Digital archive calling management method and system based on artificial intelligence

Abstract

The invention relates to the technical field of file calling management and information processing, in particular to a digital file calling management method and system based on artificial intelligence. A digital archive calling management method based on artificial intelligence comprises the following steps: S1, collecting multiple types of archive data, digitalizing the multiple types of archive data by adopting an artificial intelligence technology to generate archive fragment data, and constructing a digital archive library according to the archive fragment data; and S2, according to the digital archive library, processing archive fragment data by using a natural language processing technology, generating an archive fragment data information table, mining a relationship among different types of data, and constructing a multi-modal digital archive. According to the method, the file fragment data is recommended for the visitors by calculating the values of the file fragment data for different departments, and the multi-modal file is constructed and displayed through the virtual interaction equipment, so that the consulting efficiency and the interaction experience of the visitors are improved.

Description

Technical Field

[0001] The present invention relates to the field of file retrieval management and information processing technology, and in particular to a digital file retrieval management method and system based on artificial intelligence. Background Art

[0002] Digitization refers to the process of converting analog signals into digital signals using computer technology. Archival digitization involves the use of scanners and computer technology to digitize archives, converting them into digital images and text data stored on disks and CDs. Furthermore, the process of establishing a correlation between catalog data and digital images based on the archives' inherent connections is used.

[0003] With the advancement of archive digitization, archives now contain a vast array of data types. However, current archival systems use separate databases to store different types of archival data, such as text, audio, and video. This lacks deep semantic connections between different types of data, making it difficult for users to access them. This requires simultaneous understanding and analysis of different types of archival data, resulting in a poor interactive experience. Furthermore, due to a lack of value filtering, users must sift through vast amounts of data to find the high-value information they need, reducing query efficiency and wasting significant time.

[0004] Therefore, there is an urgent need to develop a digital archive retrieval management method and system based on artificial intelligence. Summary of the Invention

[0005] In order to overcome the shortcomings of traditional archives where different types of data are isolated and inconvenient to access, the present invention provides a digital archive retrieval management method and system based on artificial intelligence.

[0006] The technical solution is as follows: A digital archive retrieval management method based on artificial intelligence, comprising the following steps: S1: Collect multi-type archival data, digitize them using artificial intelligence technology, generate archival fragment data, and build a digital archive based on the archival fragment data; S2: Based on the digital archive, using natural language processing technology, processing the archive fragment data, generating an archive fragment data information table, mining the relationship between different types of data, and constructing a multimodal digital archive; S3: Calculate the value of the archive fragment data for different departments using the archive fragment data value calculation formula according to the archive fragment data information table, and generate an archive fragment data value table; S4: constructing an initial virtual interaction scene and a process virtual interaction scene according to the multimodal digital archive; S5: Based on the visitor information, recommend archive fragment data according to the archive fragment data value table, display the corresponding archive fragment data through the virtual reality interactive device, and provide real-time feedback on the visitor's browsing progress, and switch the virtual interactive scene of the process based on the visitor's browsing progress; S6: Record the visitor information, query content, query time and feedback information, generate an archive query backtracking table, and update the archive fragment data value table according to the visitor feedback information.

[0007] Preferably, the collecting of multi-type archive data includes: the multi-type archive data is composed of text data, image data, audio data and video data of the same archive.

[0008] Preferably, the method of digitizing multiple types of archival data using artificial intelligence technology to generate archival fragment data and constructing a digital archive based on the archival fragment data includes: Use semantic segmentation technology to process text data and segment semantically coherent text fragments; Use image recognition technology to process image data and extract text fragment data; Use speech recognition technology to convert audio data into text snippet data; Using key frame extraction technology, the video data is divided into video segments with key frames as segmentation frames. The video key frames are processed by combining image recognition technology and speech recognition technology to generate description information of the video segments.

[0009] Preferably, the digital archive is used to process archive fragment data using natural language processing technology to generate an archive fragment data information table, mine the connections between different types of data, and construct a multimodal digital archive, including: Perform knowledge extraction on the text segment data and video segment description information obtained by processing multi-type data, obtain the corresponding entity, relationship and attribute data and initial association relationship data, count the entity, relationship and attribute data of different departments, and generate the archive segment data information table; Construct semantic relationships between entities based on entity, relationship and attribute data and initial association relationship data, establish connections between text data, image data, audio data and video data, link different types of data together, and generate multimodal digital archives.

[0010] Preferably, the archival fragment data value calculation formula is used to calculate the value of the archival fragment data for different departments according to the archival fragment data information table to generate the archival fragment data value table, including: the archival fragment data value calculation formula is: ; Where, The value of the archive fragment data to the target department, 、 and is the weight, The number of entities of interest to the target department included in the archive fragment data. is the total number of entities contained in the archive fragment data. The number of relationships that the target department is concerned about included in the archive fragment data, is the total number of relations contained in the archive fragment data, The number of attributes that the target department is concerned about contained in the archive fragment data. is the total number of attributes contained in the archive fragment data, is the time decay coefficient of the archive fragment data, The data storage time of the archive fragment, Provide feedback to interviewers on archival footage data.

[0011] Preferably, the feedback from the interviewer on the archive fragment data includes: ; Where, is the average rating of the interviewees on the value of the archive fragment data, The frequency of access to archive fragment data.

[0012] Preferably, the constructing of the initial virtual interaction scene and the process virtual interaction scene based on the multimodal digital archive includes: constructing the initial virtual interaction scene based on the archive core entity, and constructing the process virtual interaction scene based on the archive image data, video key frame data and video clip data.

[0013] Preferably, based on the visitor information, archival fragment data is recommended according to the archival fragment data value table, the corresponding archival fragment data is displayed through the virtual reality interactive device, and the visitor's review progress is fed back in real time. The process virtual interaction scene is switched based on the visitor's review progress, including: the visitor's identity information and the visitor's department information; based on the visitor's department information, according to the archival fragment data value table, archival fragment data with high value for the target department is recommended, and the corresponding archival fragment data is displayed through the virtual interactive device. The starting scene is the initial virtual interaction scene, and the process virtual interaction scene is switched according to the visitor's review progress.

[0014] Preferably, the record of visitor information, review content, review time and feedback information generates an archive query backtracking table, and updates the archive fragment data value table according to the visitor feedback information, including: recalculating the archive fragment data value based on the visitor feedback information every preset statistical period, and updating the archive fragment data value table according to the calculated archive fragment data value.

[0015] An artificial intelligence-based digital archive call management system, comprising: Archive digitization module: collects multiple types of archival data, digitizes them using artificial intelligence technology, generates archival fragment data, and builds a digital archive based on the archival fragment data; Archive generation module: Based on the digital archive, using natural language processing technology, it processes the archive fragment data, generates an archive fragment data information table, mines the relationship between different types of data, and constructs a multimodal digital archive; Value calculation module: according to the archive fragment data information table, calculate the value of the archive fragment data for different departments using the archive fragment data value calculation formula, and generate an archive fragment data value table; Archive interaction module: constructs an initial virtual interaction scene and a process virtual interaction scene based on the multimodal digital archive; recommends archive fragment data according to the archive fragment data value table based on the visitor information, displays the corresponding archive fragment data through the virtual reality interaction device, and provides real-time feedback on the visitor's browsing progress, and switches the process virtual interaction scene based on the visitor's browsing progress; Record update module: records visitor information, query content, query time and feedback information, generates an archive query backtracking table, and updates the archive fragment data value table based on the visitor feedback information.

[0016] The beneficial effects of the present invention are: The present invention collects multiple types of archival data, digitizes the multiple types of archival data using artificial intelligence technology, and generates a digital archive library; utilizes natural language processing technology, based on the digital archive library, to mine the connections between different types of data and construct a multimodal digital archive; establishes the association between different types of archival data, and improves the access efficiency of visitors.

[0017] The present invention calculates the value of archive fragment data for different departments based on the archive fragment data information table and uses the archive fragment data value calculation formula to generate an archive fragment data value table, which can screen high-value archive information and provide it to visitors, saving reference time.

[0018] The present invention constructs a virtual interactive scene through multimodal digital archives, switches the virtual interactive scene based on the visitor's browsing process, simplifies the operation mode, and improves the visitor's interactive experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of the digital archive call management method based on artificial intelligence of the present invention; Figure 2 This is a structural diagram of the digital archive call management system based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0021] Example 1: A digital archive management method based on artificial intelligence, such as Figure 1 As shown, the following steps are included: S1: Collect multi-type archival data, digitize them using artificial intelligence technology, generate archival fragment data, and build a digital archive based on the archival fragment data; Multi-type archive data consists of text data, image data, audio data, and video data in the same archive.

[0022] It should be noted that text data: internal electronic documents of the organization (such as reports and emails); image data: scanned drawings, photos, signed documents; audio data: meeting recordings, oral history archives; Video data: surveillance videos, training videos; different types of data in the same file constitute the multi-type file data of the file according to "file ID-data type".

[0023] Use semantic segmentation technology to process text data and segment semantically coherent text fragments; Use image recognition technology to process image data and extract text fragment data; Use speech recognition technology to convert audio data into text snippet data; Using key frame extraction technology, the video data is divided into video segments with key frames as segmentation frames. The video key frames are processed by combining image recognition technology and speech recognition technology to generate description information of the video segments.

[0024] It should be noted that text data is processed using semantic segmentation techniques using pre-trained language models, such as the BERT model, to segment semantically coherent text fragments by topic paragraphs or logical chapters. For example, a 100-page contract can be segmented into "Party A's Obligations" and "Party B's Rights." Image data is processed using an object detection model to extract text regions from scanned documents and convert them into text fragments. For example, a technical parameter table from a drawing can be extracted. Audio data is transcribed into time-stamped text fragments using an end-to-end speech recognition model. For example, a 30-minute meeting recording can be converted into segmented text. Video data is segmented into independent video clips using keyframe extraction technology, using scene changes as segmentation points. Image and speech recognition technologies are then applied to the keyframes to generate descriptions. For example, video clip V001 is described as "2025-06-10, Zhang San demonstrates the device."

[0025] All generated archive fragment data (text fragments, image recognition results, audio transcription text, video description information) are stored in a unified format. Each fragment contains the following fields: archive ID, fragment ID, data type, content, timestamp and source.

[0026] S2: Based on the digital archive, using natural language processing technology, processing the archive fragment data, generating an archive fragment data information table, mining the relationship between different types of data, and constructing a multimodal digital archive; Perform knowledge extraction on the text segment data and video segment description information obtained by processing multi-type data, obtain the corresponding entity, relationship and attribute data and initial association relationship data, count the entity, relationship and attribute data of different departments, and generate the archive segment data information table; Construct semantic relationships between entities based on entity, relationship and attribute data and initial association relationship data, establish connections between text data, image data, audio data and video data, link different types of data together, and generate multimodal digital archives.

[0027] It should be noted that named entity recognition models are used to extract entities (such as names, locations, and equipment numbers); relationship extraction models are used to identify relationships (such as "Zhang San - Works in the Technology Department"); attribute data (such as "Equipment Number D001: Power = 200kW") is extracted using attribute annotation tools to generate initial association data. Department-specific tags are pre-set (for example, the Technology Department focuses on "Equipment Parameters" and the Legal Department focuses on "Breach of Contract Liability Clauses"). The frequency of occurrence of entities, relationships, and attributes is then counted by tag classification to generate an archive fragment data information table. Using core entities (such as "R&D Project P100") as nodes, related text fragments, image scans, meeting recordings, and video keyframes are linked to construct a multimodal digital archive.

[0028] S3: Calculate the value of the archive fragment data for different departments using the archive fragment data value calculation formula according to the archive fragment data information table, and generate an archive fragment data value table; The formula for calculating the value of archive fragment data is: ; Where, The value of the archive fragment data to the target department, 、 and is the weight, The number of entities of interest to the target department included in the archive fragment data. is the total number of entities contained in the archive fragment data. The number of relationships that the target department is concerned about included in the archive fragment data, is the total number of relations contained in the archive fragment data, The number of attributes that the target department is concerned about contained in the archive fragment data. is the total number of attributes contained in the archive fragment data, is the time decay coefficient of the archive fragment data, The data storage time of the archive fragment, Provide feedback to interviewers on archival footage data.

[0029] It should be noted that different departments have different focuses. The technical department focuses on entities (equipment numbers) and attributes (technical parameters), while the legal department focuses on relationships (contract clause associations). Therefore, the weights of different departments vary, and the entities, relationships, and attribute data that different departments focus on may overlap. The density of entities of interest to the target department contained in the archive fragment data. The density of relationships that the target department is concerned about contained in the archive fragment data, The density of attributes that the target department is concerned about contained in the archive fragment data, is the time decay coefficient of the archive fragment data. The same archive fragment data has different values ​​for different departments, and the decay coefficient of the archive fragment data over time is different. This is the feedback from the access personnel on the archive fragment data. Initially, when no one accesses the archive fragment data, the feedback is 1, which has no effect on the calculation of the archive fragment data value.

[0030] Feedback from interviewers on archival footage data, including: ; Where, is the average rating of the interviewees on the value of the archive fragment data, The frequency of access to archive fragment data.

[0031] It should be noted that The average score of the interviewers on the value of the archive fragment data is counted by department, and the average score of all interviewers in the same department is (-1, 1). The access frequency of archive fragment data is the sum of the access frequencies of all personnel in the same department during the statistical period.

[0032] S4: constructing an initial virtual interaction scene and a process virtual interaction scene according to the multimodal digital archive; The initial virtual interaction scene is constructed based on the core entity of the archive, and the process virtual interaction scene is constructed based on the archive image data, video key frame data and video clip data.

[0033] It should be noted that the virtual interactive scene is constructed based on the core entity of the archive. For example, if the core entity of the archive is equipment, the preset equipment model display is used as the initial virtual interactive scene. The process of virtual interactive scene generation is based on the following data sources: archive image data (equipment drawings), video keyframe data (screenshots of the operation process), and video clip data (installation video); image data → generate scene maps through texture mapping (for example, converting drawings into operating table backgrounds); video keyframes → extract operation step nodes and generate interactive animations (for example, demonstrating the "screw tightening sequence"); video clips → split into sub-processes and embed them into the scene timeline (for example, click "Installation Phase" to play the corresponding clip).

[0034] S5: Based on the visitor information, recommend archive fragment data according to the archive fragment data value table, display the corresponding archive fragment data through the virtual reality interactive device, and provide real-time feedback on the visitor's browsing progress, and switch the virtual interactive scene of the process based on the visitor's browsing progress; The identity information of the visitor and the information of the department to which the visitor belongs; based on the department information of the visitor and in accordance with the archive fragment data value table, the archive fragment data with high value for the target department will be recommended, and the corresponding archive fragment data will be displayed through the virtual interactive device. The starting scene is the initial virtual interactive scene, and the virtual interactive scene will be switched according to the visitor's review process.

[0035] It should be noted that the employee ID and position (e.g., technical engineer) are obtained through the system login credentials, the target department is determined (e.g., visitor E1001 → Technical Department), and the archive fragment data with high value are recommended in descending order according to the archive fragment data value table. The archive fragment data is displayed through a virtual interactive device (e.g., a head-mounted VR device), and an initial virtual interactive scene (e.g., equipment model display) is constructed based on the core entity of the archive (e.g., "equipment D001"). The scene is switched according to the visitor's browsing process, for example, switching from a device structure diagram to a maintenance video clip.

[0036] S6: Record the visitor information, query content, query time and feedback information, generate an archive query backtracking table, and update the archive fragment data value table according to the visitor feedback information.

[0037] The archive fragment data value is recalculated based on the feedback information from the visitors at every preset statistical period, and the archive fragment data value table is updated based on the calculated archive fragment data value.

[0038] It should be noted that the preset statistical period can be obtained from historical experience and the archive access situation of each department. The feedback information from the visitor includes the visitor's value score for the archive fragment data visited. The archive fragment data value is recalculated based on the feedback information from all personnel in the department during the statistical period, and the archive fragment data value table of the department is updated.

[0039] Example 2: Based on Example 1, a digital file call management system based on artificial intelligence, such as Figure 2 Shown, including: Archive digitization module: collects multiple types of archival data, digitizes them using artificial intelligence technology, generates archival fragment data, and builds a digital archive based on the archival fragment data; Archive generation module: Based on the digital archive, using natural language processing technology, it processes the archive fragment data, generates an archive fragment data information table, mines the relationship between different types of data, and constructs a multimodal digital archive; Value calculation module: according to the archive fragment data information table, calculate the value of the archive fragment data for different departments using the archive fragment data value calculation formula, and generate an archive fragment data value table; Archive interaction module: constructs an initial virtual interaction scene and a process virtual interaction scene based on the multimodal digital archive; recommends archive fragment data according to the archive fragment data value table based on the visitor information, displays the corresponding archive fragment data through the virtual reality interaction device, and provides real-time feedback on the visitor's browsing progress, and switches the process virtual interaction scene based on the visitor's browsing progress; Record update module: records visitor information, query content, query time and feedback information, generates an archive query backtracking table, and updates the archive fragment data value table based on the visitor feedback information.

[0040] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A digital archive retrieval management method based on artificial intelligence, characterized in that: The following steps are involved: S1: Collect multi-type archival data, digitize them using artificial intelligence technology, generate archival fragment data, and build a digital archive based on the archival fragment data; S2: Based on the digital archive, using natural language processing technology to process the archive fragment data, generate an archive fragment data information table, explore the relationship between different types of data, and construct a multimodal digital archive; S3: Calculate the value of the archive fragment data for different departments using the archive fragment data value calculation formula according to the archive fragment data information table, and generate an archive fragment data value table; S4: constructing an initial virtual interaction scene and a process virtual interaction scene according to the multimodal digital archive; S5: Based on the visitor information, recommend archive fragment data according to the archive fragment data value table, display the corresponding archive fragment data through the virtual reality interactive device, and provide real-time feedback on the visitor's browsing progress, and switch the virtual interactive scene of the process based on the visitor's browsing progress; S6: Record the visitor information, query content, query time and feedback information, generate an archive query backtracking table, and update the archive fragment data value table according to the visitor feedback information.

2. The method for managing digital archives based on artificial intelligence according to claim 1, characterized in that: The collecting of multi-type archive data includes: the multi-type archive data is composed of text data, image data, audio data and video data of the same archive.

3. The method for managing digital archives based on artificial intelligence according to claim 2, characterized in that: The method of using artificial intelligence technology to digitize multiple types of archival data, generate archival fragment data, and construct a digital archive based on the archival fragment data includes: Use semantic segmentation technology to process text data and segment semantically coherent text fragments; Use image recognition technology to process image data and extract text fragment data; Use speech recognition technology to convert audio data into text snippet data; Using key frame extraction technology, the video data is divided into video segments with key frames as segmentation frames. The video key frames are processed by combining image recognition technology and speech recognition technology to generate description information of the video segments.

4. The method for managing digital archives based on artificial intelligence according to claim 1, characterized in that: The digital archive is based on the natural language processing technology to process the archive fragment data, generate an archive fragment data information table, mine the relationship between different types of data, and construct a multimodal digital archive, including: Perform knowledge extraction on the text segment data and video segment description information obtained by processing multi-type data, obtain the corresponding entity, relationship and attribute data and initial association relationship data, count the entity, relationship and attribute data of different departments, and generate the archive segment data information table; Construct semantic relationships between entities based on entity, relationship and attribute data and initial association relationship data, establish connections between text data, image data, audio data and video data, link different types of data together, and generate multimodal digital archives.

5. The method for managing digital archives based on artificial intelligence according to claim 1, characterized in that: The method of calculating the value of the archive fragment data for different departments using the archive fragment data value calculation formula according to the archive fragment data information table to generate the archive fragment data value table includes: the archive fragment data value calculation formula is: ； Where, The value of the archive fragment data to the target department, 、 and is the weight, The number of entities of interest to the target department included in the archive fragment data. is the total number of entities contained in the archive fragment data. The number of relationships that the target department is concerned about included in the archive fragment data, is the total number of relations contained in the archive fragment data, The number of attributes that the target department is concerned about contained in the archive fragment data. is the total number of attributes contained in the archive fragment data, is the time decay coefficient of the archive fragment data, The data storage time of the archive fragment, Provide feedback to interviewers on archival footage data.

6. The method for managing digital archives based on artificial intelligence according to claim 5, characterized in that: The interviewer's feedback on the archival footage data includes: ； Where, is the average rating of the interviewees on the value of the archive fragment data, The frequency of access to archive fragment data.

7. The method for managing digital archives based on artificial intelligence according to claim 1, characterized in that: The constructing of the initial virtual interaction scene and the process virtual interaction scene according to the multimodal digital archive includes: constructing the initial virtual interaction scene according to the archive core entity, and constructing the process virtual interaction scene according to the archive image data, video key frame data and video clip data.

8. The method for managing digital archives based on artificial intelligence according to claim 1, characterized in that: The method includes recommending archival fragment data according to the archival fragment data value table based on the visitor information, displaying the corresponding archival fragment data through a virtual reality interactive device, and providing real-time feedback on the visitor's review progress. The method switches the process virtual interaction scene based on the visitor's review progress, including: the visitor information includes: visitor identity information and visitor department information; according to the visitor department information, archival fragment data with high value for the target department is recommended according to the archival fragment data value table, and displaying the corresponding archival fragment data through a virtual interactive device. The starting scene is an initial virtual interaction scene, and the process virtual interaction scene is switched according to the visitor's review progress.

9. The method for managing digital archives based on artificial intelligence according to claim 1, characterized in that: The method records visitor information, query content, query time and feedback information, generates an archive query backtracking table, and updates the archive fragment data value table according to the visitor feedback information, including: recalculating the archive fragment data value according to the visitor feedback information at every preset statistical period, and updating the archive fragment data value table according to the calculated archive fragment data value.

10. A digital file call management system based on artificial intelligence, according to the digital file call management method based on artificial intelligence according to any one of claims 1 to 9, characterized in that: Also includes: Archive digitization module: collects multiple types of archival data, digitizes them using artificial intelligence technology, generates archival fragment data, and builds a digital archive based on the archival fragment data; Archive generation module: Based on the digital archive, using natural language processing technology, it processes the archive fragment data, generates an archive fragment data information table, mines the relationship between different types of data, and constructs a multimodal digital archive; Value calculation module: according to the archive fragment data information table, calculate the value of the archive fragment data for different departments using the archive fragment data value calculation formula, and generate an archive fragment data value table; Archive interaction module: constructs an initial virtual interaction scene and a process virtual interaction scene based on the multimodal digital archive; recommends archive fragment data according to the archive fragment data value table based on the visitor information, displays the corresponding archive fragment data through the virtual reality interaction device, and provides real-time feedback on the visitor's browsing progress, and switches the process virtual interaction scene based on the visitor's browsing progress; Record update module: records visitor information, query content, query time and feedback information, generates an archive query backtracking table, and updates the archive fragment data value table based on the visitor feedback information.

Images