f (FIG. 5). Therefore, in this case, the f aggregating unit 12 extracts the most frequent f from the f ("node trust evaluation method" and "node information aggregation method") of each trusted node and aggregates the f by setting the average value of the parameters of the extracted f as the parameters of the aggregated f. Note that, as is clear from the calculation of the average value of the parameters, the values ​​of the parameters of the trust condition, filter, extraction method, and processing method do not matter in determining the commonality of f when identifying the most frequent f. Since the minimum number of reviews is a numerical value itself, the difference does not matter in determining the commonality. However, the method for determining the commonality of f may be changed as appropriate.

[0056] Next, the f aggregating unit 12 determines whether or not the number of "reviews" acquired in step S103 is less than the minimum number of reviews that configures the "node information aggregating method" of the own node (S105).

[0057] If the number of "reviews" is less than the minimum number of reviews (Yes in S105), that is, if the number of reviews is less than the expected number, the aggregation unit 12 determines whether the number of repetitions from step S102 onwards is less than the upper limit (S106). If the number of repetitions from step S102 onwards exceeds the upper limit (No in S106), the processing procedure in FIG. 3 ends.

[0058] If the number of repetitions of step S102 and subsequent steps is less than the upper limit (Yes in S106), the aggregation unit 12 selects, from among the "node trust evaluation methods" to be aggregated by the "node information aggregation method" of the own node, the "node trust evaluation method" with the broadest trust range (loosest trust condition) than the "node trust evaluation method" as the "node trust evaluation method" of the own node (S107), and repeats step S102 and subsequent steps. For example, if the trust condition of each "node trust evaluation method" to be aggregated is random reference, the "node trust evaluation method" with the largest number of random selections is selected. If the trust condition of each "node trust evaluation method" to be aggregated is indirect trust object degree, the "node trust evaluation method" with the largest indirect trust object degree n is selected. By adopting a "node trust evaluation method" with a relatively wide trust range, it is expected that reviews will be collected from a larger number of nodes.

[0059] On the other hand, if the number of "reviews" acquired in step S103 is equal to or greater than the minimum number of reviews configuring the "node information aggregation method" of the local node (No in S105), the review aggregator 13 determines whether or not to prioritize the aggregated f over the local node's f (S108). Whether or not to prioritize the aggregated f over the local node's f may be set in advance, or may be input by the user in step S108. If the local node's f is prioritized (No in S108), step S109 is not executed and the process proceeds to step S110. If the aggregated f is prioritized (Yes in S108), the review aggregator 13 sets the aggregated f as the local node's f (replaces the local node's f with the aggregated f) (S109) and proceeds to step S110.

[0060] In step S110, the review aggregator 13 aggregates the "reviews" acquired in step S103 based on the "node information aggregation method" of the own node. Here, if step S109 has been executed, the "node information aggregation method" of the own node is the "node information aggregation method" of the aggregated f. The "reviews" are aggregated, for example, by calculating the mode or average number of stars based on the "node information aggregation method." Furthermore, comments on the "reviews" may be aggregated, for example, by generating a list of each "review" to be aggregated.

[0061] Next, the publishing unit 14 receives from the user whether or not the f of the own node needs to be corrected, and if not corrected, whether or not to publish it (S111). At this time, the publishing unit 14 may display the contents of the f of the own node, the aggregated "reviews", etc., on the display device 106 as information for making a decision.

[0062] For example, if the trust range is too broad, it may be targeted and become a target of attack. Therefore, the user can determine whether or not f needs to be modified by checking, for example, whether the trust range indicated by the "node trust evaluation method" of their own node f is too broad (whether it exceeds the range they can tolerate) or whether it deviates from their own reliability evaluation criteria. If the user determines that no modification is necessary, they input whether or not to make it public.

[0063] If the node's f does not need to be modified and can be made public (Yes in S112), the publishing unit 14 executes processing to publish the node's f (S113). The publishing process refers to making f accessible to other nodes. It is desirable for f to be published in a state where it is highly likely to be tampered with. For example, f may be published by recording it in a distributed ledger. This ensures tamper resistance through the distributed ledger. Furthermore, the use of the "node trust evaluation method" and "node information aggregation method" using zero-knowledge proof on the distributed ledger can be kept confidential from third parties. Furthermore, by executing a smart contract, the number of times a user references and uses public information can be managed, enabling the priority of f with a high number of uses to be increased. Note that when a published f is used, a financial reward may be given to the user who published the f, providing an incentive for publishing f.

[0064] Alternatively, if there is no need to modify f of the own node and f is to be kept private (No in S112 and Yes in S114), step S113 is not executed and the processing procedure in FIG. 3 ends.

[0065] Alternatively, if f of the own node needs to be corrected (No in S112 and No in S114), the correction unit 15 corrects f in accordance with a correction instruction from the user (S115). For example, a correction to narrow the trust range is made. Alternatively, if the user determines that the trust range is too narrow, a correction to widen the trust range may be made. The type of correction to be made is at the discretion of the user.

[0066] Next, steps S102 and after are repeated based on the corrected f, resulting in a "review" based on the corrected f.

[0067] As described above, according to this embodiment, it is possible to increase the possibility of collecting highly reliable reviews (information).

[0068] In addition, the "node trust evaluation method" and "node information aggregation method" are user-defined, allowing for settings and changes to suit individual use cases. Furthermore, users can adjust the parameters themselves. As the system does not impose optimal rule definitions, it provides greater user satisfaction and freedom.

[0069] Furthermore, since f is propagated between nodes, each user (node) can propagate indirect countermeasures against attacks to the trusted range by modifying its own rules when it detects an attack.

[0070] In addition, when reviews (information) are insufficient, the range of trust can be further expanded, thereby increasing the possibility of obtaining sufficient reviews (information).

[0071] In this embodiment, a review of a certain object has been described as an example of information about the object, but this embodiment may be applied to information other than reviews about the object. For example, this embodiment may be applied to articles and various other information about a certain object.

[0072] In this embodiment, the terminal 10 is an example of an information collection device. The node information acquisition unit 11 is an example of a first acquisition unit and a second acquisition unit. The aggregation unit 12 is an example of an aggregation unit.

[0073] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various modifications and variations are possible within the scope of the gist of the present invention as described in the claims. [Explanation of symbols]

[0074] 10 devices 11 Node information acquisition unit 12 f Collecting section 13 Review Aggregation Department 14 Public Section 15 Correction section 100 Drive device 101 Recording media 102 Auxiliary storage 103 Memory Device 104 CPU 105 Interface Device 106 Display device 107 Input Device B Bus

Claims

1. a first acquisition unit configured to identify a set of nodes included in a range indicated by a condition set for a certain node among the plurality of nodes, based on a trust relationship set for each of the plurality of nodes constituting the network, indicating nodes trusted by that node, and a condition indicating a range of nodes referred to by that node in a graph based on the trust relationship, and to acquire the condition for each of the nodes belonging to the set; a second acquisition unit configured to acquire information about a certain object from a node identified based on the condition acquired by the first acquisition unit; An information collection device comprising:

2. an aggregation unit configured to aggregate the conditions acquired by the first acquisition unit, the second acquisition unit is configured to acquire the information from a node identified based on the condition aggregated by the aggregation unit.

2. The information collection device according to claim 1.

3. the first acquisition unit is further configured to acquire a method of aggregating the conditions set for each of the nodes belonging to the set, the aggregating unit is configured to aggregate the conditions acquired by the first acquisition unit based on the aggregation method acquired by the first acquisition unit.

3. The information collection device according to claim 2.

4. a correction unit configured to correct the conditions summarized by the summarizing unit in response to an instruction from a user; the first acquisition unit is configured to identify the set based on the condition modified by the modification unit.

4. The information collection device according to claim 2 or 3.

5. a first acquisition step of identifying a set of nodes included in a range indicated by a condition set for a certain node among the plurality of nodes based on a trust relationship indicating nodes trusted by the node and a condition indicating a range of nodes referenced by the node in a graph based on the trust relationship, and acquiring the condition for each of the nodes belonging to the set; a second acquisition procedure for acquiring information about a certain object from a node identified based on the condition acquired by the first acquisition procedure; An information gathering method characterized by being executed by a computer.

6. 4. A program for causing a computer to function as the information collection device according to claim 1.

Images