An intelligent retrieval system and method based on police knowledge base

Abstract

The present invention discloses an intelligent retrieval and analysis system and method based on a police knowledge base, including step S100: capturing and analyzing search trajectories generated by users based on the same information needs, and summarizing search trajectories of search processes generated by users based on different information needs; step S200: respectively, in different search processes generated by users based on different information needs, conducting regularity investigation on search trajectories contained in each search process; respectively, conducting abnormality investigation on different search processes generated by users based on different information needs; step S300: when it is determined that the user's search process is abnormal, triggering the system to capture the user's target demand information based on the browsing click record generated by the user inputting a search-type browsing; step S400: based on the captured user's target demand information and in combination with the user's account tag, realizing intelligent content push for the corresponding user account when the search process is abnormal.

Description

Technical Field

[0001] The present invention relates to the technical field of police knowledge base retrieval, and in particular to an intelligent retrieval system and method based on a police knowledge base. Background Art

[0002] At present, there are barriers to responsibilities and information between various police types and departments, and the responsibilities of different positions in the same department may not be clear. At the same time, if you want to find appropriate rules and regulations in a large number of laws and regulations, you need to go through tedious reading to obtain them. Moreover, in the actual process of handling police business, it is inevitable to touch upon knowledge areas that are not directly related to the position. At this time, due to the knowledge barriers in the position, it is even more difficult to find appropriate rules and regulations in a large number of laws and regulations. This phenomenon has affected the progress of work and increased communication costs.

[0003] Establishing a "mobile encyclopedia" through the system can realize keyword search similar to "Baidu", but there are difficulties in cumbersome and large amount of information query, and because of the emergence of knowledge barriers, users often have inaccurate expressions and non-standard descriptions when entering search formulas, which reduces search efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent retrieval system and method based on a police knowledge base to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: an intelligent retrieval and analysis method based on a police knowledge base, the method comprising:

[0006] Step S100: Based on the browsing record information generated by the user inputting the search formula, the system captures and analyzes the search tracks generated by the user based on the same information needs, and summarizes the search tracks of the user based on different information needs;

[0007] Step S200: in different search processes generated by users based on different information needs, respectively, each search track included in each search process is checked for regularity; based on the regularity checking of each search track included in each search process, respectively, different search processes generated by users based on different information needs are checked for abnormality;

[0008] Step S300: When it is determined that the user's search process is abnormal, the system is triggered to capture the user's target demand information based on the browsing click records generated by the user's input search browsing;

[0009] Step S400: Based on the captured target demand information of the user and combined with the user's account tags, implement intelligent content push for the user account corresponding to the abnormal search process.

[0010] Further, the process of calculating and capturing the search process of each user in Step S100 based on the browsing record features generated by each user through input search browsing includes:

[0011] Step S101: Corresponding all the user browsing and flipping records generated by a user's one input search formula to a search track generated by the user based on one input search formula; obtaining the set of search tracks {L1, L2,..., L n} generated by the user within the time period S; where L1, L2,..., L n respectively represent the 1st, 2nd,..., nth search tracks generated by the user in chronological order within the time period S; respectively extract search keywords from the input search formulas of each search track in the user search track set {L1, L2,..., L n} to obtain the keyword set corresponding to the search track set where, respectively represent the search keywords extracted from the input search formulas of L1, L2,..., L n ; where the number of search keywords extracted according to each input search formula is greater than or equal to 1;

[0012] Step S102: Respectively obtain the search keyword similarity W i between L i+1 and L i,i+1 ; set the first similarity threshold W1, if W i,i+1 ≥W1, determine that L i and L i+1 are associated search tracks generated by the user based on the same information needs, and L i and L i+1 belong to the same search process; if W i,i+1 <W1, set the second similarity threshold W2; obtain the set of browsing pages i generated by the user in the search track L where, respectively represent the 1st, 2nd,..., mth browsing pages generated in chronological order in the search track L i ; the browsing pages include the search result pages that appear after entering the search formula and the content pages that appear after clicking on a certain link within the search result page;

[0013] Step S103: Respectively extract keywords from each browsing page in the set of browsing pages to obtain the keyword set corresponding to the set of browsing pages Extract the search track Li Corresponding search keywords Keyword Set Each keyword in the sequence is combined to generate m combination keywords; if there is at least one combination keyword in the m combination keywords that matches the search trajectory L i+1 Corresponding keywords The similarity between them is greater than the second similarity threshold W2, and L i With L i+1 is the associated search trajectory generated by users based on the same information needs, L i With L i+1 Belong to the same search process, if there is no combination keyword in the m combination keywords and the search trajectory L i+1 Corresponding keywords The similarity between them is less than the second similarity threshold W2, and L i With L i+1 L is the independent search trajectory generated by users based on different information needs. i With L i+1 They belong to different search processes respectively;

[0014] Step S104: Classifying all search traces generated by the user within the time period S based on the search processes to which they belong, and obtaining search trace sets belonging to each search process;

[0015] Because in the actual information query process, users want to find knowledge in other fields that overlap with their own business. However, due to the lack of professional knowledge in this field and the strong standardization of various laws, regulations and other rules and regulations, users often make inaccurate and non-standard expressions when entering search terms. Therefore, the user's search process is to constantly explore the content they really want to find through repeated searches. This means that when users are looking for the same information needs, multiple search terms will appear, and because each search term is in the same information need, there is a content connection between them.

[0016] Further, step S200 includes:

[0017] Step S201: extract the search track sets contained in different search processes of the user respectively; record the search process H generated by the user based on the information demand A a The search trajectory set contained in is {a1,a2,…,a v}; where a1, a2, …, a v They represent the 1st, 2nd, …, vth search tracks generated by the user's 1st, 2nd, …, vth search expressions entered in the search box based on the query information requirement A; the corresponding search track sets {a1, a2, …, av}, obtain the browsing page set of each search trajectory in the browsing page set, calculate the average browsing time of the user corresponding to each search trajectory, and obtain the corresponding search trajectory set {a1, a2, …, a v Average browsing time of At the same time, capture the search trajectory set {a1, a2, …, a v The number of times C that the keyword similarity comparison is performed using the combined keywords in};

[0018] Step S202: If the average browsing time set If the number of average browsing times less than the average browsing time threshold is greater than the number threshold, the search process H is determined a Abnormal; if the average browsing time set The number of average browsing times less than the average browsing time threshold is less than or equal to the number threshold, and the time change index between every two adjacent search tracks is calculated:

[0019]

[0020] in, And meet

[0021]

[0022] Get the corresponding search trajectory set {a1, a2, …, a v}Time Variation Index Set

[0023] Step S203: Calculate the time variation index set The overall attenuation of:

[0024]

[0025] If C×σ<σ 阈值 , determine the search process H a Abnormal; if C×σ≥σ 阈值 , determine the search process H a normal;

[0026] When users are searching for information, they may keep searching because they do not enter the correct search formula. In this case, there is a complete lack of correct expression or almost no understanding of the content corresponding to the information they want to obtain. In this case, there is often a problem with the direction of the user's search. Timely discovery of abnormalities in the search process and providing users with auxiliary search can improve search efficiency and save time for users.

[0027] Furthermore, step S300 includes:

[0028] Step S301: Extract the user search trajectory set corresponding to the abnormal search process {L error1 ,L error2 ,…,L errorn}; where L error1 , L error2 ,…,L errorn They represent the 1st, 2nd, ..., nth search tracks generated by the user in chronological order during the abnormal search process; respectively extract the search track set {L error1 ,L error2 ,…,L errorn}, and the search keyword set {fL error1 ,fL error2 ,…,fL errorn}; where fL error1 ,fL error2 ,…,fL errorn They are respectively from the search trajectory L error1 ,L error2 ,…,L errorn The search keywords extracted from the search formula; the main keyword F1 is calculated:

[0029] F1=fL error1 ∩fL error2 ∩…∩fL errorn

[0030] Step S302: extracting the browsing page set in each search track in, They represent the 1st, 2nd, ..., mth browsing pages generated in chronological order in the xth search track in the abnormal search process; the browsing page set Each browsing page in the search results page is marked in two categories: the search results page that appears after the search formula is entered is marked in the first category, and the content page that appears after clicking a link in the search results page is marked in the second category;

[0031] Step S303: extract keywords from the browsing pages marked with the first category to obtain a keyword set in, They represent the keywords extracted from the 1st, 2nd, …, gth first-category tagged browsing pages respectively; and the keywords extracted from the browsing pages marked with the second category respectively, to obtain the keyword set in, Respectively represent the keywords extracted from the 1st, 2nd, ..., dth second-category tagged browsing pages; calculate the first subordinate keyword F2:

[0032]

[0033] Calculate the second subordinate keyword F3:

[0034]

[0035] Step S304: Performing a primary screening in the system based on the main keyword F1, taking the first subsidiary keyword F2 and the second subsidiary keyword F3 as secondary screening keywords of the main keyword F1 to complete further screening in the system, and obtaining the user's target demand information.

[0036] Furthermore, step S400 includes:

[0037] Step S401: The system collects the account tags of each logged-in user, including the department to which the user belongs, the type of police force, and the position of the user; captures the search records of the historical user account performing content searches in the system without abnormal search processes; and establishes a tag association between the search content corresponding to each search record and the user's account tag;

[0038] Step S402: Extract the account tag information of the corresponding user account when an abnormality occurs in the search process; search for each search content that has a tag association with the account tag in the system's historical search records; match each search content with the user target demand information obtained in step S300, and push and display each search content to the user in order of the content matching degree with the user target demand information from high to low.

[0039] In order to better implement the above method, an intelligent retrieval system based on the police knowledge base is also proposed. The system includes a search trajectory information capture module, a search process identification and judgment module, a search process abnormality troubleshooting module, a target demand information capture module, and a content push module;

[0040] The search track information capture module is used to capture and analyze the browsing record information generated by the user through the input search formula, and extract the search tracks generated by the user based on different information needs;

[0041] The search process identification and judgment module is used to identify and select the search tracks generated by users based on the same information needs, and to judge and summarize the search processes generated by users based on different information needs;

[0042] The search process abnormality troubleshooting module is used to conduct regular troubleshooting on the search tracks contained in different search processes generated by users based on different information needs; based on the regular troubleshooting results, the different search processes generated by users based on different information needs are respectively checked for abnormalities;

[0043] The target demand information capture module is used to receive the data in the search process abnormality troubleshooting module, and capture the user's target demand information from the browsing click record generated by inputting search-style browsing for the corresponding user account when the search process is abnormal;

[0044] The content push module is used to receive data from the target demand information capture module, extract the historical records of user accounts performing content searches in the system, and push content to users based on the data and the account tag of the corresponding user account when an abnormality occurs in the search process.

[0045] Further, the search trajectory information capturing module includes a trajectory search keyword extraction unit and a page keyword extraction unit;

[0046] The track search formula keyword extraction unit is used to extract the secondary search formula corresponding to a search track of the user to extract the search keywords; all the user browsing records generated according to the user inputting a search formula are corresponded to a search track generated by the user based on an input search formula;

[0047] A page keyword extraction unit is used to extract keywords from each browsing page generated by the user in chronological order in each search track; the browsing page includes the search result page that appears after the search formula is entered, and the content page that appears after clicking a link in the search result page;

[0048] Furthermore, the search process identification and judgment module includes a similarity comparison unit, a keyword combination management unit, and a search process judgment and identification unit;

[0049] A similarity comparison unit is used to compare the similarity of search keywords between two temporally adjacent search trajectories;

[0050] The keyword combination management unit is used to receive the data in the similarity comparison unit, and sequentially combine the search keywords of the search track located before the time sequence in two temporally adjacent search tracks with the keywords of each browsing page in the search track located after the time sequence to obtain a plurality of combined keywords;

[0051] The search process judgment and identification unit is used to receive data from the similarity comparison unit and the keyword combination management unit, judge and identify the corresponding information requirements between the search tracks based on the data, and summarize the search tracks generated based on the same information requirements into the same search process.

[0052] Furthermore, the search process anomaly troubleshooting module includes a time variation index calculation unit, an overall attenuation calculation unit, and an anomaly identification unit;

[0053] A time variation index calculation unit, used to calculate the time variation index between every two temporally adjacent search trajectories;

[0054] An overall attenuation degree calculation unit, used to receive the data in the time variation index calculation unit, and calculate the overall attenuation degree for each search process based on the data;

[0055] The abnormality identification unit is used to receive the data in the overall attenuation calculation unit, and based on the data, identify abnormalities in the user's search process.

[0056] Compared with the prior art, the beneficial effects achieved by the present invention are: the present invention can establish a "mobile encyclopedia" and can realize keyword retrieval similar to "Baidu" through the system; the present invention summarizes the search trajectories generated by users based on querying the same demand information into corresponding search processes through various search trajectories generated by users based on various search formulas input by users in the system, and performs abnormal monitoring and judgment on the search process of users. When the search process is abnormal when the user searches for a certain information demand, the user is helped to retrieve the search call records of materials with the same account tag as the user in the history, assisting the user to achieve the search purpose as soon as possible, solving the problem of low search efficiency caused by users' precise expressions and non-standard descriptions, improving search accuracy, and improving the work efficiency of police officers. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0058] Figure 1 It is a flow chart of an intelligent retrieval and analysis method based on a police knowledge base of the present invention;

[0059] Figure 2 It is a structural schematic diagram of an intelligent retrieval and analysis system based on a police knowledge base of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] See also Figure 1-Figure 2 The present invention provides a technical solution: an intelligent retrieval and analysis method based on a police knowledge base, the method comprising:

[0062] Step S100: The system captures and analyzes the search trajectories generated by the user based on the same information needs according to the browsing and flipping record information generated by the user through inputting search terms, and summarizes the search trajectories of the search processes generated by the user based on different information needs respectively;

[0063] Among them, the process of calculating and capturing the search process of each user in step S100 based on the flipping record characteristics generated by each user through inputting search terms for browsing includes:

[0064] Step S101: Correspond all the user browsing and flipping records generated by the user inputting a search term once to a search trajectory generated by the user based on one input search term; Obtain the set of search trajectories {L1, L2, …, L n} generated by the user within the time period S; where L1, L2, …, L n respectively represent the 1st, 2nd, …, nth search trajectories generated by the user in chronological order within the time period S; Extract search keywords from the input search terms of each search trajectory in the user search trajectory set {L1, L2, …, L n} respectively to obtain the keyword set corresponding to the search trajectory set Among them, respectively represent the search keywords extracted from the input search terms of L1, L2, …, L n ; where the number of search keywords extracted according to each input search term is greater than or equal to 1;

[0065] Step S102: Obtain the search keyword similarity W i between L i+1 and L i,i+1 respectively; Set the first similarity threshold W1, if W i,i+1 ≥W1, determine that L i and L i+1 are associated search trajectories generated by the user based on the same information needs, and L i and L i+1 belong to the same search process; if W i,i+1 <W1, set the second similarity threshold W2; Obtain the set of browsing pages generated by the user in the search trajectory L i Among them, respectively represent the 1st, 2nd, …, mth browsing pages generated in chronological order in the search trajectory L i ; The browsing pages include the search result pages that appear after inputting the search term and the content pages that appear after clicking on a certain link in the search result page;

[0066] Step S103: Respectively for the set of browsing pages Extract keywords from each browsing page to get the keyword set corresponding to the browsing page set The search trajectory L i Corresponding search keywords Keyword Set Each keyword in the sequence is combined to generate m combination keywords; if there is at least one combination keyword in the m combination keywords that matches the search trajectory L i+1 Corresponding keywords The similarity between them is greater than the second similarity threshold W2, and L i With L i+1 is the associated search trajectory generated by users based on the same information needs, L i With L i+1 Belong to the same search process, if there is no combination keyword in the m combination keywords and the search trajectory L i+1 Corresponding keywords The similarity between them is less than the second similarity threshold W2, and L i With L i+1 L is the independent search trajectory generated by users based on different information needs. i With L i+1 They belong to different search processes respectively;

[0067] Step S104: Classifying all search traces generated by the user within the time period S based on the search processes to which they belong, and obtaining search trace sets belonging to each search process;

[0068] Step S200: in different search processes generated by users based on different information needs, respectively, each search track included in each search process is checked for regularity; based on the regularity checking of each search track included in each search process, respectively, different search processes generated by users based on different information needs are checked for abnormality;

[0069] Wherein, step S200 includes:

[0070] Step S201: extract the search track sets contained in different search processes of the user respectively; record the search process H generated by the user based on the information demand A a The search trajectory set contained in is {a1,a2,…,a v}; where a1, a2, …, a v They represent the 1st, 2nd, …, vth search tracks generated by the user's 1st, 2nd, …, vth search expressions entered in the search box based on the query information requirement A; the corresponding search track sets {a1, a2, …, a v}, obtain the browsing page set of each search trajectory in the browsing page set, calculate the average browsing time of the user corresponding to each search trajectory, and obtain the corresponding search trajectory set {a1, a2, …, a v Average browsing time of At the same time, capture the search trajectory set {a1, a2, …, a v The number of times C that the keyword similarity comparison is performed using the combined keywords in};

[0071] For example, a search process includes search track 1, search track 2, search track 3, and search track 4; the search track 1 and search track 2, and the search track 3 and search track 4 all satisfy that the search keyword similarity is greater than or equal to the first similarity threshold W1=0.8;

[0072] The search keyword similarity between search track 2 and search track 3 is less than the first similarity threshold value W1=0.8, but the search keyword of search track 2 is combined with the keyword of a browsing page existing in search track 2 to form a combination keyword 1, and the keyword similarity between the combination keyword 1 and the search keyword of search track 3 is greater than the second similarity threshold value W2=0.68, and it is determined that the number of times the combination keyword is used to perform keyword similarity comparison in this search process is C=1;

[0073] Step S202: If the average browsing time set If the number of average browsing times less than the average browsing time threshold is greater than the number threshold, the search process H is determined a Abnormal; if the average browsing time set The number of average browsing times less than the average browsing time threshold is less than or equal to the number threshold, and the time change index between every two adjacent search tracks is calculated:

[0074]

[0075] in, And meet

[0076]

[0077] Get the corresponding search trajectory set {a1, a2, …, a v}Time Variation Index Set

[0078] Step S203: Calculate the time variation index set The overall attenuation of:

[0079]

[0080] If C×σ<σ 阈值, determine the search process H a Abnormal; if C×σ≥σ 阈值 , determine the search process H a normal

[0081] Step S300: When it is determined that the user's search process is abnormal, the system is triggered to capture the user's target demand information based on the browsing click records generated by the user's input search browsing;

[0082] Wherein, step S300 includes:

[0083] Step S301: Extract the user search trajectory set corresponding to the abnormal search process {L error1 ,L error2 ,…,L errorn}; where L error1 , L error2 ,…,L errorn They represent the 1st, 2nd, ..., nth search tracks generated by the user in chronological order during the abnormal search process; respectively extract the search track set {L error1 ,L error2 ,…,L errorn}, and the search keyword set {fL error1 ,fL error2 ,…,fL errorn}; where fL error1 ,fL error2 ,…,fL errorn They are respectively from the search trajectory L error1 ,L error2 ,…,L errorn The search keywords extracted from the search formula; the main keyword F1 is calculated:

[0084] F1=fL error1 ∩fL error2 ∩…∩fL errorn

[0085] Step S302: extracting the browsing page set in each search track in, They represent the 1st, 2nd, ..., mth browsing pages generated in chronological order in the xth search track in the abnormal search process; the browsing page set Each browsing page in the search results page is marked in two categories: the search results page that appears after the search formula is entered is marked in the first category, and the content page that appears after clicking a link in the search results page is marked in the second category;

[0086] Step S303: extract keywords from the browsing pages marked with the first category to obtain a keyword set in, They represent the keywords extracted from the 1st, 2nd, …, gth first-category tagged browsing pages respectively; and the keywords extracted from the browsing pages marked with the second category respectively, to obtain the keyword set in, Respectively represent the keywords extracted from the 1st, 2nd, ..., dth second-category tagged browsing pages; calculate the first subordinate keyword F2:

[0087]

[0088] Calculate the second subordinate keyword F3:

[0089]

[0090] Step S304: Perform a primary screening in the system based on the main keyword F1, and use the first subsidiary keyword F2 and the second subsidiary keyword F3 as secondary screening keywords of the main keyword F1 to complete further screening in the system to obtain the user's target demand information

[0091] Step S400: Based on the captured user's target demand information and in combination with the user's account tag, intelligent content push is implemented to the corresponding user account when an abnormality occurs in the search process;

[0092] Wherein, step S400 includes:

[0093] Step S401: The system collects the account tags of each logged-in user, including the department to which the user belongs, the type of police force, and the position of the user; captures the search records of the historical user account performing content searches in the system without abnormal search processes; and establishes a tag association between the search content corresponding to each search record and the user's account tag;

[0094] Step S402: Extract the account tag information of the corresponding user account when an abnormality occurs in the search process; search for each search content that has a tag association with the account tag in the system's historical search records; match each search content with the user target demand information obtained in step S300, and push and display each search content to the user in order of the content matching degree with the user target demand information from high to low.

[0095] In order to better implement the above method, an intelligent retrieval system based on the police knowledge base is also proposed. The system includes a search trajectory information capture module, a search process identification and judgment module, a search process abnormality troubleshooting module, a target demand information capture module, and a content push module;

[0096] The search track information capture module is used to capture and analyze the browsing record information generated by the user through the input search formula, and extract the search tracks generated by the user based on different information needs;

[0097] The search process identification and judgment module is used to identify and select the search tracks generated by users based on the same information needs, and to judge and summarize the search processes generated by users based on different information needs;

[0098] Among them, the search process identification and judgment module includes a similarity comparison unit, a keyword combination management unit, and a search process judgment and identification unit;

[0099] A similarity comparison unit is used to compare the similarity of search keywords between two temporally adjacent search trajectories;

[0100] The keyword combination management unit is used to receive the data in the similarity comparison unit, and sequentially combine the search keywords of the search track located before the time sequence in two temporally adjacent search tracks with the keywords of each browsing page in the search track located after the time sequence to obtain a plurality of combined keywords;

[0101] A search process judgment and identification unit, used to receive data from the similarity comparison unit and the keyword combination management unit, judge and identify the corresponding information requirements between the search tracks based on the data, and classify the search tracks generated based on the same information requirements into the same search process;

[0102] Among them, the search trajectory information capture module includes a trajectory search keyword extraction unit and a page keyword extraction unit;

[0103] The track search formula keyword extraction unit is used to extract the secondary search formula corresponding to a search track of the user to extract the search keywords; all the user browsing records generated according to the user inputting a search formula are corresponded to a search track generated by the user based on an input search formula;

[0104] A page keyword extraction unit is used to extract keywords from each browsing page generated by the user in chronological order in each search track; the browsing page includes the search result page that appears after the search formula is entered, and the content page that appears after clicking a link in the search result page;

[0105] The search process abnormality troubleshooting module is used to conduct regular troubleshooting on the search tracks contained in different search processes generated by users based on different information needs; based on the regular troubleshooting results, the different search processes generated by users based on different information needs are respectively checked for abnormalities;

[0106] Among them, the search process abnormality troubleshooting module includes a time change index calculation unit, an overall attenuation calculation unit, and an abnormality identification unit;

[0107] A time variation index calculation unit, used to calculate the time variation index between every two temporally adjacent search trajectories;

[0108] An overall attenuation degree calculation unit, used to receive the data in the time variation index calculation unit, and calculate the overall attenuation degree for each search process based on the data;

[0109] An abnormality identification unit, used to receive data from the overall attenuation calculation unit, and identify abnormalities in the user's search process based on the data;

[0110] The target demand information capture module is used to receive the data in the search process abnormality troubleshooting module, and capture the user's target demand information from the browsing click record generated by inputting search-style browsing for the corresponding user account when the search process is abnormal;

[0111] The content push module is used to receive data from the target demand information capture module, extract the historical records of user accounts performing content searches in the system, and push content to users based on the data and the account tag of the corresponding user account when an abnormality occurs in the search process.

[0112] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0113] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. 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 intelligent retrieval and analysis method based on a police knowledge base, characterized in that: The method comprises: Step S100: Based on the browsing record information generated by the user inputting the search formula, the system captures and analyzes the search tracks generated by the user based on the same information needs, and summarizes the search tracks of the user based on different information needs; Step S200: in different search processes generated by users based on different information needs, respectively, each search track included in each search process is checked for regularity; based on the regularity checking of each search track included in each search process, respectively, different search processes generated by users based on different information needs are checked for abnormality; Step S300: when it is determined that the user's search process is abnormal, the system is triggered to crawl the user's target demand information based on the browsing click records generated by the user inputting the search-style browsing; Step S400: Based on the captured user's target demand information and in combination with the user's account tag, intelligent content push is implemented to the corresponding user account when an abnormality occurs in the search process; Step S200 includes: Step S201: extract the search track sets contained in different search processes of the user respectively; record the search process H generated by the user based on the information demand A a The search trajectory set contained in is {a1,a2,…,a v }; where a1, a2, …, a v They represent the 1st, 2nd, …, vth search tracks generated by the user's 1st, 2nd, …, vth search expressions entered in the search box based on the query information requirement A; the corresponding search track sets {a1, a2, …, a v }, obtain the browsing page set of each search trajectory in the browsing page set, calculate the average browsing time of the user corresponding to each search trajectory, and obtain the corresponding search trajectory set {a1, a2, …, a v Average browsing time of At the same time, capture the search trajectory set {a1, a2, …, a v The number of times C that the keyword similarity comparison is performed using the combined keywords in}; Step S202: If the average browsing time set If the number of average browsing times less than the average browsing time threshold is greater than the number threshold, the search process H is determined a Abnormal; if the average browsing time set The number of average browsing times less than the average browsing time threshold is less than or equal to the number threshold, and the time change index between every two adjacent search tracks is calculated: in, And meet Get the corresponding search trajectory set {a1, a2, …, a v }Time Variation Index Set Step S203: Calculate the time variation index set The overall attenuation of: If C×σ<σ 阈值 , determine the search process H a Abnormal; if C×σ≥σ 阈值 , determine the search process H a normal; The step S400 includes: Step S401: The system collects the account tags of each logged-in user, including the department to which the user belongs, the type of police force, and the position of the user; captures the search records of the historical user account performing content searches in the system without abnormal search processes; and establishes a tag association between the search content corresponding to each search record and the user's account tag; Step S402: extract the account tag information of the corresponding user account when an abnormality occurs in the search process; search for each search content that has a tag association with the account tag in the system's historical search records; match each search content with the user target demand information obtained in step S300, and push and display each search content to the user in order of the content matching degree with the user target demand information from high to low.

2. According to claim 1, the intelligent retrieval and analysis method based on the police knowledge base is characterized in that: The process of calculating and capturing the search progress of each user based on the browsing record features generated by each user through inputting search-style browsing in step S100 includes: Step S101: All user browsing records generated by a user inputting a search formula are mapped to a search track generated by the user based on an input search formula; a search track set {L1, L2, ..., L n }; where L1, L2, …, L n They represent the 1st, 2nd, …, nth search trajectories generated by the user in time period S in chronological order; the user search trajectory set {L1, L2, …, L n } Extract the search keywords from the input search formula of each search track in the search path, and get the keyword set corresponding to the search track set in, Respectively represent from L1, L2, ..., L n The search keywords extracted by the search formula input in; wherein the number of search keywords extracted according to each search formula input is greater than or equal to 1; Step S102: Obtain L respectively i and L i+1 The search keyword similarity W between them i,i+1 ; Set the first similarity threshold W1. If W i,i+1 ≥W1, it is determined that L i and L i+1 are associated search trajectories generated by the user based on the same information needs, and L i and L i+1 belong to the same search process; If W i,i+1 <W1, set the second similarity threshold W2; Obtain the set of browsing pages generated by the user in the search trajectory L i Among them, where respectively represent the 1st, 2nd,..., mth browsing pages generated in chronological order in the search trajectory L i ; The browsing pages include the search result pages that appear after entering the search formula, and the content pages that appear after clicking on a certain link within the search result page; Step S103: Browse the page sets separately Extract keywords from each browsing page to get the keyword set corresponding to the browsing page set The search trajectory L i Corresponding search keywords Keyword Set Each keyword in the sequence is combined to generate m combination keywords; if there is at least one combination keyword in the m combination keywords that matches the search trajectory L i+1 Corresponding keywords The similarity between them is greater than the second similarity threshold W2, and L i With L i+1 is the associated search trajectory generated by users based on the same information needs, L i With L i+1 Belong to the same search process, if there is no combination keyword in the m combination keywords and the search trajectory L i+1 Corresponding keywords The similarity between them is less than the second similarity threshold W2, and L i With L i+1 L is the independent search trajectory generated by users based on different information needs. i With L i+1 They belong to different search processes respectively; Step S104: All search traces generated by the user within the time period S are classified based on the search processes to which they belong, and a search trace set belonging to each search process is obtained.

3. According to claim 2, the intelligent retrieval and analysis method based on the police knowledge base is characterized in that: The step S300 includes: Step S301: Extract the user search trajectory set corresponding to the abnormal search process {L error1 ,L error2 ,…,L errorn }; where L error1 , L error2 ,…,L errorn They represent the 1st, 2nd, ..., nth search tracks generated by the user in chronological order during the abnormal search process; respectively extract the search track set {L error1 ,L error2 ,…,L errorn }, and the search keyword set {fL error1 ,fL error2 ,…,fL errorn }; where fL error1 ,fL error2 ,…,fL errorn They are respectively from the search trajectory L error1 ,L error2 ,…,L errorn The search keywords extracted from the search formula; the main keyword F1 is calculated: F1=fL error1 ∩fL error2 ∩…∩fL errorn Step S302: extracting the browsing page set in each search track in, They respectively represent the 1st, 2nd, ..., mth browsing pages generated in chronological order in the xth search track in the abnormal search process; the browsing page set Each browsing page in the search results page is marked in two categories: the search results page that appears after the search formula is entered is marked in the first category, and the content page that appears after clicking a link in the search results page is marked in the second category; Step S303: extract keywords from the browsing pages marked with the first category to obtain a keyword set in, They represent the keywords extracted from the 1st, 2nd, …, gth first-category tagged browsing pages respectively; and the keywords extracted from the browsing pages marked with the second category respectively, to obtain the keyword set in, Respectively represent the keywords extracted from the 1st, 2nd, ..., dth second-category tagged browsing pages; calculate the first subordinate keyword F2: Calculate the second subordinate keyword F3: Step S304: Performing a primary screening in the system based on the main keyword F1, taking the first subsidiary keyword F2 and the second subsidiary keyword F3 as secondary screening keywords of the main keyword F1 to complete further screening in the system, and obtaining the user's target demand information.

4. An intelligent retrieval system based on a police knowledge base applied to the intelligent retrieval and analysis method based on a police knowledge base according to any one of claims 1 to 3, characterized in that: The system includes a search trajectory information capture module, a search process identification and judgment module, a search process abnormality troubleshooting module, a target demand information capture module, and a content push module; The search track information capturing module is used to capture and analyze the browsing record information generated by the user inputting the search formula, and respectively extract the search tracks generated by the user based on different information needs; The search process identification and judgment module is used to identify and select the search tracks generated by users based on the same information needs, and to judge and summarize the search processes generated by users based on different information needs; The search process abnormality troubleshooting module is used to perform regularity troubleshooting on the search tracks contained in different search processes generated by users based on different information needs; based on the regularity troubleshooting results, perform abnormality troubleshooting on different search processes generated by users based on different information needs respectively; The target demand information capture module is used to receive the data in the search process abnormality troubleshooting module, and capture the user's target demand information from the browsing click record generated by inputting search-style browsing for the corresponding user account when the search process is abnormal; The content push module is used to receive data from the target demand information capture module, extract historical records of content searches by user accounts in the system, and push content to users by combining the data with the account tag of the corresponding user account when an abnormality occurs in the search process.

5. According to claim 4, the intelligent retrieval system based on the police knowledge base is characterized in that: The search track information capture module includes a track search keyword extraction unit and a page keyword extraction unit; The track search formula keyword extraction unit is used to extract the secondary search formula corresponding to a search track of the user to extract search keywords; all user browsing records generated according to the user inputting a search formula are corresponded to a search track generated by the user based on an input search formula; The page keyword extraction unit is used to extract keywords from the browsing pages generated by the user in chronological order in each search track; the browsing pages include the search result page that appears after entering the search formula and the content page that appears after clicking a link in the search result page.

6. The intelligent retrieval system based on the police knowledge base according to claim 4 is characterized in that: The search process identification and judgment module includes a similarity comparison unit, a keyword combination management unit, and a search process judgment and identification unit; The similarity comparison unit is used to compare the similarity of search keywords between two temporally adjacent search tracks; The keyword combination management unit is used to receive the data in the similarity comparison unit, and sequentially combine the search keywords of the search track located before the time sequence in two temporally adjacent search tracks with the keywords of each browsing page in the search track located after the time sequence to obtain a plurality of combined keywords; The search process judgment and identification unit is used to receive data from the similarity comparison unit and the keyword combination management unit, judge and identify the corresponding information requirements between the search tracks based on the data, and summarize the search tracks generated based on the same information requirements into the same search process.

7. The intelligent retrieval system based on the police knowledge base according to claim 4 is characterized in that: The search process abnormality troubleshooting module includes a time change index calculation unit, an overall attenuation calculation unit, and an abnormality identification unit; The time variation index calculation unit is used to calculate the time variation index between every two temporally adjacent search trajectories; The overall attenuation degree calculation unit is used to receive the data in the time change index calculation unit, and calculate the overall attenuation degree for each search process based on the data; The abnormality identification unit is used to receive the data in the overall attenuation degree calculation unit, and based on the data, identify abnormalities in the user's search process.

Images