An artificial intelligence information processing method

Abstract

The present invention relates to the field of artificial intelligence information security technology, and in particular to an artificial intelligence information processing method, the method comprising: allocating a simulated role with a corresponding authority range according to received user login information; receiving user access information; retrieving data corresponding to the access information in a database according to the access information; issuing an alarm prompt, issuing an interception prompt or turning on a defense mode according to the access information; tracking the access path information of the simulated role in a virtual host, determining that the simulated role belongs to a mis-retrieval or a vulnerability attack and repairing the vulnerability; building a next virtual host and entering virtual information; the present invention adopts the above method, by analyzing the received user access information, judging whether the information the user intends to access is within the user's authority range, and by counting the frequency of occurrence of sensitive words, effectively improving the accuracy of access information analysis, and further improving the security and practicality of artificial intelligence information processing.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence information security technology, and in particular to an artificial intelligence information processing method. Background Art

[0002] The application of authentication and authorization fields such as face recognition, voice control and fingerprint began before artificial intelligence. Many applications have not yet been equipped with technical prevention measures to deal with artificial intelligence deep fakes, resulting in security risks in authentication and insecure access to personal data.

[0003] Chinese Patent Publication Number: CN118200021A. A method, device, equipment, medium and program product for preventing network attacks are disclosed. The invention discloses a method, device, equipment, medium and program product for preventing network attacks, which can be applied to the fields of artificial intelligence and information security technology. The method includes: using a pre-trained user input recognition model to identify a user access request to obtain a recognition result, wherein the recognition result includes that the user access request belongs to one of a security category, a risk category and an unsafe category, wherein the security levels corresponding to the security category, the risk category and the unsafe category decrease in sequence; when the user access request belongs to the risk category, the user access request is matched with the keywords in the vulnerability vocabulary library; if the match is successful, the user access request is intercepted, otherwise the user access request is released.

[0004] Chinese Patent Publication No.: CN114528397A. A method, system, medium, computer device and terminal for defending against adversarial text attacks are disclosed. The invention discloses a method, system, medium, computer device and terminal for defending against adversarial text attacks, including: preprocessing text data; defense against attacks of the same type of characters; defense against character-level adversarial samples; defense against word-level adversarial samples; generating new defensive text. The present invention processes the text at the source without obtaining and modifying the model structure. It can not only target various forms of adversarial text attacks, but also retain complete semantic information and achieve a higher defense effect without obtaining model parameters, thereby simply and efficiently improving the defense performance of text adversarial samples. The present invention processes text with adversarial disturbances through three different modules, restores the semantics of the text itself to eliminate the disturbance and achieve a defense effect. It can efficiently defend against mixed adversarial text attacks of different granularities without obtaining model parameters and data and without retraining the model.

[0005] It can be seen that the above technical solution has the following problems: simply matching vulnerability words cannot effectively limit vulnerability attacks, and using characters to counter attacks can easily lead to misjudgment. Summary of the invention

[0006] To this end, the present invention provides an artificial intelligence information processing method to overcome the problem in the prior art that simply matching vulnerability words cannot effectively limit vulnerability attacks, and using characters to counter attacks is very likely to cause misjudgment.

[0007] To achieve the above object, the present invention provides an artificial intelligence information processing method, comprising:

[0008] Determine the authority scope of the user according to the received user login information and assign a simulation role corresponding to the authority scope;

[0009] Receiving access information of a user and determining whether the access information falls within the authority scope of the simulated role;

[0010] Retrieving data corresponding to the access information within the authority scope of the simulated role in a database according to the access information;

[0011] issuing an alarm, an interception prompt or turning on a defense mode according to the extent to which the access information exceeds the scope of the authority, and, when the defense mode is turned on, assigning the simulated role to a virtual host and providing virtual information;

[0012] Tracking the access path information of the simulated role in the virtual host, determining that the simulated role is a mis-retrieval or a vulnerability attack and repairing the vulnerability, and isolating the virtual host attacked by the vulnerability;

[0013] Using a plurality of simulated roles in the isolated virtual hosts, constructing a next virtual host and entering the virtual information;

[0014] Among them, the simulated role is an access user with a range of permissions; the alarm prompt includes a prompt that the search content exceeds the permission; the interception prompt includes a prompt that the user is prohibited from searching; the defense mode is to prohibit the simulated role from searching in the database; the virtual host is a false database built to defend against the vulnerability attack; the virtual information is false information provided by the virtual host corresponding to the access information.

[0015] Furthermore, the steps of determining the authority scope of the user according to the received login information of the user and assigning a simulated role corresponding to the authority scope include:

[0016] According to the preset scope of authority of each user, upon receiving the login information of the user, determining the scope of authority of the user corresponding to the login information;

[0017] Setting a simulated role within the scope of the user's authority;

[0018] The login information, the simulated role and the user authority range are set in a one-to-one correspondence.

[0019] Furthermore, the step of receiving the user's access information and determining whether the access information belongs to the authority range of the simulated role includes:

[0020] Receiving user access information and splitting the access information into a plurality of original keywords;

[0021] The number of sensitive keywords matched by the original keyword is counted and recorded as the number of sensitive words;

[0022] Determining whether the access information falls within the authority scope of the simulated role according to the number of sensitive words;

[0023] The sensitive keywords are keywords that are not within the authority scope of the simulated character.

[0024] Further, whether the access information belongs to the authority range of the simulated role is determined according to the comparison result of the number of sensitive words and the number of preset keywords, wherein:

[0025] If the number of sensitive words is less than the number of preset keywords, it is determined that the access information belongs to the authority range of the simulated role;

[0026] If the number of the sensitive words is not less than the number of the preset keywords, it is determined that the access information does not fall within the authority scope of the simulated role;

[0027] Among them, the number of preset keywords is negatively correlated with the level of user authority range.

[0028] Further, in a state where it is determined that the access information belongs to the authority scope of the simulated role, data corresponding to the access information is retrieved in the authority scope of the simulated role in a database according to the access information;

[0029] If the data corresponding to the access information cannot be retrieved, it is prompted that the access information needs to be modified and then searched again.

[0030] Further, when it is determined that the access information does not belong to the authority scope of the simulated role, the number of characters of the sensitive keywords matched in the access information is analyzed,

[0031] Among them, the number of characters corresponding to several identical sensitive keywords is recorded as the number of characters corresponding to a single sensitive keyword.

[0032] Further, according to the comparison results of the number of characters of the sensitive keyword with the first preset number of characters and the second preset number of characters, the extent to which the access information exceeds the scope of the authority is determined and an alarm prompt is issued, an interception prompt is issued, or a defense mode is turned on, wherein:

[0033] If the number of characters in the sensitive keyword is less than the first preset number of characters, it is determined that the degree to which the access information exceeds the scope of the authority is the first level of exceeding the scope and an alarm is issued to the user;

[0034] If the number of characters of the sensitive keyword is not less than the first preset number of characters and less than the second preset number of characters, it is determined that the degree to which the access information exceeds the scope of authority is the second level of exceeding the scope and an interception prompt is issued to the user;

[0035] If the number of characters of the sensitive keyword is not less than the second preset number of characters, it is determined that the degree to which the access information exceeds the scope of authority is the third level and a defense mode is activated;

[0036] The first preset number of characters and the second preset number of characters are negatively correlated with the user authority range level.

[0037] Further, in a state where the defense mode is turned on, the simulated character is assigned to a virtual host and virtual information is provided;

[0038] Wherein, the access information input by the simulated character in the virtual host can retrieve corresponding virtual information.

[0039] Further, the access path information of the simulated role is tracked in the virtual host, and it is determined that the simulated role belongs to the mis-retrieval or the vulnerability attack and the vulnerability is repaired, wherein:

[0040] If the number of clicks on the access information beyond the authority range of the simulated character in the virtual host is less than the preset number of clicks, it is determined that the simulated character belongs to the erroneous retrieval;

[0041] If the number of clicks on the access information of the simulated character that exceeds the scope of authority in the virtual host is not less than the preset number of clicks, it is determined that the simulated character is subject to the vulnerability attack and the vulnerability is repaired;

[0042] The preset number of clicks is negatively correlated with the level of the user's authority range.

[0043] Further, the steps of using a plurality of simulated roles in the isolated virtual hosts to construct the next virtual host and input the virtual information include:

[0044] Collect access information of several simulated roles in the isolated virtual host, set the proportion of access information corresponding to the first level over-range, the second level over-range and the third level over-range respectively to build the virtual host;

[0045] The virtual information of the access information corresponding to the first level out-of-range, the second level out-of-range and the third level out-of-range respectively is entered.

[0046] Compared with the prior art, the beneficial effect of the present invention lies in determining the user's authority range based on the user's login information received and allocating a simulated role with the corresponding authority range. By setting the simulated role to search in the database, the user is prevented from directly accessing the database. While preliminarily isolating the direct information interaction between the database and the user, the security and practicality of artificial intelligence information processing are effectively improved.

[0047] Furthermore, by analyzing the access information received from the user, it is determined whether the information the user intends to access is within the scope of the user's authority. By counting the frequency of occurrence of sensitive words, the accuracy of the access information analysis is effectively improved. While effectively improving the accuracy of judging whether it is within the scope of user authority, it further improves the security and practicality of artificial intelligence information processing.

[0048] Furthermore, by analyzing the number of sensitive characters, the degree to which access to information exceeds the authority range can be determined, and an alarm prompt or interception prompt can be issued or a defense mode can be turned on based on the judgment result. When the defense mode is turned on, a simulated role is assigned to a virtual subject and virtual information is provided, which effectively improves the ability to block vulnerability attacks while further enhancing the security and practicality of artificial intelligence information processing.

[0049] Furthermore, a virtual host is constructed according to the access information corresponding to different degrees of exceeding the scope of authority, and virtual information is entered. While effectively updating the virtual host, the security and practicality of artificial intelligence information processing are further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flowchart of the artificial intelligence information processing method of the present invention;

[0051] Figure 2 A schematic diagram of determining access information according to an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of determining sensitive keywords in an embodiment of the present invention;

[0053] Figure 4 Schematic diagram of determining vulnerability attacks according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0055] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0056] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0057] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0058] See also Figure 1 As shown, it is a flow chart of the artificial intelligence information processing method of the present invention, an artificial intelligence information processing method, comprising:

[0059] Step S1, determining the user's authority range according to the received user's login information and assigning a simulation role corresponding to the authority range;

[0060] Step S2, receiving the user's access information and determining whether the access information falls within the authority range of the simulated role;

[0061] Step S3, retrieving data corresponding to the access information within the authority scope of the simulated role in the database according to the access information;

[0062] Step S4, issuing an alarm prompt, issuing an interception prompt or turning on a defense mode according to the extent to which the access information exceeds the authority range, and, when the defense mode is turned on, assigning a simulated role to a virtual host and providing virtual information;

[0063] Step S5, tracking the access path information of the simulated role in the virtual host, determining that the simulated role is a mis-search or a vulnerability attack and repairing the vulnerability, and isolating the virtual host attacked by the vulnerability;

[0064] Step S6, using a number of simulated roles in the isolated virtual hosts to build the next virtual host and input virtual information;

[0065] Among them, the simulated role is an access user with a range of permissions; the alarm prompt includes a prompt that the search content exceeds the permission; the interception prompt includes a prompt that the user is prohibited from searching; the defense mode is to prohibit the simulated role from searching in the database; the virtual host is a false database built to defend against vulnerability attacks; the virtual information is false information corresponding to the access information provided by the virtual host.

[0066] Specifically, in step S1, the steps of determining the authority range of the user according to the received login information of the user and assigning a simulated role corresponding to the authority range include:

[0067] Step S11, according to the preset permission range of each user, when receiving the login information of the user, determining the user permission range corresponding to the login information;

[0068] Step S12, setting a simulated role within the user's authority range;

[0069] Among them, the login information, simulation role and user authority range are set in a one-to-one correspondence.

[0070] It is understandable that the user's authority range is set in advance in the user's login information, and when setting the user to log in to the database, a simulated role is set to access the database instead of the user.

[0071] Specifically, the user's authority range is determined based on the user's login information and a simulated role with the corresponding authority range is assigned. By setting the simulated role to search the database, the user is prevented from directly accessing the database. While initially isolating the direct information interaction between the database and the user, it effectively improves the security and practicality of artificial intelligence information processing.

[0072] See also Figure 2 As shown, it is a schematic diagram of determining access information in an embodiment of the present invention. In step S2, the steps of receiving the user's access information and determining whether the access information belongs to the authority range of the simulated role include:

[0073] Step S21, receiving the user's access information and splitting the access information into a number of original keywords;

[0074] Step S22, counting the number of sensitive keywords matched by the original keyword and recording it as the number of sensitive words;

[0075] Step S23, determining whether the access information falls within the authority scope of the simulated role according to the number of sensitive words;

[0076] Among them, sensitive keywords are keywords that are not within the authority scope of the simulated role.

[0077] Specifically, in step S23, it is determined whether the access information belongs to the authority range of the simulated role according to the comparison result between the number of sensitive words and the number of preset keywords, wherein:

[0078] If the number of sensitive words is less than the preset number of keywords, it is determined that the access information falls within the authority scope of the simulated role;

[0079] If the number of sensitive words is not less than the preset number of keywords, it is determined that the access information does not fall within the authority scope of the simulated role;

[0080] Embodiment 1: In this embodiment, the number of preset keywords is 5. If the number of sensitive words is 3, which is less than the number of preset keywords, 5, it is determined that the access information belongs to the authority range of the simulated role;

[0081] If the number of sensitive words (6) is greater than the preset number of keywords (5), it is determined that the access information does not fall within the authority scope of the simulated role.

[0082] Among them, the number of preset keywords is negatively correlated with the level of user authority range.

[0083] It is understandable that the higher the level of user authority, the more data can be accessed and the fewer sensitive keywords are involved.

[0084] Optionally, when the user authority range level is high, medium and low in descending order, if the user authority range level is high, the number of preset keywords is 3;

[0085] If the user's authority range level is intermediate, the number of preset keywords is 5;

[0086] If the user's permission range level is low, the preset number of keywords is 7.

[0087] Specifically, by analyzing the access information received from users, we determine whether the information the user intends to access is within the scope of user authority. By counting the frequency of occurrence of sensitive words, we effectively improve the accuracy of access information analysis. While effectively improving the accuracy of judging whether it is within the scope of user authority, it further improves the security and practicality of artificial intelligence information processing.

[0088] Specifically, in step S3, when it is determined that the access information belongs to the authority scope of the simulated role, data corresponding to the access information is retrieved in the database within the authority scope of the simulated role according to the access information;

[0089] If the data corresponding to the access information cannot be retrieved, it will prompt that the access information needs to be modified and then searched again.

[0090] Specifically, when it is determined that the access information does not fall within the authority scope of the simulated role, the number of characters of the sensitive keywords matched in the access information is analyzed.

[0091] Among them, the number of characters corresponding to several identical sensitive keywords is recorded as the number of characters corresponding to the sensitive keyword once.

[0092] See also Figure 3 As shown, it is a schematic diagram of determining sensitive keywords in an embodiment of the present invention. In step S4, according to the comparison results of the number of characters of the sensitive keyword with the first preset number of characters and the second preset number of characters, the degree to which the access information exceeds the authority range is determined and an alarm prompt is issued, an interception prompt is issued, or a defense mode is turned on, wherein:

[0093] If the number of characters in the sensitive keyword is less than the first preset number of characters, the degree to which the access information exceeds the scope of authority is determined to be the first level of over-range and an alarm is issued to the user;

[0094] If the number of characters in the sensitive keyword is not less than the first preset number of characters and less than the second preset number of characters, the degree to which the access information exceeds the scope of authority is determined to be the second level of excess and an interception prompt is issued to the user;

[0095] If the number of characters in the sensitive keyword is not less than the second preset number of characters, the degree of the access information exceeding the scope of authority is determined to be the third level of exceeding the scope and the defense mode is turned on;

[0096] Embodiment 2: Based on Embodiment 1, this embodiment is different from the embodiment 1 in that the first preset number of characters is 4 bytes, the second preset number of characters is 8 bytes, and if the number of characters of the sensitive keyword is 2 bytes, which is less than the first preset number of characters, 4 bytes, then the degree to which the access information exceeds the scope of authority is determined to be the first level of over-range and an alarm prompt is issued to the user;

[0097] If the number of characters of the sensitive keyword is 6 bytes, which is greater than the first preset number of characters, 4 bytes, and less than the second preset number of characters, 8 bytes, then the degree of the access information exceeding the scope of authority is determined to be the second level of over-range and an interception prompt is issued to the user;

[0098] If the number of characters of the sensitive keyword is 12 bytes, which is greater than the second preset number of characters, 8 bytes, then the degree of the access information exceeding the scope of authority is determined to be the third level of excess and the defense mode is turned on;

[0099] The first preset number of characters and the second preset number of characters are negatively correlated with the user authority range level.

[0100] It is understandable that the higher the level of user authority range, the more data that can be accessed, the fewer sensitive keywords involved, and the fewer the first preset number of characters and the second preset number of characters.

[0101] Optionally, when the user authority range levels are high, medium and low from high to low, if the user authority range level is high, the first preset number of characters is 2 bytes, and the second preset number of characters is 6 bytes;

[0102] If the user authority range level is intermediate, the first preset number of characters is 4 bytes, and the second preset number of characters is 8 bytes;

[0103] If the user authority range level is low, the first preset number of characters is 6 bytes, and the second preset number of characters is 10 bytes.

[0104] Specifically, the degree to which access to information exceeds the authority range is determined by analyzing the number of sensitive characters, and an alarm or interception prompt is issued or a defense mode is turned on based on the judgment result. When the defense mode is turned on, a simulated role is assigned to a virtual subject and virtual information is provided, which effectively blocks vulnerability attacks while further improving the security and practicality of artificial intelligence information processing.

[0105] Specifically, in the state where the defense mode is turned on, the simulated role is assigned to the virtual host and virtual information is provided;

[0106] Among them, the access information entered by the simulated character in the virtual host can retrieve the corresponding virtual information.

[0107] See also Figure 4 As shown, it is a schematic diagram of determining vulnerability attacks in an embodiment of the present invention. In step S5, the access path information of the simulated role is tracked in the virtual host, and it is determined that the simulated role belongs to a mis-retrieval or a vulnerability attack and the vulnerability is repaired, wherein:

[0108] If the number of clicks on the access information that exceeds the authority range of the simulated role in the virtual host is less than the preset number of clicks, the simulated role is judged to be a mis-retrieval;

[0109] If the number of clicks on the access information that exceeds the permission range of the simulated role in the virtual host is not less than the preset number of clicks, the simulated role is judged to be a vulnerability attack and the vulnerability is repaired;

[0110] Embodiment 3: Based on Embodiment 2, this embodiment is different from the embodiment 2 in that the preset number of clicks is 3 times. If the number of clicks of the simulated character in the virtual host that exceeds the scope of authority is 2 times and is less than the preset number of clicks 3 times, it is determined that the simulated character is a mis-retrieval;

[0111] If the number of clicks on access information that exceeds the scope of the authority of the simulated role in the virtual host is 4 times, which is greater than the preset number of clicks of 3 times, the simulated role is judged to be a vulnerability attack and the vulnerability is repaired;

[0112] The number of preset clicks is negatively correlated with the level of user authority scope.

[0113] It is understandable that the higher the level of user authority, the more data can be accessed, the fewer sensitive keywords are involved, the fewer the number of false searches during the search process, and the lower the preset number of clicks.

[0114] Optionally, when the user authority range level is high, medium and low in descending order, if the user authority range level is high, the preset number of clicks is 2;

[0115] If the user's authority range level is intermediate, the preset number of clicks is 3;

[0116] If the user authority range level is low, the preset number of clicks is 4 times.

[0117] Specifically, in step S6, the steps of using a plurality of simulated roles in isolated virtual hosts to construct the next virtual host and input virtual information include:

[0118] Step S61, collecting access information of several simulated roles in the isolated virtual host, setting the proportion of access information corresponding to the first level out-of-range, the second level out-of-range and the third level out-of-range respectively to build the virtual host;

[0119] Step S62, respectively input virtual information of the access information corresponding to the first level out-of-range, the second level out-of-range and the third level out-of-range.

[0120] It is understandable that when a simulated role is used to construct a virtual host, a virtual database is constructed by calling the simulated role's access information, access path information, and retrieval file and other related features to shuffle, insert error codes, and delete codes.

[0121] It is understandable that in the constructed virtual database, different proportions of virtual information are configured according to different levels of retrieval content, making the constructed virtual database more reliable.

[0122] Specifically, a virtual host is built based on the access information corresponding to the degree of exceeding the authority scope, and virtual information is entered. This effectively updates the virtual host and further improves the security and practicality of artificial intelligence information processing.

[0123] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0124] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An artificial intelligence information processing method, characterized in that: include: Determine the authority scope of the user according to the received user login information and assign a simulation role corresponding to the authority scope; Receiving access information of a user and determining whether the access information falls within the authority scope of the simulated role; Retrieving data corresponding to the access information within the authority scope of the simulated role in a database according to the access information; issuing an alarm, an interception prompt or turning on a defense mode according to the extent to which the access information exceeds the scope of the authority, and, when the defense mode is turned on, assigning the simulated role to a virtual host and providing virtual information; Tracking the access path information of the simulated role in the virtual host, determining that the simulated role is a mis-retrieval or a vulnerability attack and repairing the vulnerability, and isolating the virtual host attacked by the vulnerability; Using access information corresponding to different degrees of exceeding the authority of a plurality of simulated roles in the isolated virtual hosts, constructing a next virtual host and entering the virtual information; The simulated role is an access user with a range of permissions; the alarm prompt includes a prompt for exceeding the permissions for searching content; the interception prompt includes a prompt for prohibiting the user from searching; the defense mode prohibits the simulated role from searching in the database; the virtual host is a false database built to defend against the vulnerability attack; The virtual information is false information provided by the virtual host and corresponding to the access information.

2. The artificial intelligence information processing method according to claim 1, characterized in that: The steps of determining the authority scope of the user according to the received login information of the user and assigning a simulated role corresponding to the authority scope include: According to the preset scope of authority of each user, upon receiving the login information of the user, determining the scope of authority of the user corresponding to the login information; Setting a simulated role within the scope of the user's authority; The login information, the simulated role and the user authority range are set in a one-to-one correspondence.

3. The artificial intelligence information processing method according to claim 2, characterized in that: The steps of receiving the user's access information and determining whether the access information belongs to the authority range of the simulated role include: Receiving user access information and splitting the access information into a plurality of original keywords; Counting the number of sensitive words of the sensitive keyword matched by the original keyword; Determining whether the access information falls within the authority scope of the simulated role according to the number of sensitive words; The sensitive keywords are keywords that are not within the authority scope of the simulated character.

4. The artificial intelligence information processing method according to claim 3, characterized in that: According to the comparison result between the number of sensitive words and the number of preset keywords, it is determined whether the access information belongs to the authority scope of the simulated role, wherein: If the number of sensitive words is less than the number of preset keywords, it is determined that the access information belongs to the authority range of the simulated role; If the number of the sensitive words is not less than the number of the preset keywords, it is determined that the access information does not fall within the authority scope of the simulated role; Among them, the number of preset keywords is negatively correlated with the level of user authority range.

5. The artificial intelligence information processing method according to claim 4, characterized in that: In a state where it is determined that the access information belongs to the authority scope of the simulated role, data corresponding to the access information is retrieved in a database within the authority scope of the simulated role according to the access information; If the data corresponding to the access information cannot be retrieved, it is prompted that the access information needs to be modified and then searched again.

6. The artificial intelligence information processing method according to claim 5, characterized in that: In a state where it is determined that the access information does not belong to the authority scope of the simulated role, the number of characters of the matched sensitive keywords in the access information is analyzed, Among them, the number of characters corresponding to several identical sensitive keywords is recorded as the number of characters corresponding to a single sensitive keyword.

7. The artificial intelligence information processing method according to claim 6, characterized in that: According to the comparison results of the number of characters of the sensitive keyword with the first preset number of characters and the second preset number of characters, the extent to which the access information exceeds the scope of the authority is determined and an alarm prompt is issued, an interception prompt is issued, or a defense mode is turned on, wherein: If the number of characters of the sensitive keyword is less than the first preset number of characters, it is determined that the degree to which the access information exceeds the scope of the authority is a first level over-range situation and an alarm is issued to the user; If the number of characters of the sensitive keyword is not less than the first preset number of characters and less than the second preset number of characters, it is determined that the degree to which the access information exceeds the scope of the authority is a second level over-range situation and an interception prompt is issued to the user; If the number of characters of the sensitive keyword is not less than the second preset number of characters, it is determined that the degree to which the access information exceeds the scope of authority is a third level over-range situation and a defense mode is turned on; The first preset number of characters and the second preset number of characters are negatively correlated with the user authority range level.

8. The artificial intelligence information processing method according to claim 7, characterized in that: In a state where the defense mode is turned on, the simulated role is assigned to a virtual host and virtual information is provided; Wherein, the access information input by the simulated character in the virtual host can retrieve corresponding virtual information.

9. The artificial intelligence information processing method according to claim 8, characterized in that: Tracking the access path information of the simulated role in the virtual host, determining that the simulated role belongs to the mis-retrieval or the vulnerability attack and repairing the vulnerability, wherein: If the number of clicks on the access information beyond the authority range of the simulated character in the virtual host is less than the preset number of clicks, it is determined that the simulated character belongs to the erroneous retrieval; If the number of clicks on the access information of the simulated character that exceeds the scope of authority in the virtual host is not less than the preset number of clicks, it is determined that the simulated character is subject to the vulnerability attack and the vulnerability is repaired; The preset number of clicks is negatively correlated with the level of the user's authority range.

10. The artificial intelligence information processing method according to claim 9, characterized in that: The steps of using access information corresponding to different degrees of exceeding the authority range of a plurality of simulated roles in the isolated virtual hosts to construct a next virtual host and input the virtual information include: Collect access information of a number of simulated roles in the isolated virtual host, set the proportion of the access information corresponding to the first level out-of-range situation, the second level out-of-range situation and the third level out-of-range situation in the virtual host, and build the virtual host using the access information according to the proportion; The virtual information of the access information corresponding to the first level out-of-range situation, the second level out-of-range situation and the third level out-of-range situation is respectively entered.

Images