A blockchain-assisted electronic case management system

Abstract

The present invention discloses a blockchain-assisted electronic case management system, which relates to the technical field of auxiliary case management. The system realizes all-round management and intelligent processing of case evidence through the organic combination of modules such as a collection unit, an identification unit, an evaluation unit, an early warning unit, a strategy unit and a blockchain storage unit. The collection unit is responsible for collecting case information and corresponding evidence data, and cleaning and preprocessing them; the identification unit identifies and analyzes different forms of evidence through text recognition, video processing and voice recognition technology; the evaluation unit evaluates the recognition result and generates a corresponding evaluation result; the early warning unit issues an early warning instruction when the evaluation result does not meet the standard; the strategy unit generates and executes the corresponding strategy according to the evaluation result; the blockchain storage unit is responsible for archiving qualified case evidence data on the blockchain to ensure the security and non-tamperability of the data.

Description

Technical Field

[0001] The present invention relates to the technical field of auxiliary case file management, and specifically to a blockchain-assisted electronic case file management system. Background Art

[0002] With the continuous development of the information age, the electronic case management system has become an indispensable part of the judicial field. The traditional case management method has been unable to meet the needs of reviewing and managing evidence, especially with the diversification and increase in the number of evidence forms, such as voice, video and written evidence, which puts higher requirements on the management system. Therefore, it is imperative to use advanced technical means to build a blockchain-assisted electronic case management system.

[0003] In the traditional electronic case management system, although the case evidence can be stored and retrieved, the intelligent analysis and management capabilities of unstructured data such as voice and video are relatively weak. To address this problem, blockchain technology is introduced as an auxiliary means to effectively solve the problems of information security and data immutability, and improve the credibility and stability of the system. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a blockchain-assisted electronic case management system to solve the problem of assisting the analysis and review of voice, video and written evidence mentioned in the background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a blockchain-assisted electronic case management system, including a collection unit, a first identification unit, a second identification unit, a third identification unit, an evaluation unit, an early warning unit, a strategy unit and a blockchain storage unit;

[0006] The collection unit is used to collect case file information and corresponding case file evidence data to form a case file set, and clean and pre-process the case file set to remove invalid information and duplicate data, and then convert the case file set into a unified standard format; the case file evidence data includes written materials, video recording materials, audio recording materials and digital evidence;

[0007] The first recognition unit is used to extract the written material, identify the text content of the extracted written material through text recognition technology, identify the number of events in the quantified written evidence, compare the number of events quantified in the written evidence with the number of events in the original document of the litigation document, and obtain the written evidence completeness index SZs; when the written evidence completeness index SZs is lower than a first threshold, the early warning unit generates and issues a first early warning instruction; when the written evidence completeness index SZs is higher than or equal to the first threshold, a first qualified mark is generated;

[0008] The second identification unit is used to extract video recording materials, identify metadata of video files, obtain resolution fbl, contrast dbd and frame number zs, compare the continuity method between video frames for verification, obtain video integrity wzd and clarity index Qx, and use video processing technology to identify case targets, vehicles and event behaviors in the video, obtain the number of event occurrences sjcxcs, duration cxsj and the position interval difference jgcz, and construct an event index SJzs; and extract key frames or important clips in the video, and generate a video summary, and associate the video integrity wzd, clarity index Qx and event index SJzs, obtain the video comprehensive coefficient Zh, and perform a first mark in the video summary;

[0009] The third recognition unit is used to extract the recording material, and is used to use the PESQ recognition tool to identify and obtain the voice duration yycs, the average voice amplitude zf and the signal-to-noise ratio xzb, and after converting the recording material into text, extract the event keywords, obtain the event keyword times sjcs, and construct the recording quality coefficient Zlx;

[0010] The evaluation unit is used to evaluate the video comprehensive coefficient Zh and the recording quality coefficient Zlx respectively, obtain the second evaluation result and the third evaluation result, and the second evaluation result and the third evaluation result;

[0011] The strategy unit is used to generate and execute corresponding strategies according to the first evaluation result, the second evaluation result and the third evaluation result, and when the first evaluation result, the second evaluation result and the third evaluation result are all qualified, the qualified case file evidence data is archived in the corresponding case file through the blockchain storage unit.

[0012] Preferably, the first recognition unit includes a semantic analysis module and a first calculation module;

[0013] The semantic analysis module is used to perform semantic analysis on the written materials using natural language processing technology, and to identify the number of events in the quantified written evidence through keyword extraction, obtain the number of events sz1, and calculate the average interval time pjjg between events in the case file; compare the number of events sz1 quantified in the written evidence with the number of events Bz in the original document of the litigation document, and identify the repeated event patterns and event groups with strong correlation in the written materials, and obtain the associated event confidence C (A→B) of event A and event B through the following formula:

[0014]

[0015] In the formula, s(A∪B) represents the number of times event A and event B appear at the same time; s(A) represents the number of times event A appears alone; and C(A→B) higher than the preset correlation threshold is classified as a correlated event, and the number of correlated events xglsj is obtained by counting;

[0016] The first calculation module is used to calculate the written evidence integrity index SZs by the following formula after dimensionless processing of the event number sz1, the event number Bz of the original litigation document, the number of associated events xglsj and the average interval time pjjg:

[0017]

[0018] Wherein, 0<w1<1, 0<w2<1, 0<w3<1, and w1+w2+w3=1, w1, w2 and w3 are weights, and their specific values ​​are adjusted and set by the user; A is a first constant correction coefficient; the first recognition unit includes a semantic analysis module and a first calculation module;

[0019] The semantic analysis module is used to perform semantic analysis on the written materials using natural language processing technology, and to identify the number of events in the quantified written evidence through keyword extraction, obtain the number of events sz1, and calculate the average interval time pjjg between events in the case file; compare the number of events sz1 quantified in the written evidence with the number of events Bz in the original document of the litigation document, and identify the repeated event patterns and event groups with strong correlation in the written materials, and obtain the associated event confidence C (A→B) of event A and event B through the following formula:

[0020]

[0021] In the formula, s(A∪B) represents the number of times event A and event B appear at the same time; s(A) represents the number of times event A appears alone; and C(A→B) higher than the preset correlation threshold is classified as a correlated event, and the number of correlated events xglsj is obtained by counting;

[0022] The first calculation module is used to calculate the written evidence integrity index SZs by the following formula after dimensionless processing of the event number sz1, the event number Bz of the original litigation document, the number of associated events xglsj and the average interval time pjjg:

[0023]

[0024] Among them, 0<w1<1, 0<w2<1, 0<w3<1, and w1+w2+w3=1, w1, w2 and w3 are weights, and their specific values ​​are adjusted and set by the user; A is the first constant correction coefficient.

[0025] Preferably, the first identification unit further includes a first judgment unit, which is used to compare and judge the written evidence integrity index SZs with a first threshold value to obtain a first evaluation result, including:

[0026] When the written evidence integrity index SZs is less than the first threshold, it means that the written evidence is unqualified, and the warning unit generates and issues a first warning instruction;

[0027] When the written evidence completeness index SZs ≥ the first threshold, a first qualified mark is generated.

[0028] Preferably, the second recognition unit includes a second recognition module and an event behavior recognition module;

[0029] The second recognition module is used to extract video recording materials, identify the metadata of the video file through the FFmpeg video processing tool, and verify the continuity method between video frames, and generate the video integrity wzd through the following formula:

[0030]

[0031] In the formula, zjp represents the standard deviation of the time interval between video frames, and α represents the adjustment parameter, which is used to adjust the influence of the standard deviation on the video integrity. The smaller the standard deviation zjp is, the more stable the time interval between video frames is, and the better the continuity of the video is.

[0032] The clarity index Qx is obtained by extracting the resolution fbl, the contrast dbd and the number of frames zs, and after dimensionless processing, the clarity index Qx is generated by the following formula:

[0033] Qx=w4*fbl+w5*dbd*w6*zs+B

[0034] Wherein, 0<w4<1, 0<w5<1, 0<w6<1, and w4+w5+w6=1, w4, w5 and w6 are weights, and their specific values ​​are adjusted and set by the user; B is the second constant correction coefficient;

[0035] The event behavior recognition module is used to use the target detection and tracking algorithms in computer vision technology, including YOLO and FasterR-CNN, to identify the case target and the target positioning position of the vehicle in the video, and track the movement trajectory of the target in the video. Based on the target detection and tracking results, the traffic accident behavior and the gathering behavior of people are identified and counted as event features;

[0036] The number of times the event features are identified is counted, and the number of event occurrences sjcxcs, duration cxsj, and location interval difference jgcz are obtained. After dimensionless processing, the event index SJzs is generated using the following formula:

[0037] SJzs=w7*sjcxcs+w8*cxsj*w9*jgcz

[0038] Among them, 0<w7<1, 0<w8<1, 0<w9<1, and w7+w8+w9=1, w7, w8 and w9 are weights, and their specific values ​​are adjusted and set by the user.

[0039] Preferably, the second identification unit further includes an association module and a marking module;

[0040] The association module is used to generate a video comprehensive coefficient Zh by combining the video completeness wzd, the clarity index Qx and the event index SJzs through the following association consensus:

[0041] Zh=x*wzd+y*Qx*z*SJzs

[0042] Wherein, x, y and z represent the preset proportional coefficients of the video completeness wzd, the clarity index Qx and the event index SJzs, respectively, 0.25≤x≤0.33, 0.20≤y≤0.25, 0.35≤z≤0.42, and 0.80≤x+y+z≤1.0;

[0043] The marking module is used to extract key frames or important segments in the video and generate a video summary, and to first mark the video comprehensive coefficient Zh on the first page of the video summary.

[0044] Preferably, the recording quality coefficient Zlx is obtained by extracting the speech duration yycs, the average speech amplitude zf, the signal-to-noise ratio xzb and the number of event keywords sjcs, and after dimensionless processing, the recording quality coefficient Zlx is generated by the following formula:

[0045]

[0046] Among them, bz2 represents the preset standard speech duration threshold, bz3 represents the preset speech average amplitude threshold, bz4 represents the preset signal-to-noise ratio threshold, bz5 represents the preset time keyword number threshold, q, w, e and r represent weight values, and 0<q<1, 0<w<1, 0<e<1, 0<r<1 and q+w+e+r≤1.0; its specific value is adjusted and set by the user; C is the third constant correction coefficient.

[0047] Preferably, the evaluation unit comprises a first evaluation module and a second evaluation module;

[0048] The first evaluation module is used to compare the video comprehensive coefficient Zh with the second threshold value to obtain a second evaluation result, including:

[0049] When the video comprehensive coefficient Zh ≥ the second threshold value, it indicates that the video quality is qualified, and a second qualified mark is generated;

[0050] When the video comprehensive coefficient Zh is less than the second threshold value, indicating that the video is unqualified, the warning unit generates and issues a second warning instruction;

[0051] The second evaluation module is used to compare the recording quality coefficient Zlx with the third threshold to obtain a third evaluation result, including:

[0052] When the recording quality coefficient Zlx ≥ the third threshold, it indicates that the recording quality is qualified, and a third qualified mark is generated;

[0053] When the recording quality coefficient Zlx is less than the third threshold value, indicating that the recording quality is unqualified, the warning unit generates and issues a third warning instruction.

[0054] Preferably, the strategy unit is used to generate and execute corresponding strategies according to the first evaluation result, the second evaluation result and the third evaluation result, including:

[0055] Generate a first strategy according to the first warning instruction, the first strategy including: re-examining the written evidence to find out possible defects or incompleteness, and correcting or supplementing them as soon as possible; and sending a reminder notice to the relevant responsible person to remind him to pay attention to the integrity of the written evidence and urge him to take action to correct it as soon as possible;

[0056] Generate a second strategy according to the second warning instruction, the second strategy comprising: re-recording the video, adjusting the environmental optical fiber, camera settings and shooting angle during the recording process, and using video processing software to correct and optimize the recorded video, clipping and editing the recorded video, deleting non-event parts, retaining the event content, and adding annotations and descriptions;

[0057] A third strategy is generated based on the third warning instruction, and the third strategy includes: re-recording the audio, installing noise jammers in the recording equipment and environment, and using audio processing software to post-process the recorded audio to perform noise reduction, equalization, gain adjustment and processing to optimize the quality and clarity of the audio, and marking the event description or testimony content in the corrected audio.

[0058] Preferably, it also includes a fourth identification unit, which retrieves and extracts emails under specified conditions from the digital evidence through an email crawler, and identifies information in the email text through natural language processing technology and text analysis tools, obtains the email sending frequency fspl, the size of the sent attachment fjdx, the total number of emails sent and received zsl, the email reply rate hfl and the number of email citations yycs, and generates the email interaction behavior index Xwzs through the following formula after dimensionless processing:

[0059]

[0060] Wherein, a, k, d, b and h represent the preset proportional coefficients of the email sending frequency fspl, the total number of emails sent and received zsl, the email reply rate hfl, the number of email citations yycs and the size of the sent attachments fjdx, respectively, and 0.17≤a≤0.21, 0.10≤k≤0.12, 0.15≤d≤0.19, 0.16≤b≤0.18, 0.22≤h≤0.30, and 0.80≤a+k+d+b+h≤1.0; D is the fourth constant correction coefficient;

[0061] Comparing the email interaction behavior index Xwzs with the fourth threshold value to obtain a fourth evaluation result, including: when the email interaction behavior index Xwzs ≥ the fourth threshold value, it indicates that the email interaction behavior activity value is qualified, and a more in-depth investigation and analysis is conducted on the case files with relatively active email interaction behaviors; and the case files are reviewed as the first priority;

[0062] When the email interaction behavior index Xwzs is less than the fourth threshold, it indicates that the email interaction behavior is relatively unqualified and is ranked as the second priority for attention and review.

[0063] Preferably, it also includes an intelligent access unit and an intelligent log monitoring unit;

[0064] The intelligent access unit is used to monitor the access behavior of system users, including login, operation and file access behavior, identify abnormal behavior and potential threats through real-time monitoring and analysis of access logs, and take corresponding measures in time; the intelligent access unit interacts with the evaluation unit and the early warning unit to generate corresponding strategies and early warning instructions according to the monitoring results;

[0065] The intelligent log monitoring unit is used to monitor and analyze the system's log data in real time, including user operation logs, system logs and security logs. It detects abnormal events and potential threats through technologies such as machine learning algorithms and rule engines, and responds in real time.

[0066] The present invention provides a blockchain-assisted electronic case management system, which has the following beneficial effects:

[0067] (1) This blockchain-assisted electronic case management system realizes the all-round management and intelligent processing of case evidence through the organic combination of modules such as the collection unit, identification unit, evaluation unit, early warning unit, strategy unit and blockchain storage unit. The collection unit is responsible for collecting case information and corresponding evidence data, and cleaning and preprocessing them; the identification unit uses technologies such as text recognition, video processing and voice recognition to identify and analyze different forms of evidence; the evaluation unit evaluates the recognition results and generates corresponding evaluation results; the early warning unit issues early warning instructions when the evaluation results do not meet the standards; the strategy unit generates and executes corresponding strategies based on the evaluation results; the blockchain storage unit is responsible for archiving qualified case evidence data on the blockchain to ensure the security and non-tamperability of the data.

[0068] (2) The written evidence completeness index SZs takes into account the number of events in the written evidence, the relevance of the events, and the uniformity of the event distribution, thereby providing a comprehensive assessment of the completeness of the written evidence in the case file. Through this intelligent assessment method, the completeness of the written evidence in the case file can be assessed more objectively and accurately, providing a more reliable reference for judicial personnel. This assessment method can not only improve the efficiency of judicial work, but also ensure the fairness and accuracy of judicial decisions, thus bringing substantial benefits to the informatization of the judicial industry.

[0069] (3) In this blockchain-assisted electronic case management system, the second recognition module calculates the video integrity wzd by extracting metadata of the video recording material and verifying the continuity between video frames. This evaluation helps to determine the availability and integrity of the video recording, providing an important reference for the credibility of the case evidence. The second recognition module also calculates the clarity index Qx of the video by extracting information such as the resolution, contrast and number of frames of the video. Clarity is one of the important factors affecting video quality. By evaluating it, we can better understand the video content and improve our understanding and analysis of the case. The event behavior recognition module uses target detection and tracking algorithms in computer vision technology, such as YOLO and FasterR-CNN, to identify case targets, vehicles and related behaviors in the video. This helps to quickly locate and identify key events, providing more accurate and comprehensive information for the analysis of case evidence. The number of identified event features, such as the number of event occurrences, duration and location interval difference, is counted and calculated using the event index SJzs formula. This helps to quantify the importance and relevance of events in case evidence, providing an objective basis for the analysis and evaluation of the case.

[0070] (4) The recording quality coefficient Zlx takes into account multiple factors, including speech duration, average speech amplitude, signal-to-noise ratio, and the number of event keywords. Through the comprehensive evaluation of these factors, the quality and credibility of the recording can be judged more comprehensively, which helps to improve the accuracy of the analysis and evaluation of recording evidence;

[0071] The fourth recognition unit uses natural language processing technology and text analysis tools to recognize and analyze email texts, thereby generating an email interaction behavior index Xwzs, which can objectively evaluate the level of email interaction activity involved in the case file and provide important clues for case investigation and analysis. By comparing the generated email interaction behavior index with the set fourth threshold, the case files can be divided into different priorities according to the level of email interaction activity, so as to better determine the priority of review and investigation. This helps to optimize the allocation of investigation resources and improve investigation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 This is a flowchart diagram of a blockchain-assisted electronic case management system of the present invention. DETAILED DESCRIPTION

[0073] 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.

[0074] Example 1

[0075] See also Figure 1 , the present invention provides a blockchain-assisted electronic case management system. To achieve the above purpose, the present invention is implemented through the following technical solutions: a blockchain-assisted electronic case management system, including a collection unit, a first identification unit, a second identification unit, a third identification unit, an evaluation unit, an early warning unit, a strategy unit and a blockchain storage unit;

[0076] The collection unit is used to collect case file information and corresponding case file evidence data to form a case file set, and clean and pre-process the case file set to remove invalid information and duplicate data, and then convert the case file set into a unified standard format; the case file evidence data includes written materials, video recording materials, audio recording materials and digital evidence;

[0077] The first recognition unit is used to extract the written material, identify the text content of the extracted written material through text recognition technology, identify the number of events in the quantified written evidence, compare the number of events quantified in the written evidence with the number of events in the original document of the litigation document, and obtain the written evidence completeness index SZs; when the written evidence completeness index SZs is lower than a first threshold, the early warning unit generates and issues a first early warning instruction; when the written evidence completeness index SZs is higher than or equal to the first threshold, a first qualified mark is generated;

[0078] The second identification unit is used to extract video recording materials, identify metadata of video files, obtain resolution fbl, contrast dbd and frame number zs, compare the continuity method between video frames for verification, obtain video integrity wzd and clarity index Qx, and use video processing technology to identify case targets, vehicles and event behaviors in the video, obtain the number of event occurrences sjcxcs, duration cxsj and the position interval difference jgcz, and construct an event index SJzs; and extract key frames or important clips in the video, and generate a video summary, and associate the video integrity wzd, clarity index Qx and event index SJzs, obtain the video comprehensive coefficient Zh, and perform a first mark in the video summary;

[0079] The third recognition unit is used to extract the recording material, and is used to use the PESQ recognition tool to identify and obtain the voice duration yycs, the average voice amplitude zf and the signal-to-noise ratio xzb, and after converting the recording material into text, extract the event keywords, obtain the event keyword times sjcs, and construct the recording quality coefficient Zlx;

[0080] The evaluation unit is used to evaluate the video comprehensive coefficient Zh and the recording quality coefficient Zlx respectively, obtain the second evaluation result and the third evaluation result, and the second evaluation result and the third evaluation result;

[0081] The strategy unit is used to generate and execute corresponding strategies according to the first evaluation result, the second evaluation result and the third evaluation result, and when the first evaluation result, the second evaluation result and the third evaluation result are all qualified, the qualified case file evidence data is archived in the corresponding case file through the blockchain storage unit.

[0082] In this embodiment, the blockchain-assisted electronic case management system proposed by the present invention realizes the all-round management and intelligent processing of case evidence through the organic combination of modules such as the collection unit, the identification unit, the evaluation unit, the early warning unit, the strategy unit and the blockchain storage unit. The collection unit is responsible for collecting case information and corresponding evidence data, and cleaning and preprocessing; the identification unit identifies and analyzes different forms of evidence through technologies such as text recognition, video processing and voice recognition; the evaluation unit evaluates the recognition results and generates corresponding evaluation results; the early warning unit issues an early warning instruction when the evaluation results do not meet the standards; the strategy unit generates and executes corresponding strategies based on the evaluation results; the blockchain storage unit is responsible for archiving qualified case evidence data on the blockchain to ensure the security and non-tamperability of the data.

[0083] Example 2: This example is an explanation of Example 1. Figure 1 ,Specifically, the first recognition unit includes a semantic analysis module and a first calculation module;

[0084] The semantic analysis module is used to perform semantic analysis on the written materials using natural language processing technology, and to identify the number of events in the quantified written evidence through keyword extraction, obtain the number of events sz1, and calculate the average interval time pjjg between events in the case file; compare the number of events sz1 quantified in the written evidence with the number of events Bz in the original document of the litigation document, and identify the repeated event patterns and event groups with strong correlation in the written materials, and obtain the associated event confidence C (A→B) of event A and event B through the following formula:

[0085]

[0086] In the formula, s(A∪B) represents the number of times event A and event B appear at the same time; s(A) represents the number of times event A appears alone; and C(A→B) higher than the preset correlation threshold is classified as a correlated event, and the number of correlated events xglsj is obtained by counting;

[0087] The first calculation module is used to calculate the written evidence integrity index SZs by the following formula after dimensionless processing of the event number sz1, the event number Bz of the original litigation document, the number of associated events xglsj and the average interval time pjjg:

[0088]

[0089] Wherein, 0<w1<1, 0<w2<1, 0<w3<1, and w1+w2+w3=1, w1, w2 and w3 are weights, and their specific values ​​are adjusted and set by the user; A is a first constant correction coefficient; the first recognition unit includes a semantic analysis module and a first calculation module;

[0090] The semantic analysis module is used to perform semantic analysis on the written materials using natural language processing technology, and to identify the number of events in the quantified written evidence through keyword extraction, obtain the number of events sz1, and calculate the average interval time pjjg between events in the case file; compare the number of events sz1 quantified in the written evidence with the number of events Bz in the original document of the litigation document, and identify the repeated event patterns and event groups with strong correlation in the written materials, and obtain the associated event confidence C (A→B) of event A and event B through the following formula:

[0091]

[0092] In the formula, s(A∪B) represents the number of times event A and event B appear at the same time; s(A) represents the number of times event A appears alone; and C(A→B) higher than the preset correlation threshold is classified as a correlated event, and the number of correlated events xglsj is obtained by counting;

[0093] The first calculation module is used to calculate the written evidence integrity index SZs by the following formula after dimensionless processing of the event number sz1, the event number Bz of the original litigation document, the number of associated events xglsj and the average interval time pjjg:

[0094]

[0095] Among them, 0<w1<1, 0<w2<1, 0<w3<1, and w1+w2+w3=1, w1, w2 and w3 are weights, and their specific values ​​are adjusted and set by the user; A is the first constant correction coefficient.

[0096] In this embodiment, through semantic analysis, the system can deeply understand the written materials, so as to accurately identify and quantify the number of events therein, and ensure the comprehensiveness and accuracy of the evaluation results. The use of natural language processing technology automates the evaluation process, greatly improves processing efficiency, and reduces the burden of manual operations. By identifying related events and repeated event patterns, it helps users to promptly discover potential risks and take corresponding measures.

[0097] The written evidence integrity index SZs takes into account the number of events, the relevance of events, and the uniformity of event distribution, thus providing a comprehensive assessment of the integrity of the case file. This intelligent assessment method can more objectively and accurately assess the integrity of the written evidence in the case file, providing a more reliable reference for judicial personnel. This assessment method can not only improve judicial work efficiency, but also ensure the fairness and accuracy of judicial decisions, thus bringing substantial benefits to the informatization of the judicial industry.

[0098] Example 3, this example is an explanation of Example 2, please refer to Figure 1 Specifically, the first identification unit further includes a first judgment unit, which is used to compare and judge the written evidence integrity index SZs with a first threshold value to obtain a first evaluation result, including:

[0099] When the written evidence integrity index SZs is less than the first threshold, it means that the written evidence is unqualified, and the warning unit generates and issues a first warning instruction;

[0100] When the written evidence completeness index SZs ≥ the first threshold, a first qualified mark is generated.

[0101] In this embodiment, when the written evidence integrity index SZs is lower than the pre-set first threshold, it means that the integrity of the written evidence does not meet the requirements and there may be potential problems or risks. At this time, the early warning unit will generate a first early warning instruction and issue it immediately to remind relevant personnel to take necessary measures in a timely manner. The operation of the first judgment unit realizes the automated evaluation and processing of the integrity of written evidence, reduces the need for manual intervention, and improves processing efficiency and accuracy. The generation of the first qualified mark represents that the integrity of the written evidence reaches or exceeds the expected level and meets the requirements of the system. This helps to optimize the case management process and improve management efficiency and quality. Timely warnings and qualified marks can help institutions effectively reduce potential risks and reduce the use of erroneous or incomplete evidence, thereby saving judicial resources and improving the reliability of case management.

[0102] Example 4: This example is an explanation of Example 1. Figure 1 ,Specifically, the second recognition unit includes a second recognition module and an event behavior recognition module;

[0103] The second recognition module is used to extract video recording materials, identify the metadata of the video file through the FFmpeg video processing tool, and verify the continuity method between video frames, and generate the video integrity wzd through the following formula:

[0104]

[0105] In the formula, zjp represents the standard deviation of the time interval between video frames, and α represents the adjustment parameter, which is used to adjust the influence of the standard deviation on the video integrity. The smaller the standard deviation zjp is, the more stable the time interval between video frames is, and the better the continuity of the video is.

[0106] The clarity index Qx is obtained by extracting the resolution fbl, the contrast dbd and the number of frames zs, and after dimensionless processing, the clarity index Qx is generated by the following formula:

[0107] Qx=w4*fbl+w5*dbd*w6*zs+B

[0108] Wherein, 0<w4<1, 0<w5<1, 0<w6<1, and w4+w5+w6=1, w4, w5 and w6 are weights, and their specific values ​​are adjusted and set by the user; B is the second constant correction coefficient;

[0109] The event behavior recognition module is used to use the target detection and tracking algorithms in computer vision technology, including YOLO and FasterR-CNN, to identify the case target and the target positioning position of the vehicle in the video, and track the movement trajectory of the target in the video. Based on the target detection and tracking results, the traffic accident behavior and the gathering behavior of people are identified and counted as event features;

[0110] The number of times the event features are identified is counted, and the number of event occurrences sjcxcs, duration cxsj, and location interval difference jgcz are obtained. After dimensionless processing, the event index SJzs is generated using the following formula:

[0111] SJzs=w7*sjcxcs+w8*cxsj*w9*jgcz

[0112] Among them, 0<w7<1, 0<w8<1, 0<w9<1, and w7+w8+w9=1, w7, w8 and w9 are weights, and their specific values ​​are adjusted and set by the user.

[0113] In this embodiment, the second recognition module calculates the video integrity wzd by extracting the metadata of the video recording material and verifying the continuity between the video frames. This evaluation helps to determine the availability and integrity of the video recording, and provides an important reference for the credibility of the case evidence. The second recognition module also calculates the clarity index Qx of the video by extracting information such as the resolution, contrast and number of frames of the video. Clarity is one of the important factors affecting video quality. By evaluating it, the video content can be better understood and the ability to understand and analyze the case can be improved. The event behavior recognition module uses target detection and tracking algorithms in computer vision technology, such as YOLO and FasterR-CNN, to identify case targets, vehicles and related behaviors in the video. This helps to quickly locate and identify key events and provide more accurate and comprehensive information for the analysis of case evidence. The number of identified event features, such as the number of event occurrences, duration and location interval difference, is counted and calculated by the event index SJzs formula. This helps to quantify the importance and relevance of events in case evidence and provides an objective basis for the analysis and evaluation of cases.

[0114] Example 5. This example is an explanation of Example 4. Please refer to Figure 1 ,Specifically, the second recognition unit also includes an association module and a marking module;

[0115] The association module is used to generate a video comprehensive coefficient Zh by combining the video completeness wzd, the clarity index Qx and the event index SJzs through the following association consensus:

[0116] Zh=x*wzd+y*Qx*z*SJzs

[0117] Wherein, x, y and z represent the preset proportional coefficients of the video completeness wzd, the clarity index Qx and the event index SJzs, respectively, 0.25≤x≤0.33, 0.20≤y≤0.25, 0.35≤z≤0.42, and 0.80≤x+y+z≤1.0;

[0118] The marking module is used to extract key frames or important segments in the video and generate a video summary, and to first mark the video comprehensive coefficient Zh on the first page of the video summary.

[0119] In this embodiment, the association module generates a video comprehensive coefficient Zh by associating the video completeness wzd, clarity index Qx and event index SJzs through the associated consensus. This comprehensive coefficient comprehensively considers the integrity, clarity and key events contained in the video, and provides a comprehensive reference for the analysis and evaluation of subsequent case evidence. The marking module extracts key frames or important clips in the video based on the generated video comprehensive coefficient and generates a video summary. The video summary is a concise summary of the video content, which helps users quickly understand the main content and important events of the video and improves the efficiency of information communication.

[0120] Example 6, this example is an explanation of Example 5, please refer to Figure 1 Specifically, the recording quality coefficient Zlx is obtained by extracting the speech duration yycs, the average speech amplitude zf, the signal-to-noise ratio xzb and the number of event keywords sjcs, and after dimensionless processing, the recording quality coefficient Zlx is generated by the following formula:

[0121]

[0122] Among them, bz2 represents the preset standard speech duration threshold, bz3 represents the preset speech average amplitude threshold, bz4 represents the preset signal-to-noise ratio threshold, bz5 represents the preset time keyword number threshold, q, w, e and r represent weight values, and 0<q<1, 0<w<1, 0<e<1, 0<r<1 and q+w+e+r≤1.0; its specific value is adjusted and set by the user; C is the third constant correction coefficient.

[0123] In this embodiment, the recording quality coefficient Zlx comprehensively considers multiple factors such as speech duration, average speech amplitude, signal-to-noise ratio, and the number of event keywords. Through the comprehensive evaluation of these factors, the quality and credibility of the recording can be judged more comprehensively, which helps to improve the accuracy of analysis and evaluation of recording evidence. Using a dimensionless processing method, the recording quality indicators of different dimensions are unified to the same dimension, which is convenient for comparison and comprehensive evaluation. This processing method eliminates the dimensional differences between different indicators, so that the impact of each indicator on the final quality coefficient can be reasonably weighed.

[0124] Example 7. This example is an explanation of Example 6. Please refer to Figure 1 ,Specifically, the evaluation unit includes a first evaluation module and a second evaluation module;

[0125] The first evaluation module is used to compare the video comprehensive coefficient Zh with the second threshold value to obtain a second evaluation result, including:

[0126] When the video comprehensive coefficient Zh ≥ the second threshold value, it indicates that the video quality is qualified, and a second qualified mark is generated;

[0127] When the video comprehensive coefficient Zh is less than the second threshold value, indicating that the video is unqualified, the warning unit generates and issues a second warning instruction;

[0128] The second evaluation module is used to compare the recording quality coefficient Zlx with the third threshold to obtain a third evaluation result, including:

[0129] When the recording quality coefficient Zlx ≥ the third threshold, it indicates that the recording quality is qualified, and a third qualified mark is generated;

[0130] When the recording quality coefficient Zlx is less than the third threshold value, indicating that the recording quality is unqualified, the warning unit generates and issues a third warning instruction.

[0131] In this embodiment, by comparing the video comprehensive coefficient Zh and the recording quality coefficient Zlx with the preset threshold value, the video quality and the recording quality can be evaluated in a standardized manner. This standardized processing helps to ensure the objectivity and consistency of the evaluation results, and improves the reliability and credibility of the case management system. When the video or recording quality does not meet the expected standards, the evaluation module will promptly generate and issue corresponding early warning instructions to remind relevant personnel to pay attention to the existence of the problem. This timely early warning can help managers discover and solve problems in a timely manner, avoid adverse effects on case handling, and ensure the smooth progress of judicial work.

[0132] Example 8. This example is an explanation of Example 7. Please refer to Figure 1 Specifically, the strategy unit is used to generate and execute corresponding strategies according to the first evaluation result, the second evaluation result and the third evaluation result, including:

[0133] Generate a first strategy according to the first warning instruction, the first strategy including: re-examining the written evidence to find out possible defects or incompleteness, and correcting or supplementing them as soon as possible; and sending a reminder notice to the relevant responsible person to remind him to pay attention to the integrity of the written evidence and urge him to take action to correct it as soon as possible;

[0134] Generate a second strategy according to the second warning instruction, the second strategy comprising: re-recording the video, adjusting the environmental optical fiber, camera settings and shooting angle during the recording process, and using video processing software to correct and optimize the recorded video, clipping and editing the recorded video, deleting non-event parts, retaining the event content, and adding annotations and descriptions;

[0135] A third strategy is generated based on the third warning instruction, and the third strategy includes: re-recording the audio, installing noise jammers in the recording equipment and environment, and using audio processing software to post-process the recorded audio to perform noise reduction, equalization, gain adjustment and processing to optimize the quality and clarity of the audio, and marking the event description or testimony content in the corrected audio.

[0136] In this embodiment, the first strategy proposes a correction or supplement plan for possible defects or incompleteness in the written evidence. By re-examining the written evidence and sending reminder notifications to the relevant responsible persons, problems can be discovered in time and prompted to make corrections, thereby improving the integrity and accuracy of the written evidence. The second strategy mainly targets the quality issues of video recording materials. By re-recording the video and correcting, optimizing, clipping and editing the recorded video, the clarity and continuity of the video can be improved, the quality and availability of the video can be improved, and a more reliable basis for the review and analysis of the case can be provided. The third strategy focuses on solving the quality problems that may exist in the recording materials. By re-recording the audio, installing noise jammers in the recording equipment and environment, and using audio processing software for post-processing, it can reduce noise, equalize, adjust gain and process, thereby improving the quality and clarity of the audio, making the event description or testimony content clearer and more recognizable, which is helpful for the investigation and trial of the case.

[0137] Example 9. This example is an explanation of Example 1. Please refer to Figure 1 Specifically, it also includes a fourth identification unit. The fourth identification unit retrieves and extracts emails under specified conditions for the digital evidence through an email crawler, and identifies information in the email text through natural language processing technology and text analysis tools to obtain the email sending frequency fspl, the size of the sent attachment fjdx, the total number of emails sent and received zsl, the email reply rate hfl and the number of email citations yycs. After dimensionless processing, the email interaction behavior index Xwzs is generated through the following formula:

[0138]

[0139] Wherein, a, k, d, b and h represent the preset proportional coefficients of the email sending frequency fspl, the total number of emails sent and received zsl, the email reply rate hfl, the number of email citations yycs and the size of the sent attachments fjdx, respectively, and 0.17≤a≤0.21, 0.10≤k≤0.12, 0.15≤d≤0.19, 0.16≤b≤0.18, 0.22≤h≤0.30, and 0.80≤a+k+d+b+h≤1.0; D is the fourth constant correction coefficient;

[0140] Comparing the email interaction behavior index Xwzs with the fourth threshold value to obtain a fourth evaluation result, including: when the email interaction behavior index Xwzs ≥ the fourth threshold value, it indicates that the email interaction behavior activity value is qualified, and a more in-depth investigation and analysis is conducted on the case files with relatively active email interaction behaviors; and the case files are reviewed as the first priority;

[0141] When the email interaction behavior index Xwzs is less than the fourth threshold, it indicates that the email interaction behavior is relatively unqualified and is ranked as the second priority for attention and review.

[0142] In this embodiment, the fourth recognition unit uses natural language processing technology and text analysis tools to recognize and analyze email text, thereby generating an email interaction behavior index Xwzs, which can objectively evaluate the active level of email interaction involved in the case file and provide important clues for case investigation and analysis. By comparing the generated email interaction behavior index with the set fourth threshold, the case file can be divided into different priorities according to the active level of email interaction, so as to better determine the priority of review and investigation. This helps to optimize the allocation of investigation resources and improve investigation efficiency.

[0143] Example 10. This example is an explanation of Example 1. Please refer to Figure 1 ,Specifically, it also includes an intelligent access unit and an intelligent log monitoring unit;

[0144] The intelligent access unit is used to monitor the access behavior of system users, including login, operation and file access behavior, identify abnormal behavior and potential threats through real-time monitoring and analysis of access logs, and take corresponding measures in time; the intelligent access unit interacts with the evaluation unit and the early warning unit to generate corresponding strategies and early warning instructions according to the monitoring results;

[0145] The intelligent log monitoring unit is used to monitor and analyze the system's log data in real time, including user operation logs, system logs and security logs. It detects abnormal events and potential threats through technologies such as machine learning algorithms and rule engines, and responds in real time.

[0146] In this embodiment, the intelligent access unit and the intelligent log monitoring unit can monitor the access behavior of system users and system log data in real time. This enables the system to detect abnormal events and potential threats in a timely manner, helping to prevent security vulnerabilities and improper behavior. By analyzing access behavior and log data, the intelligent unit can identify abnormal access patterns and unusual operating behaviors. This helps system administrators to detect possible security issues in a timely manner and take appropriate measures to deal with them.

[0147] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A blockchain-assisted electronic case management system, characterized by: It includes a collection unit, a first identification unit, a second identification unit, a third identification unit, an evaluation unit, an early warning unit, a strategy unit and a blockchain storage unit; The collection unit is used to collect case file information and corresponding case file evidence data to form a case file set, and clean and pre-process the case file set to remove invalid information and duplicate data, and then convert the case file set into a unified standard format; the case file evidence data includes written materials, video recording materials, audio recording materials and digital evidence; The first recognition unit is used to extract the written material, identify the text content of the extracted written material through text recognition technology, identify the number of events in the quantified written evidence, compare the number of events quantified in the written evidence with the number of events in the original document of the litigation document, and obtain the written evidence completeness index SZs; when the written evidence completeness index SZs is lower than a first threshold, the early warning unit generates and issues a first early warning instruction; when the written evidence completeness index SZs is higher than or equal to the first threshold, a first qualified mark is generated; The second identification unit is used to extract video recording materials, identify metadata of video files, obtain resolution fbl, contrast dbd and frame number zs, compare the continuity method between video frames for verification, obtain video integrity wzd and clarity index Qx, and use video processing technology to identify case targets, vehicles and event behaviors in the video, obtain the number of event occurrences sjcxcs, duration cxsj and the position interval difference jgcz, and construct an event index SJzs; and extract key frames or important clips in the video, and generate a video summary, and associate the video integrity wzd, clarity index Qx and event index SJzs, obtain the video comprehensive coefficient Zh, and perform a first mark in the video summary; The third recognition unit is used to extract the recording material, and is used to use the PESQ recognition tool to identify and obtain the voice duration yycs, the average voice amplitude zf and the signal-to-noise ratio xzb, and after converting the recording material into text, extract the event keywords, obtain the event keyword times sjcs, and construct the recording quality coefficient Zlx; The evaluation unit is used to evaluate the video comprehensive coefficient Zh and the recording quality coefficient Zlx respectively, obtain the second evaluation result and the third evaluation result, and the second evaluation result and the third evaluation result; The strategy unit is used to generate and execute corresponding strategies according to the first evaluation result, the second evaluation result and the third evaluation result, and when the first evaluation result, the second evaluation result and the third evaluation result are all qualified, the qualified case file evidence data is archived in the corresponding case file through the blockchain storage unit.

2. A blockchain-assisted electronic case management system according to claim 1, characterized in that: The first recognition unit includes a semantic analysis module and a first calculation module; The semantic analysis module is used to perform semantic analysis on the written materials using natural language processing technology, and to identify the number of events in the quantified written evidence through keyword extraction, obtain the number of events sz1, and calculate the average interval time pjjg between events in the case file; compare the number of events sz1 quantified in the written evidence with the number of events Bz in the original document of the litigation document, and identify the repeated event patterns and event groups with strong correlation in the written materials, and obtain the associated event confidence of event A and event B through the following formula : In the formula, Indicates the number of times event A and event B occur simultaneously; Indicates the number of times event A occurs alone; and will Those higher than the preset correlation threshold are classified as correlation events, and the number of correlation events xglsj is obtained by counting; The first calculation module is used to calculate the written evidence integrity index SZs by the following formula after dimensionless processing of the event number sz1, the event number Bz of the original litigation document, the number of associated events xglsj and the average interval time pjjg: Among them, 0<w1<1, 0<w2<1, 0<w3<1, and w1+w2+w3=1, w1, w2 and w3 are weights, and their specific values ​​are adjusted and set by the user; A is the first constant correction coefficient.

3. According to claim 1, a blockchain-assisted electronic case management system is characterized by: The first identification unit further includes a first judgment unit, which is used to compare and judge the written evidence integrity index SZs with a first threshold value to obtain a first evaluation result, including: When the written evidence integrity index SZs is less than the first threshold, it means that the written evidence is unqualified, and the warning unit generates and issues a first warning instruction; When the written evidence completeness index SZs ≥ the first threshold, a first qualified mark is generated.

4. A blockchain-assisted electronic case management system according to claim 1, characterized in that: The second recognition unit includes a second recognition module and an event behavior recognition module; The second recognition module is used to extract video recording materials, identify the metadata of the video file through the FFmpeg video processing tool, and verify the continuity method between video frames, and generate the video integrity wzd through the following formula: In the formula, zjp represents the standard deviation of the time interval between video frames, and α represents the adjustment parameter, which is used to adjust the influence of the standard deviation on the video integrity. The smaller the standard deviation zjp is, the more stable the time interval between video frames is, and the better the continuity of the video is. The clarity index Qx is obtained by extracting the resolution fbl, the contrast dbd and the number of frames zs, and after dimensionless processing, the clarity index Qx is generated by the following formula: Wherein, 0<w4<1, 0<w5<1, 0<w6<1, and w4+w5+w6=1, w4, w5 and w6 are weights, and their specific values ​​are adjusted and set by the user; B is the second constant correction coefficient; The event behavior recognition module is used to use the target detection and tracking algorithms in computer vision technology, including YOLO and FasterR-CNN, to identify the case target and the target positioning position of the vehicle in the video, and track the movement trajectory of the target in the video. Based on the target detection and tracking results, the traffic accident behavior and the gathering behavior of people are identified and counted as event features; The number of times the event features are identified is counted, and the number of event occurrences sjcxcs, duration cxsj, and location interval difference jgcz are obtained. After dimensionless processing, the event index SJzs is generated using the following formula: Among them, 0<w7<1, 0<w8<1, 0<w9<1, and w7+w8+w9=1, w7, w8 and w9 are weights, and their specific values ​​are adjusted and set by the user.

5. According to claim 1, a blockchain-assisted electronic case management system is characterized by: The second recognition unit also includes an association module and a marking module; The association module is used to generate a video comprehensive coefficient Zh by combining the video completeness wzd, the clarity index Qx and the event index SJzs through the following association consensus: Wherein, x, y and z represent the preset proportional coefficients of the video completeness wzd, the clarity index Qx and the event index SJzs, respectively, 0.25≤x≤0.33, 0.20≤y≤0.25, 0.35≤z≤0.42, and 0.80≤x+y+z≤1.0; The marking module is used to extract key frames or important segments in the video and generate a video summary, and to first mark the video comprehensive coefficient Zh on the first page of the video summary.

6. A blockchain-assisted electronic case management system according to claim 1, characterized in that: The recording quality coefficient Zlx is obtained by extracting the speech duration yycs, the average speech amplitude zf, the signal-to-noise ratio xzb and the number of event keywords sjcs, and after dimensionless processing, the recording quality coefficient Zlx is generated by the following formula: Among them, bz2 represents the preset standard speech duration threshold, bz3 represents the preset speech average amplitude threshold, bz4 represents the preset signal-to-noise ratio threshold, bz5 represents the preset time keyword number threshold, q, w, e and r represent weight values, and 0<q<1, 0<w<1, 0<e<1, 0<r<1 and q+w+e+r≤1.0; its specific value is adjusted and set by the user; C is the third constant correction coefficient.

7. A blockchain-assisted electronic case management system according to claim 1, characterized in that: The evaluation unit includes a first evaluation module and a second evaluation module; The first evaluation module is used to compare the video comprehensive coefficient Zh with the second threshold value to obtain a second evaluation result, including: When the video comprehensive coefficient Zh ≥ the second threshold value, it indicates that the video quality is qualified, and a second qualified mark is generated; When the video comprehensive coefficient Zh is less than the second threshold value, indicating that the video is unqualified, the warning unit generates and issues a second warning instruction; The second evaluation module is used to compare the recording quality coefficient Zlx with the third threshold to obtain a third evaluation result, including: When the recording quality coefficient Zlx ≥ the third threshold, it indicates that the recording quality is qualified, and a third qualified mark is generated; When the recording quality coefficient Zlx is less than the third threshold value, indicating that the recording quality is unqualified, the warning unit generates and issues a third warning instruction.

8. A blockchain-assisted electronic case management system according to claim 1, characterized in that: The strategy unit is used to generate and execute corresponding strategies according to the first evaluation result, the second evaluation result and the third evaluation result, including: Generate a first strategy according to the first warning instruction, the first strategy including: re-examining the written evidence to find out possible defects or incompleteness, and correcting or supplementing them as soon as possible; and sending a reminder notice to the relevant responsible person to remind him to pay attention to the integrity of the written evidence and urge him to take action to correct it as soon as possible; Generate a second strategy according to the second warning instruction, the second strategy comprising: re-recording the video, adjusting the environmental optical fiber, camera settings and shooting angle during the recording process, and using video processing software to correct and optimize the recorded video, clipping and editing the recorded video, deleting non-event parts, retaining event content, and adding annotations and descriptions; A third strategy is generated based on the third warning instruction, and the third strategy includes: re-recording the audio, installing noise jammers in the recording equipment and environment, and using audio processing software to post-process the recorded audio to perform noise reduction, equalization, gain adjustment and processing to optimize the quality and clarity of the audio, and marking the event description or testimony content in the corrected audio.

9. A blockchain-assisted electronic case management system according to claim 1, characterized in that: The fourth identification unit is further included. The fourth identification unit retrieves and extracts emails under specified conditions from the digital evidence through an email crawler, and identifies information in the email text through natural language processing technology and text analysis tools to obtain the email sending frequency fspl, the size of the sent attachment fjdx, the total number of emails sent and received zsl, the email reply rate hfl and the number of email citations yycs. After dimensionless processing, the email interaction behavior index Xwzs is generated through the following formula: Wherein, a, k, d, b and h represent the preset proportional coefficients of the email sending frequency fspl, the total number of emails sent and received zsl, the email reply rate hfl, the number of email citations yycs and the size of the sent attachments fjdx, respectively, and 0.17≤a≤0.21, 0.10≤k≤0.12, 0.15≤d≤0.19, 0.16≤b≤0.18, 0.22≤h≤0.30, and 0.80≤a+k+d+b+h≤1.0; D is the fourth constant correction coefficient; Comparing the email interaction behavior index Xwzs with the fourth threshold value to obtain a fourth evaluation result, including: when the email interaction behavior index Xwzs ≥ the fourth threshold value, it indicates that the email interaction behavior activity value is qualified, and a more in-depth investigation and analysis is conducted on the case files with relatively active email interaction behaviors; and the case files are reviewed as the first priority; When the email interaction behavior index Xwzs is less than the fourth threshold, it indicates that the email interaction behavior is relatively unqualified and is ranked as the second priority for attention and review.

10. A blockchain-assisted electronic case management system according to claim 1, characterized in that: It also includes an intelligent access unit and an intelligent log monitoring unit; The intelligent access unit is used to monitor the access behavior of system users, including login, operation and file access behavior, identify abnormal behavior and potential threats through real-time monitoring and analysis of access logs, and take corresponding measures in time; the intelligent access unit interacts with the evaluation unit and the early warning unit to generate corresponding strategies and early warning instructions according to the monitoring results; The intelligent log monitoring unit is used to monitor and analyze the system's log data in real time, including user operation logs, system logs and security logs. It detects abnormal events and potential threats through machine learning algorithms and rule engine technology, and responds in real time.

Images