A method and system for counter-user profiling

Abstract

The present invention relates to the technical field of user data analysis and management, and discloses a method for counter-user profiling, comprising the following steps: step S1, reading the content input by the user; step S2, performing semantic analysis on the content input by the user, and generating a series of requests that are opposite, contradictory, and irrelevant to the user's actual request; step S3, recording the five-tuple information of the user's original request and the request generated by the counter-user profiling system, and sending all the requests to the service platform after recording; step S4, receiving all the request data returned by the service platform, comparing the five-tuple information of each response, removing the response data triggered by the request generated by the counter-user profiling system itself, and returning the response data of the user's actual request to the user. The present invention also discloses a system for counter-user profiling, comprising a proxy access module, a reverse request data generation module, a request data marking and sending module, and a return data filtering module. The present invention is used to solve the problem of user profiling abuse.

Description

Technical Field

[0001] The present invention relates to the technical field of user data analysis and management, and specifically, to an anti-user profiling method and system for solving the problem of user profiling abuse. Background Art

[0002] With the rise of the Internet, e-commerce, taxi-hailing software, travel and other Internet service providers provide users with various services, which bring many conveniences. At the same time, users' various online operations are also recorded by the service platform, and user portraits are generated based on the collected user access data and access habits to provide users with personalized services. However, due to the abuse of user portraits, problems such as big data price discrimination, automatic recommendation of high-priced goods, and inducing consumption have also emerged. For these problems, users have no active choice, and whether they are abused or not depends entirely on the service provider.

[0003] In view of this, in order to solve the problem that the existing technology only focuses on how to reasonably utilize user portraits but does not involve preventing user portraits from being collected, the present invention provides an anti-user portrait method and system. When users access various service platforms, they are first processed by this system through a proxy. While forwarding the user's normal request, this system automatically generates a request that is inconsistent or inconsistent with the actual data based on the user's actual access data, and sends it to the platform together. Due to the existence of this system, the user's access data collected by the platform is confused and contradictory, making it difficult to generate a usable user portrait, thereby solving the problem of user portrait abuse. Summary of the Invention

[0004] The purpose of the present invention is to provide an anti-user profiling method to solve the problem of user profiling abuse.

[0005] The present invention is implemented through the following technical solution: A method for anti-user profiling, comprising the following steps:

[0006] Step S1, reading the content input by the user;

[0007] Step S2: semantically analyze the content input by the user to generate a series of requests that are opposite, contradictory, or irrelevant to the user's actual request;

[0008] Step S3: Record the five-tuple information of the user's original request and the request generated by the reverse user profiling system, and send all requests to the service platform after recording;

[0009] Step S4: Receive all request data returned by the service platform, compare the five-tuple information of each response, remove the response data triggered by the request generated by the anti-user profiling system itself, and return the response data of the user's actual request to the user.

[0010] In order to better implement the present invention, further, step S1 includes:

[0011] The anti-user profiling system reads the user's browsing data and the user's input requests by accessing the user's Internet traffic.

[0012] To better implement the present invention, further, the method of generating a series of requests that are opposite to or contradictory to the actual requests of the user in step S2 includes:

[0013] Use word segmentation technology to segment the content entered by the user and split the request content entered by the user into individual words;

[0014] The segmented words are input into the antonym API platform to obtain the antonym API of each word, and then the antonyms are obtained, and the antonyms are combined again to form content that is opposite and contradictory to the content requested by the user.

[0015] In order to better implement the present invention, further, the method of generating a series of requests unrelated to the user's actual request in step S2 includes a method of injecting a large number of random words into the request input by the user and a method of shifting and flipping the request input by the user.

[0016] In order to better implement the present invention, further, the word segmentation technology includes jieba word segmentation technology.

[0017] In order to better implement the present invention, further, the antonym API platform in step S3 includes an aggregated data platform, a binary data platform and a FREE API platform.

[0018] In order to better implement the present invention, further, the five-tuple information in step S3 includes source IP information, source port information, destination IP information, destination port information and protocol information.

[0019] In order to better implement the present invention, further, step S4 includes:

[0020] Compare the five-tuple information of each response obtained with the five-tuple information of each recorded request. If the five-tuple information is inconsistent with the user request, it is determined that the response is triggered by the request generated by this system and does not need to be returned to the user and is directly discarded. On the contrary, if the five-tuple information is the same as the user request, it is determined that the response is triggered by the user request, and the response data is returned to the user normally.

[0021] In order to better implement the present invention, the present invention further provides a system for anti-user profiling, including a proxy access module, a reverse request data generation module, a request data marking and sending module, and a return data filtering module, wherein:

[0022] The proxy access module is used to access the user's Internet traffic, obtain the user's browsing data to read the content entered by the user, and send the read content to the reverse request data generation module;

[0023] The reverse request data generation module is used to perform semantic analysis on the user's browsing data and generate a series of requests that are opposite, contradictory, or irrelevant to the user's actual request;

[0024] The request data marking and sending module is used to record the five-tuple information of the user's original request and the request generated by the reverse user profiling system, and then send all requests to the service platform after recording;

[0025] The return data filtering module is used to receive all request data returned by the service platform, compare the five-tuple information of each response, eliminate the response data triggered by the request generated by the anti-user profiling system itself, and return the response data of the user's actual request to the user.

[0026] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0027] (1) The present invention weakens and hides the user's actual Internet habits by constructing a large number of requests that are opposite, contradictory, or irrelevant to the user's actual requests, making it impossible for service providers to effectively classify and profile users based on their Internet behavior. The control of user profiles is given to the users themselves rather than the Internet service providers, thus preventing the abuse of user profiles.

[0028] (2) The present invention generates a large number of reverse or irrelevant requests based on the user's actual browsing requests, making it difficult to effectively collect user profiles;

[0029] (3) The present invention filters the response data by marking, so that the user will not receive the response data of the request not initiated by himself, which will not affect the normal use of the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The present invention is further described in conjunction with the following drawings and embodiments, and all concepts and innovations of the present invention should be regarded as disclosed contents and the protection scope of the present invention.

[0031] Figure 1 A flowchart of a method for countering user profiling provided by the present invention.

[0032] Figure 2 A structural diagram of an anti-user profiling system provided by the present invention. DETAILED DESCRIPTION

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. It should be understood that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments, and therefore should not be regarded as limiting the scope of protection. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technical personnel in this field without making creative work are within the scope of protection of the present invention.

[0034] In the description of the present invention, it should be noted that user portrait refers to a labeled user model abstracted based on information such as user attributes, user preferences, living habits, and user behavior.

[0035] The API platform is an API platform that provides data services.

[0036] Jieba is a word segmentation library for Python.

[0037] Example 1:

[0038] A method of counter-user profiling in this embodiment, such as Figure 1 As shown, the present invention weakens and hides the user's actual Internet habits by constructing a large number of requests that are opposite, contradictory, and irrelevant to the user's actual requests, making it impossible for service providers to effectively label and classify users and create user profiles based on their Internet behaviors. The control of user profiles is given to the users themselves rather than the Internet service providers, thereby avoiding the abuse of user profiles.

[0039] Example 2:

[0040] This example further optimizes Example 1. In this example, jieba is a Python library that splits a sentence into individual words, such as "The weather is very good today" into "today," "weather," and "very good." Jieba internally matches word prefixes. Because a sentence doesn't have antonyms, we need to use it to split the user-entered sentence into individual words and then generate antonyms for the corresponding words.

[0041] The rest of this embodiment is the same as that of embodiment 1, so it will not be described again.

[0042] Example 3:

[0043] This embodiment further optimizes the above-mentioned embodiments 1 or 2. In this embodiment, two API platforms, Aggregate Data and Binary Data, provide methods for obtaining antonyms for individual words. We only use one of them. The API platform is only used to obtain antonyms, in order to generate data that is completely opposite to the original data.

[0044] The rest of this embodiment is the same as that of the above-mentioned embodiment 1 or 2, and thus will not be described in detail.

[0045] Example 4:

[0046] This embodiment further optimizes any one of the above embodiments 1-3. In this embodiment, irrelevant content is added by injecting a large number of random words into the original content, shifting the original content, and flipping it. Specifically, the process is as follows:

[0047] First, these methods can be implemented within the anti-user profiling system. For example, if the user inputs "The weather is good today"

[0048] Method 1: A large number of random words. This only requires randomly generating characters. Each programming language provides a way to generate random data. For example, the random data "Chinese xxx English asbadf characters! @###￥￥%￥%" is generated, and then randomly injected into the original sentence. The final sentence may be like this: "Today xxx weather astronomy##!@The weather is very abcd English good", achieving the purpose of being almost unrelated to the original sentence.

[0049] Method 2: Content displacement: the position of each word is shuffled, such as "The weather is very good today" becomes "The weather is very good today".

[0050] That is, this anti-user profiling system can directly inject a large number of random words into the original content, displace the original content, flip it, etc., without using API receipts. This is to generate data unrelated to the original data. There are two purposes for this: one is to generate opposite requests, and the other is to generate unrelated but not necessarily opposite requests. Opposite requests require the help of the API platform, while unrelated requests are randomly generated. The random algorithm can be called at the code level. The following example uses Python to implement the idea:

[0051] First, call the character library (Python's string library) to generate a set of all characters (all characters include uppercase and lowercase English letters, numbers, punctuation marks, etc.);

[0052] Second, call the random library (python's random library), which provides a method for randomly selecting characters, random.choice(), which can randomly select characters from the character set in 1 to form a string of any length;

[0053] Third, randomly concatenate the random string generated in the second point into the user's original sentence, and you will get a new sentence of "original sentence + random string";

[0054] Fourth, repeat the second and third points several times to get a new statement that is more complex and longer and more different from the original statement.

[0055] The rest of this embodiment is the same as any of the above embodiments 1-3, so it will not be repeated here.

[0056] Example 5:

[0057] This embodiment is further optimized based on any one of the above embodiments 1-4. This embodiment provides a system for anti-user profiling, such as Figure 2 As shown, module A is a proxy access module, which is used to access the user's Internet traffic, obtain the user's browsing data and send it to module B. Module B is a reverse request data generation module, which is used to perform semantic analysis on the user's browsing data and generate a series of requests that are opposite, contradictory, or irrelevant to the user's actual request. Specifically:

[0058] For antonyms, we first use word segmentation technologies such as Jieba to split the user's request into individual words. We then call API platforms such as Aggregate Data, Binary Data, and Free API to obtain the antonym API to obtain the antonym of each word. We then combine the antonyms into content that is opposite or contradictory to the user's request.

[0059] For irrelevant content, we can inject a large number of random words into the original content, or transpose or flip the original content. For irrelevant requests, we don't need to segment the words; we can simply randomize and transpose the entire sentence to create a new sentence that is almost unrelated to the original information.

[0060] Module C is the request data tagging and sending module. It records the five-tuple information (source IP, source port, destination IP, destination port, and protocol) of the user's original request and the request generated by this system. After recording, all requests are sent to the service platform. The five-tuple information of the connection generated by each request is inconsistent, mainly the source port inconsistency.

[0061] Module D is the return data filtering module. It receives all request data returned by the service platform, compares the five-tuple information of each response, removes the responses triggered by the request generated by the proxy system itself, and returns the response data of the user's actual request to the user. Specifically, the five-tuple information of each response is compared with the five-tuple information of each request recorded in module C (request data marking and sending module). If the five-tuple information is inconsistent with the five-tuple information of the user request, it is determined that the response is triggered by the request generated by this system and does not need to be returned to the user. It is directly discarded. On the contrary, if the five-tuple information is the same as the user request, it is determined that the response is triggered by the user request, and the response data is returned to the user normally.

[0062] The rest of this embodiment is the same as any of the above embodiments 1-4, so it will not be repeated here.

[0063] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention falls within the scope of protection of the present invention.

Claims

1. A method for counter-user profiling, characterized in that: The following steps are involved: Step S1, reading the content input by the user; Step S2: semantically analyze the content input by the user to generate a series of requests that are opposite, contradictory, or irrelevant to the user's actual request; Step S3: Record the five-tuple information of the user's original request and the request generated by the reverse user profiling system, and send all requests to the service platform after recording; Step S4: Receive all request data returned by the service platform, compare the five-tuple information of each response, remove the response data triggered by the request generated by the anti-user profiling system itself, and return the response data of the user's actual request to the user; The method of generating a series of requests that are opposite to or contradictory to the actual requests of the user in step S2 includes: Use word segmentation technology to segment the content entered by the user and split the request content entered by the user into individual words; Input the segmented words into the antonym API platform to obtain the antonym API of each word, and then obtain the antonym. The antonym is recombined into content that is opposite or contradictory to the user's request content; The method of generating a series of requests unrelated to the actual request of the user in step S2 includes a method of injecting a large number of random words into the request input by the user and a method of shifting or flipping the request input by the user; The step S4 comprises: Compare the five-tuple information of each response obtained with the five-tuple information of each recorded request. If the five-tuple information is inconsistent with the user request, it is determined that the response is triggered by the request generated by this system and does not need to be returned to the user and is directly discarded. On the contrary, if the five-tuple information is the same as the user request, it is determined that the response is triggered by the user request, and the response data is returned to the user normally.

2. The method for anti-user profiling according to claim 1, characterized in that: The step S1 comprises: The anti-user profiling system reads the user's browsing data and the user's input requests by accessing the user's Internet traffic.

3. The method for anti-user profiling according to claim 1, characterized in that: The word segmentation technology includes jieba word segmentation technology.

4. The method for anti-user profiling according to claim 1, characterized in that: The antonym API platform in step S3 includes an aggregated data platform, a binary data platform and a FREE API platform.

5. The method for anti-user profiling according to claim 1, characterized in that: The five-tuple information in step S3 includes source IP information, source port information, destination IP information, destination port information and protocol information.

6. A system for anti-user profiling, configured to execute the anti-user profiling method according to claim 1; characterized in that: It includes a proxy access module, a reverse request data generation module, a request data marking and sending module, and a return data filtering module, among which: The proxy access module is used to access the user's Internet traffic, obtain the user's browsing data to read the content entered by the user, and send the read content to the reverse request data generation module; The reverse request data generation module is used to perform semantic analysis on the user's browsing data and generate a series of requests that are opposite, contradictory, or irrelevant to the user's actual request; The request data marking and sending module is used to record the five-tuple information of the user's original request and the request generated by the reverse user profiling system, and then send all requests to the service platform after recording; The return data filtering module is used to receive all request data returned by the service platform, compare the five-tuple information of each response, eliminate the response data triggered by the request generated by the anti-user profiling system itself, and return the response data of the user's actual request to the user.

Images