A method and system for generating a depression intervention program based on a nostalgia guide

Abstract

The application discloses a nostalgic guide-based depression intervention scheme generation method and system, and belongs to the technical field of computer-aided psychological intervention. A nostalgic guide digital representation uploaded by a user terminal is acquired, entity object identifiers and scene context identifiers are deeply analyzed and extracted therefrom, the identifiers are matched with preset life period division axes, corresponding personal historical period labels are determined, a positive emotion arousal keyword set is recalled from an emotion association mapping table based on the labels, an intervention clue sequence containing multi-modal prompts is constructed, a structured intervention scheme having a progressive relationship is generated, and the structured intervention scheme is pushed to a user, and the order or content of unexecuted clues is dynamically adjusted according to real-time feedback signals after execution by the user. Through fine-grained semantic analysis of the nostalgic guide, accurate mapping of the personal historical period and a feedback-driven dynamic updating mechanism, the degree of individualization, the depth of semantic understanding and the clinical response flexibility of the depression intervention scheme are significantly improved.

Description

Technical Field

[0001] This invention discloses a method and system for generating depression intervention programs based on nostalgia guidance, belonging to the field of computer-aided psychological intervention technology. Background Technology

[0002] In recent years, computer-assisted intervention technologies for mood disorders such as depression have gradually gained attention. Nostalgia therapy, as a psychological intervention that guides individuals to recall positive past experiences to improve mood, has shown good results in clinical practice. However, traditional nostalgia interventions heavily rely on therapists' experience and the preparation of handmade materials. The intervention content is mostly general nostalgic themes, making it difficult to generate real-time, personalized intervention plans based on the specific nostalgic items or scenarios carried by each user. With the widespread use of mobile devices, users can record their personal nostalgic guides through photography, audio recording, etc., but existing computer-assisted intervention systems cannot effectively analyze these heterogeneous digital representations and automatically convert them into intervention steps with a temporal progression.

[0003] Currently, some technical solutions attempt to randomly push emotionally evocative materials using digital media libraries, or recommend nostalgic content with fixed themes after assessing users' emotional states through pre-set questionnaires. These methods generally suffer from the following drawbacks: First, they simply store user-uploaded nostalgic guides, lacking structured parsing capabilities for image foreground and background, audio semantics, and text syntax, making it impossible to accurately extract entity objects and scene context. Second, intervention plans are mostly generated based on static rule bases, failing to dynamically link individual user age with the historical popularity of items and scenes, resulting in coarse attribution to personal historical periods and misalignment between emotionally evocative keywords and users' actual experiences. Third, intervention clues are mostly organized in parallel, lacking a gradient ranking mechanism based on emotional arousal intensity, and once a plan is generated, the order of clues cannot be adjusted, content replaced, or keywords iterated based on real-time user feedback, resulting in a one-way output characteristic in the intervention process.

[0004] The aforementioned shortcomings result in significant deficiencies in existing computer-assisted nostalgia intervention systems regarding personalization, semantic understanding depth, and dynamic response flexibility, making it difficult to meet the practical needs of depression interventions that emphasize user-centeredness and dynamic process optimization. Therefore, there is a need for an intervention generation method capable of fine-grained analysis of nostalgia cues, precise mapping of individual historical periods, automatic orchestration of emotional arousal gradients, and support for feedback-driven program updates, thereby enhancing the intelligence and effectiveness of computerized nostalgia therapy. Summary of the Invention

[0005] To achieve the above objectives, this application provides the following technical solution: A method for generating a depression intervention program based on nostalgia guidance includes the following steps: 101: Obtain the digital representation of the nostalgic guide uploaded by the user's interactive terminal, wherein the digital representation of the nostalgic guide includes image files, audio files, or text descriptions; 102: Analyze the digital representation of the nostalgic guide and extract the entity object identifiers and scene context identifiers contained therein; 103: Based on the entity object identifier and the scene context identifier, match the corresponding personal historical period label on the preset life time period division axis; 104: Based on the personal historical period tags, retrieve the set of positive emotional arousal keywords associated with that period from the emotion association mapping table; 105: Based on the aforementioned set of positive emotion arousal keywords, construct an intervention cue sequence that includes narrative cues, sensory stimulation cues, and behavioral suggestion cues; 106: Sort the intervention cue sequence according to the emotional arousal gradient to generate a nostalgia intervention step arrangement with a time progression relationship; 107: The nostalgia intervention steps are arranged and packaged into a structured intervention plan data package and pushed to the user interaction terminal; 108: In response to the solution execution feedback signal returned by the user interaction terminal, adjust the order of unexecuted clues in the intervention clue sequence or replace the clue content, rearrange the adjusted intervention clue sequence and update the intervention solution data package.

[0006] Furthermore, the image file of the digital representation of the nostalgic guide in 101 is obtained through the following method: Receive photos of physical vintage items taken by a user interaction terminal, extract the shooting timestamp and geolocation tag of the photos, and append the shooting timestamp and geolocation tag to the metadata field of the image file.

[0007] Furthermore, the specific steps in step 102 for parsing the digital representation of the nostalgic guide and extracting entity object identifiers and scene context identifiers are as follows: 1021: If the digital representation of the nostalgic guide is an image file, then the image file is segmented to identify the foreground subject outline and background environment outline in the image; 1022: Compare the outline of the foreground subject with the outline features of a pre-set library of nostalgic items, and extract the names of items with a matching degree exceeding a threshold as entity object identifiers; 1023: Compare the background environment outline with the features of the pre-set scene type library to identify the architectural style, natural landscape or interior layout type in the environmental elements and generate scene context identifiers; 1024: If the nostalgic guide digital representation is an audio file, then the audio file is transcribed into speech, and noun phrases are extracted from the transcribed text as entity object identifiers, and noun phrases representing locations are extracted as scene context identifiers. 1025: If the nostalgic guide's digital representation is a text description, then perform syntactic dependency analysis on the text description, extract the subject headword as the entity object identifier, and extract the place words in the adverbial or complement as the scene context identifier.

[0008] Furthermore, the specific steps in step 103 for matching corresponding personal historical period tags on a preset life time segmentation axis are as follows: 1031: Combine entity object identifiers with scene context identifiers into composite search criteria; 1032: Query the constructed nostalgia mapping index table using composite search criteria. This index table contains an item-scene combination field and the corresponding era field. 1033: When a composite search condition matches the item-scene combination field in the index table, extract the corresponding era stage field value as a candidate period label; 1034: If 1033 does not find any record, then query the entity object identifier and the scene context identifier as single search conditions, and take the intersection of the era stage field values ​​that they hit. If the intersection is not empty, then take the intersection as the candidate period label. 1035: Map the candidate period labels to the life period division axis. The life period division axis is pre-divided into five segments according to age range: childhood, adolescence, youth, middle age and old age. Determine the age segment to which the candidate period label belongs based on the year range indicated by the candidate period label, and output the name of the age segment as the personal historical period label.

[0009] Furthermore, the 104 emotion association mapping tables are constructed and invoked in the following manner: Pre-configure at least three positive emotional evoking keywords for each individual's historical period tag. These keywords are adjective phrases that describe feelings of pleasure, pride, emotion, or tranquility. Based on the personal historical period tags output by 103, perform a hash lookup in the emotion association mapping table to extract all positive emotion arousal keywords corresponding to the tag, and sort them in descending order according to the frequency of user responses triggered by the keywords in historical interaction data. Take the top five keywords as the set of positive emotion arousal keywords.

[0010] Furthermore, the specific steps for constructing the intervention cue sequence in step 105 are as follows: 1051: Select one keyword from the set of positive emotional arousal keywords as the core emotional tone of the current cue; 1052: Based on the core emotional tone, nostalgic narrative short articles that match the personal historical period tags and are emotionally oriented with the keyword are retrieved from the narrative template library to form narrative prompts; 1053: Based on the core emotional tone, retrieve image or audio samples associated with the entity object identifier or scene context identifier from the multimedia material library to form a sensory stimulation cue; 1054: Based on the core emotional tone, extract simple physical activities or creative expression task descriptions that match the personal historical period labels from the behavior suggestion rule base to form behavior suggestion prompts; 1055: Combine narrative cues, sensory stimulus cues, and behavioral suggestion cues in a fixed order to form an intervention cue unit. Repeat steps 1051 to 1054 until all keywords in the positive emotional arousal keyword set have been processed, resulting in multiple intervention cue units. The set forms an intervention cue sequence.

[0011] Furthermore, the specific steps for sorting according to the emotional arousal gradient in step 106 are as follows: Each intervention cue unit is assigned an arousal intensity coefficient, which is determined based on the level of the positive emotional arousal keywords used in that unit in a preset intensity level table. The intervention cue units in the intervention cue sequence are arranged from low to high according to the arousal intensity coefficient, forming a progressive structure in which emotional arousal transitions from gentle to intense. The arranged intervention clue units are assigned sequential numbers, and these sequential numbers serve as the basis for the temporal progression of the nostalgia intervention steps.

[0012] Furthermore, the specific steps in step 108 for adjusting the intervention cue sequence in response to the scheme execution feedback signal returned by the user interaction terminal are as follows: 1081: Receive the solution execution feedback signal returned by the user interaction terminal. The feedback signal includes the user's satisfaction rating and emotion self-rating scale score for the completed intervention cue unit. 1082: Compare the satisfaction score and the score of the emotion self-rating scale with the preset satisfaction threshold and emotion improvement threshold respectively to generate an execution effect level label, wherein the execution effect level label includes high approval, average approval and low approval; 1083: If the execution effect level label is highly approved, the original order and content of the unexecuted intervention cue units remain unchanged; if the execution effect level label is moderately approved, the positive emotional arousal keywords used by the intervention cue unit corresponding to the feedback signal are identified, other units using the same keywords are found in the unexecuted intervention cue units, and the sorting position of the other unit is adjusted forward by one position; if the execution effect level label is lowly approved, the positive emotional arousal keywords used by the intervention cue unit corresponding to the feedback signal are identified, the keywords are deleted from the set of positive emotional arousal keywords, and according to 104, unused positive emotional arousal keywords are retrieved again from the emotion association mapping table, and subsequent intervention cue units are regenerated according to 105 to 106 and the original unexecuted units are replaced. 1084: Integrate the adjusted sequence of intervention clue units with the replaced unit content to form an adjusted sequence of intervention clues.

[0013] Furthermore, after rearranging the adjusted intervention cue sequence and updating the intervention protocol data package in step 108, the following steps are also included: Recalculate the mean and standard deviation of the arousal intensity coefficients of all intervention cue units in the updated intervention protocol data package, and write the mean and standard deviation into the metadata field of the intervention protocol data package; Generate a version change record, which includes the timestamp of this adjustment, the number of leads before the adjustment, the number of leads after the adjustment, and the execution effect level label field. Append this version change record to the end of the intervention plan data package.

[0014] Furthermore, the personal history period tags, positive emotion arousal keyword sets, intervention clue sequences, and intervention program data packages used in 101 to 108 are all transmitted between the user interaction terminal and the server in encrypted JSON data format. During the transmission process, the entity object identifiers and scene context identifiers are desensitized by replacing the specific names in the identifiers with the generic names of the categories to which those names belong.

[0015] According to a second aspect of the present invention, the present invention claims protection for a depression intervention program generation system based on nostalgia guidance, comprising: One or more processors; A memory storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the aforementioned method for generating a depression intervention program based on nostalgia guidance.

[0016] This invention discloses a method and system for generating depression intervention programs based on nostalgia-guided prompts, belonging to the field of computer-aided psychological intervention technology. The method involves acquiring digital representations of nostalgia-guided prompts uploaded by the user's terminal, deeply analyzing and extracting entity object identifiers and scene context identifiers, matching these identifiers with preset life time period division axes to determine corresponding personal historical period tags, and retrieving a set of positive emotional arousal keywords from an emotion association mapping table based on these tags. This constructs an intervention cue sequence containing multimodal cues, generates a structured intervention program with a progressive relationship, and pushes it to the user. The order or content of unexecuted cues is dynamically adjusted based on real-time feedback signals after user execution. This invention significantly improves the personalization, semantic understanding depth, and clinical response flexibility of depression intervention programs through fine-grained semantic analysis of nostalgia-guided prompts, precise mapping of personal historical periods, and a feedback-driven dynamic update mechanism. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for generating a depression intervention program based on nostalgia guidance, as claimed in an embodiment of the present invention. Figure 2 A flowchart illustrating the matching of personal historical period tags in a method for generating a depression intervention program based on nostalgia guidance, as claimed in an embodiment of the present invention. Figure 3 The flowchart illustrates the construction of an intervention cue sequence for a method of generating a depression intervention program based on nostalgia as claimed in an embodiment of the present invention. Figure 4 The flowchart illustrates the process of adjusting the intervention cue sequence using the execution feedback signal of a method for generating a depression intervention program based on nostalgia guidelines, as claimed in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0019] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of those features. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications in the embodiments of this application, such as up, down, left, right, front, back, etc., are only used to explain the relative positional relationships and movements between components in a specific orientation as shown in the accompanying drawings. If the specific orientation changes, the directional indications will change accordingly. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0020] References to embodiments herein mean that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0021] According to a first embodiment of the present invention, the present invention claims protection for a method for generating a depression intervention program based on nostalgia guidance, referring to... Figure 1 This includes the following steps: 101: Obtain the digital representation of the nostalgic guide uploaded by the user's interactive terminal, wherein the digital representation of the nostalgic guide includes image files, audio files, or text descriptions; 102: Analyze the digital representation of the nostalgic guide and extract the entity object identifiers and scene context identifiers contained therein; 103: Based on the entity object identifier and the scene context identifier, match the corresponding personal historical period label on the preset life time period division axis; 104: Based on the personal historical period tags, retrieve the set of positive emotional arousal keywords associated with that period from the emotion association mapping table; 105: Based on the aforementioned set of positive emotion arousal keywords, construct an intervention cue sequence that includes narrative cues, sensory stimulation cues, and behavioral suggestion cues; 106: Sort the intervention cue sequence according to the emotional arousal gradient to generate a nostalgia intervention step arrangement with a time progression relationship; 107: The nostalgia intervention steps are arranged and packaged into a structured intervention plan data package and pushed to the user interaction terminal; 108: In response to the solution execution feedback signal returned by the user interaction terminal, adjust the order of unexecuted clues in the intervention clue sequence or replace the clue content, rearrange the adjusted intervention clue sequence and update the intervention solution data package.

[0022] This embodiment uses Mr. Zhang, a middle-aged male user in a state of mild to moderate depression who has a preference for reminiscing but lacks the habit of actively recalling memories, as the intervention subject to fully illustrate the implementation process of the depression intervention program generation method based on nostalgia-guided materials. Mr. Zhang used a dedicated intervention application installed on his personal smartphone to take a photo of a sewing machine that his mother used when she was young as a nostalgia-guided material.

[0023] The application receives a photo of a sewing machine taken by Zhang. Based on the file header identifier, it determines that the file is a JPEG image. It then encapsulates this image file, along with the automatically acquired timestamp and geolocation information, into HTTP multipart form data and uploads it to the backend intervention scheme generation server via an encrypted transmission channel. Upon receiving the image, the server stores it in a temporary working directory and assigns it a unique session identifier.

[0024] The server-side calls the image parsing module to process the photo. First, the sewing machine in the foreground is separated from the old-fashioned wooden floor and corner lines in the background using an image segmentation algorithm. The foreground contour features are extracted and matched with the height of the treadle sewing machine template in the nostalgic item contour library, and the output entity object is identified as a sewing machine. The texture features of the background are extracted and matched with the interior scenes of residential buildings from the 1980s and 1990s in the scene feature library, and the output scene context is identified as an old-fashioned residential living room.

[0025] The query uses a combination of the physical identifier "sewing machine" and the scene identifier "old-fashioned living room" as a composite condition to retrieve a nostalgia mapping index. The index records show that sewing machines were widely popular between 1975 and 1990, while old-fashioned living rooms were a typical interior style between 1980 and 1995. The overlap between these two periods is 1980 to 1990. The user profile shows that Zhang was born in 1970, and the overlap of 1980 to 1990 corresponds to his age of 10 to 20, thus mapping to adolescence as a personal history period tag.

[0026] Using adolescence as the key, an emotion association mapping table is retrieved. This table pre-sets several positive emotion words for each period, and each word is accompanied by a historical trigger frequency counter. The retrieved list of associated words for that period includes pride, warmth, curiosity, being cared for, and a sense of accomplishment. After sorting by frequency, the top five words are selected to form a set of positive emotion-evoking keywords, specifically pride, warmth, being cared for, a sense of accomplishment, and curiosity.

[0027] Iterate through the keyword set above, constructing intervention cue units for each keyword in turn. Taking pride as an example, retrieve a narrative template matching adolescence and the emotion of pride from the narrative template library. The template content is: "Do you remember the scene when your family praised you for doing a good job the first time you mastered {entity}?" Replace the placeholder {entity} with the entity identifier sewing machine to generate a narrative cue. Retrieve audio materials related to sewing machines or old-fashioned living rooms from the multimedia material library. Find a popular light music track from 1985 to generate a sensory stimulus cue. Retrieve behavioral suggestions matching adolescent pride from the behavioral suggestion rule library. Imitate the small skill you learned at that time, even if it's just simple threading a needle, to generate a behavioral suggestion cue. Combine these three elements with the current keyword pride to form an intervention cue unit. After processing five keywords in this way, obtain an intervention cue sequence containing five units.

[0028] Each intervention cue unit was assigned an emotional arousal intensity value. By consulting a keyword intensity mapping table, pride corresponded to an intensity value of 4, warmth to 3, being cared for to 3, a sense of accomplishment to 4, and curiosity to 2. The units were arranged in ascending order of intensity value, resulting in the sequence: Curiosity 2 → Warmth 3 → Being Cared For 3 → Pride 4 → Achievement 4. Each unit was then assigned step numbers 1 to 5, forming the nostalgia intervention step sequence.

[0029] The above-mentioned arrangement sequence is encapsulated into a JSON-formatted structured data packet, containing fields such as step number, narrative text, audio Uniform Resource Identifier, and behavior suggestion text. It is then sent to Zhang's mobile application via push service and presented in the form of a card list.

[0030] Mr. Zhang followed the intervention steps sequentially. After completing the second step, "warmth," the application displayed a satisfaction rating and an emotional self-assessment scale. Mr. Zhang rated the feedback as "neutral." The server received the feedback signal, parsed the corresponding step number 2, identified the keyword used in this step as "warmth," and determined the feedback label to be "neutral approval." Subsequently, it searched for other units using the same keyword "warmth" in the unexecuted sequence steps 3 to 5. No other units were found, so the backup logic for sorting adjustment in the "neutral approval" processing was executed: the order of all subsequent unexecuted units remained unchanged, and only this feedback was recorded. If a unit with the same keyword existed here, it was moved forward one position. Since there were no units with the same keyword, the original sequence was maintained. The server repackaged the result of this feedback processing—that is, maintaining the original sequence—into an updated intervention plan data package and pushed it to the terminal again, while simultaneously writing information such as the sequence length and feedback label into the metadata.

[0031] Furthermore, the image file of the digital representation of the nostalgic guide in 101 is obtained through the following method: Receive photos of physical vintage items taken by a user interaction terminal, extract the shooting timestamp and geolocation tag of the photos, and append the shooting timestamp and geolocation tag to the metadata field of the image file.

[0032] In this embodiment, when Zhang takes a photo of the sewing machine, the application automatically calls the system time service to convert the local time of the shooting moment into a string formatted as 2025-04-07 14:32:18 and writes it into the creation time tag of the photo metadata. Simultaneously, it calls the Global Positioning System module to obtain the current location's latitude and longitude, and parses it using an offline reverse geocoding database to determine the address as "XYZ, China, Hangzhou, Xihu District, Xihu District, Xihu District," writing this address string into the geographic location tag of the photo metadata. After receiving this information, the server explicitly reads these two pieces of information from the metadata and stores them in association with the image file itself. In subsequent 103 matching of personal historical periods, if multiple candidate period tags exist in the entity and scene retrieval results, the server prioritizes the year indicated by the shooting timestamp as an auxiliary verification condition: if the difference between the shooting timestamp year and the representative year of a candidate period tag is less than five years, that tag is selected first. In this embodiment, the photo was taken in 2025. The sewing machine in the photo is obviously a vintage item rather than a modern one. Therefore, the timestamp does not have a substantial impact on the determination of the period. However, this mechanism provides the correct age correction capability for future users when taking photos of newly acquired vintage items.

[0033] Furthermore, the specific steps in step 102 for parsing the digital representation of the nostalgic guide and extracting entity object identifiers and scene context identifiers are as follows: 1021: If the digital representation of the nostalgic guide is an image file, then the image file is segmented to identify the foreground subject outline and background environment outline in the image; 1022: Compare the outline of the foreground subject with the outline features of a pre-set library of nostalgic items, and extract the names of items with a matching degree exceeding a threshold as entity object identifiers; 1023: Compare the background environment outline with the features of the pre-set scene type library to identify the architectural style, natural landscape or interior layout type in the environmental elements and generate scene context identifiers; 1024: If the nostalgic guide digital representation is an audio file, then the audio file is transcribed into speech, and noun phrases are extracted from the transcribed text as entity object identifiers, and noun phrases representing locations are extracted as scene context identifiers. 1025: If the nostalgic guide's digital representation is a text description, then perform syntactic dependency analysis on the text description, extract the subject headword as the entity object identifier, and extract the place words in the adverbial or complement as the scene context identifier.

[0034] In this embodiment, for the sewing machine image uploaded by Zhang, the image is converted into a grayscale matrix, and the segmentation threshold is determined using the Otsu's method to obtain the foreground mask. Connected component labeling is performed on the mask, and the connected component with the largest area is retained as the sewing machine outline. The outline edge point set is extracted, and after thinning using the Douglas-Puk algorithm, 47 feature points remain. The Hausdorff distance is calculated for each of these point sets with templates such as sewing machines, bicycles, radios, and clocks in the nostalgic item outline library. The point with the smallest distance, which is below a preset first threshold, is output as the entity identifier "sewing machine". HSV color histograms and local binary pattern texture features are extracted from the background region and concatenated into a 576-dimensional feature vector. Euclidean distance is calculated between this vector and descriptors such as old-fashioned living rooms, old-fashioned kitchens, old-fashioned bedrooms, and school playgrounds in the scene feature library. The point with the smallest distance, which is below a second threshold, is output as the scene identifier "old-fashioned residential living room".

[0035] Secondly, suppose another user, Li, uploads an audio file containing a fragment of a dialect nursery rhyme hummed by his grandmother. The server reads the audio stream, removes the first and last silent segments through speech activity detection, and sends it to an automatic speech recognition engine to convert it into text: moon, grandmother, light the lamp, I, do, brother, you, sit; after word segmentation based on a dictionary, auxiliary words and modal particles are filtered out, retaining the nouns moon, grandmother, light the lamp, and brother; after frequency statistics, the top three words, moon, grandmother, and brother, are selected as entity objects; the text is scanned for location nouns, and if no match is found, an unrecognized location placeholder is output.

[0036] Next, suppose user Wang directly inputs the text description: "When I was little, there was an osmanthus tree in my grandmother's yard, and the yard was filled with fragrance in autumn." Dependency parsing of this text yields the core predicates: subject headword "yard," object headword "osmanthus tree." The modifier "grandmother's house" is extracted from the attributive clause governing "yard." Locative word retrieval reveals that "yard" is an adverbial phrase of location, outputting the scene identifier "yard." The entity identifier set contains "osmanthus tree," "grandmother," and "yard." All the above parsing results are then entered into the subsequent 103-period matching process.

[0037] Furthermore, referring to Figure 2 The specific steps for matching the corresponding personal historical period tags on the preset life time division axis in step 103 are as follows: 1031: Combine entity object identifiers with scene context identifiers into composite search criteria; 1032: Query the constructed nostalgia mapping index table using composite search criteria. This index table contains an item-scene combination field and the corresponding era field. 1033: When a composite search condition matches the item-scene combination field in the index table, extract the corresponding era stage field value as a candidate period label; 1034: If 1033 does not find any record, then query the entity object identifier and the scene context identifier as single search conditions, and take the intersection of the era stage field values ​​that they hit. If the intersection is not empty, then take the intersection as the candidate period label. 1035: Map the candidate period labels to the life period division axis. The life period division axis is pre-divided into five segments according to age range: childhood, adolescence, youth, middle age and old age. Determine the age segment to which the candidate period label belongs based on the year range indicated by the candidate period label, and output the name of the age segment as the personal historical period label.

[0038] This embodiment fully presents multiple branch paths for period tag matching.

[0039] Scenario 1: Combination Hit. In Zhang's case, the physical sewing machine and the scenes "old-fashioned house" and "living room" are not placeholders. The combination key "sewing machine|old-fashioned house living room" directly hits the nostalgia mapping index table. The index table pre-stores the period label corresponding to this combination as adolescence, so it is directly output without calculation.

[0040] Scenario 2: Retrieve the intersection separately. Suppose a user uploads a photo of a military water bottle with a school playground in the background. The water bottle dates from 1965 to 1985, and the school playground dates from 1970 to 1990. The intersection is 1970 to 1985. The user was born in 1960, so the intersection represents the year 1977, corresponding to age 17, which is considered adolescence.

[0041] Scenario 3: One-sided hit. If the physical mobile phone is between 1990 and 2000, and the beach scene has no date record, then the median of the physical range, 1995, is taken. The user was born in 1970 and is 25 years old, so they are judged to be in their youth.

[0042] Scenario 4: No records or no intersection. The user uploads a modern selfie. The latest model of the physical mobile phone has no historical record, and the café scene has no clear start or end point of popularity. Therefore, the user's actual age is calculated based on the current year and the year of birth. The user's age is 55 years old, which is considered old age.

[0043] In this embodiment, all date calculations use the system's standard time function, and the birth year is read from the profile filled in by the user during registration, ensuring that the period tags strictly correspond to the user's actual life course.

[0044] Furthermore, the 104 emotion association mapping tables are constructed and invoked in the following manner: Pre-configure at least three positive emotional evoking keywords for each individual's historical period tag. These keywords are adjective phrases that describe feelings of pleasure, pride, emotion, or tranquility. Based on the personal historical period tags output by 103, perform a hash lookup in the emotion association mapping table to extract all positive emotion arousal keywords corresponding to the tag, and sort them in descending order according to the frequency of user responses triggered by the keywords in historical interaction data. Take the top five keywords as the set of positive emotion arousal keywords.

[0045] In this embodiment, the process of constructing and calling the emotion association mapping table is as follows: In the initial construction phase, psychology experts entered 8 to 12 positive emotion words for each of the five period tags: childhood, adolescence, youth, middle age, and old age, with the initial trigger frequency of each word set to 0. The system periodically scans user feedback logs, and whenever a user rates an intervention cue unit as satisfied or its emotion score exceeds a threshold, the trigger frequency of the corresponding keyword for that unit is incremented by 1.

[0046] In this embodiment, after Zhang completed the "pride" keyword unit, he gave feedback of high approval, and the system increased the "pride" trigger frequency by 1. When 104 searches using the "adolescence" tag, the frequency of "pride" in the corresponding list of the emotion association mapping table is higher than other words, and it ranks first after sorting. The top five words are output as a set of positive emotion arousal keywords. If there are fewer than five related words for a certain period tag, all existing words are output without any padding or completion. In this embodiment, there are 12 related words for adolescence. After frequency sorting, the top five are "pride," "warmth," "being cared for," "sense of accomplishment," and "curiosity." The output order of the set is strictly descending according to frequency, and if the frequencies are the same, they are ascending according to the alphabetical order of the word's pinyin.

[0047] Furthermore, referring to Figure 3 The specific steps for constructing the intervention cue sequence in step 105 are as follows: 1051: Select one keyword from the set of positive emotional arousal keywords as the core emotional tone of the current cue; 1052: Based on the core emotional tone, nostalgic narrative short articles that match the personal historical period tags and are emotionally oriented with the keyword are retrieved from the narrative template library to form narrative prompts; 1053: Based on the core emotional tone, retrieve image or audio samples associated with the entity object identifier or scene context identifier from the multimedia material library to form a sensory stimulation cue; 1054: Based on the core emotional tone, extract simple physical activities or creative expression task descriptions that match the personal historical period labels from the behavior suggestion rule base to form behavior suggestion prompts; 1055: Combine narrative cues, sensory stimulus cues, and behavioral suggestion cues in a fixed order to form an intervention cue unit. Repeat steps 1051 to 1054 until all keywords in the positive emotional arousal keyword set have been processed, resulting in multiple intervention cue units. The set forms an intervention cue sequence.

[0048] In this embodiment, the complete process of constructing the intervention cue sequence is as follows: Create an empty array L in the server's memory; Take the first keyword "proud" sequentially from the keyword set; The narrative template library is stored as XML files, with each template having a `period` attribute and an `emotion` attribute. The server executes an XPath query ` / templates / template[@period='teenage' and @emotion='pride']`, finding a unique template with the text: "Remember the surprised looks from those around you when you first independently manipulated {entity}? That was your shining moment." Replacing {entity} with the entity identifier "sewing machine" generates the narrative hint T1. The multimedia resource library uses a CSV index to record the mapping relationship between items, scenes, and resource paths. The server reads the index, retrieves the row that matches both the item "sewing machine" and the scene "old-fashioned living room," and finds a record with the resource path " / audio / 80s_sewing_machine_ambient.mp3." This is encapsulated as a Uniform Resource Identifier and used to generate a sensory cue R1. The behavior suggestion rule base is in JSON array format. Each rule contains the fields period, keyword, action, and duration. Rules with the period "adolescence" and keyword "pride" are filtered, and the rule with the most detailed action description is selected. The action description is: "Find an old object in your home and try to polish it, just like you did when you first maintained it." The duration is a suggested time of 5 minutes. After replacing the placeholders, behavior suggestion A1 is generated. Create a structure U1 containing T1, R1, A1, and the keyword "proud", and append it to array L.

[0049] Repeat the above steps to process warmth, care, sense of accomplishment, and curiosity, resulting in five structures U2 to U5. The array L is now in its original order, processed according to the keyword order. Output arrays L to 106.

[0050] Furthermore, the specific steps for sorting according to the emotional arousal gradient in step 106 are as follows: Each intervention cue unit is assigned an arousal intensity coefficient, which is determined based on the level of the positive emotional arousal keywords used in that unit in a preset intensity level table. The intervention cue units in the intervention cue sequence are arranged from low to high according to the arousal intensity coefficient, forming a progressive structure in which emotional arousal transitions from gentle to intense. The arranged intervention clue units are assigned sequential numbers, and these sequential numbers serve as the basis for the temporal progression of the nostalgia intervention steps.

[0051] In this embodiment, the emotion arousal gradient ranking is implemented as follows: When the system starts, it loads a keyword intensity mapping table, which resides in memory as a hash structure. The table stores the intensity values: curiosity (2), warmth (3), being cared for (3), pride (4), and a sense of accomplishment (4).

[0052] When 106 is executed, the array L is traversed, and the keyword fields from U1 to U5 are read sequentially. The intensity value is obtained through hash lookup and written into the newly added intensity field of each structure. At this time, the intensity of U1 is 4, the intensity of U2 is 3, the intensity of U3 is 3, the intensity of U4 is 4, and the intensity of U5 is 2.

[0053] Sort array L in stable ascending order, with the intensity field as the sorting key. The sorting process preserves the original relative order of elements with the same intensity value. In the original order, U2 (Warmth) appeared before U3 (Care), both with an intensity of 3. After sorting, U2 remains before U3. The resulting array order is: U5 (Curiosity) 2, U2 (Warmth) 3, U3 (Care) 3, U1 (Pride) 4, U4 (Achievement) 4.

[0054] Starting from 1, assign a sequential number to each element of the sorted array and write it into the step_id field. The final intervention steps are arranged as follows: Step 1 Curiosity, Step 2 Warmth, Step 3 Being Cared For, Step 4 Pride, Step 5 Sense of Accomplishment.

[0055] Furthermore, referring to Figure 4 The specific steps in step 108 for adjusting the intervention cue sequence in response to the scheme execution feedback signal returned by the user interaction terminal are as follows: 1081: Receive the solution execution feedback signal returned by the user interaction terminal. The feedback signal includes the user's satisfaction rating and emotion self-rating scale score for the completed intervention cue unit. 1082: Compare the satisfaction score and the score of the emotion self-rating scale with the preset satisfaction threshold and emotion improvement threshold respectively to generate an execution effect level label, wherein the execution effect level label includes high approval, average approval and low approval; 1083: If the execution effect level label is highly approved, the original order and content of the unexecuted intervention cue units remain unchanged; if the execution effect level label is moderately approved, the positive emotion arousal keywords used by the intervention cue unit corresponding to the feedback signal are identified, other units using the same keywords are found in the unexecuted intervention cue units, and the sorting position of the other unit is adjusted forward by one position; if the execution effect level label is lowly approved, the positive emotion arousal keywords used by the intervention cue unit corresponding to the feedback signal are identified, the keywords are deleted from the set of positive emotion arousal keywords, and according to 104, unused positive emotion arousal keywords are retrieved again from the emotion association mapping table, and subsequent intervention cue units are regenerated according to 105 to 106 and the original unexecuted units are replaced. 1084: Integrate the adjusted sequence of intervention clue units with the replaced unit content to form an adjusted sequence of intervention clues.

[0056] In this embodiment, the feedback-driven adjustment mechanism is implemented in detail as follows: After completing step 1, Zhang showed a satisfaction score of 4.8 and an emotion score of 4.2, both higher than the preset satisfaction threshold of 4.0 and emotion threshold of 3.0, generating a "high approval" label. Step 1088 was executed, maintaining the original order of subsequent steps 2 to 5 without any adjustments.

[0057] After completing step 2 ("warmth"), Zhang gave a satisfaction score of 2.5 and an emotion score of 2.8, meeting the acceptable criteria. The score was below the satisfaction threshold but above half of it. Execution 1086: Locate the current cursor Cur=2, and search array L for cells in the keyword field equal to "warmth" starting from index 2 (step 3). The search revealed that the keywords for step 3 were "being cared for," for step 4 "pride," and for step 5 "sense of accomplishment," with no matching cells. Therefore, no swap operation was performed, and the subsequent sequence remained unchanged; only the feedback was recorded.

[0058] After Zhang completed step 3 and received care, his satisfaction score was 1.8 and his emotional score was 1.5, which met the low approval criteria and was below half the satisfaction threshold. Step 1087: Remove "being cared for" from the positive emotion keyword set C. At this point, the remaining keywords in C, such as pride, warmth, sense of accomplishment, and curiosity, have already been processed but are still in the set; deletion does not affect the already processed parts. Step 104 is re-executed with the adolescent tag. The most frequent word from the list of words not deleted is selected from the corresponding list in the emotion association mapping table. "Pleasant" is a candidate and has not been used; its frequency is similar to "curiosity." "Pleasant" is selected as a supplementary word and added to the end of C.

[0059] Clear elements from index 3 onwards in array L (i.e., steps 4 and onwards). Using the new keyword set C as input, re-execute steps 105 to 106: Construct a new intervention cue unit for "pleasure," and sort it by intensity along with the original "pride" and "achievement" cues, as well as the previously executed "curiosity" and "warmth." The new sequence is: Curiosity 2—Warmth 3—Pleasure 3—Pride 4—Achievement 4. Overwrite the original array L from index 3 to the end with the new sequence, and output the adjusted sequence.

[0060] Furthermore, after rearranging the adjusted intervention cue sequence and updating the intervention protocol data package in step 108, the following steps are also included: Recalculate the mean and standard deviation of the arousal intensity coefficients of all intervention cue units in the updated intervention protocol data package, and write the mean and standard deviation into the metadata field of the intervention protocol data package; Generate a version change record, which includes the timestamp of this adjustment, the number of leads before the adjustment, the number of leads after the adjustment, and the execution effect level label field. Append this version change record to the end of the intervention plan data package.

[0061] This embodiment details the additional operations performed when updating the intervention plan data package: Receive the adjusted sequence array L_adjusted from the 108 output. At this time, the sequence contains five units with intensity values ​​of 2, 3, 3, 4, and 4 respectively. Iterate through L_adjusted and store the intensity values ​​in a list ValList; Calculate the arithmetic mean of the ValList; the sum of the five values ​​is 16, divided by 5, we get 3.2, rounded to two decimal places, 3.20; Calculate the square of the difference between each value and the mean: 2-3.2²=1.44, 3-3.2²=0.04, 3-3.2²=0.04, 4-3.2²=0.64, 4-3.2²=0.64, sum = 2.8, divide by 5 to get 0.56, take the square root to get 0.75, round to two decimal places: 0.75; Add the key-value pair mean_intensity:3.20, std_intensity:0.75 to the metadata sub-object of the data packet JSON object; Get the server's current UNIX timestamp 1744051200 and convert it to ISO 8601 format 2025-04-08T10:00:00Z; Read the original sequence length OldLen=5 from the original data packet, and the current new sequence length NewLen=5; The feedback tag generated by reading 108 indicates low approval. Create a version change history structure containing timestamp 2025-04-08T10:00:00Z, OldLen=5, NewLen=5, FeedbackLabel=low_acceptance, serialize it into a JSON string, Base64 encoded it, and append it to the end of the data packet JSON object with the key name version_history.

[0062] The updated data package fully records the statistical characteristics and version trajectory of this adjustment, which can be used for subsequent analysis.

[0063] Furthermore, the personal history period tags, positive emotion arousal keyword sets, intervention clue sequences, and intervention program data packages used in 101 to 108 are all transmitted between the user interaction terminal and the server in encrypted JSON data format. During the transmission process, the entity object identifiers and scene context identifiers are desensitized by replacing the specific names in the identifiers with the generic names of the categories to which those names belong.

[0064] In this embodiment, before generating the intervention plan data package, the server inputs the entity identifier "sewing machine" and the scene identifier "old-fashioned residential living room" extracted by 102 into the desensitization module. The desensitization module loads a category generalization dictionary containing entries such as "sewing machine: household sewing equipment" and "old-fashioned residential living room: living space." After a perfect match, the entity identifier is replaced with "household sewing equipment," and the scene identifier is replaced with "living space."

[0065] In the subsequent replacement of the 105 narrative prompt placeholder, the narrative text becomes: Remember when you first operated {entity} independently... actually filling the household sewing equipment.

[0066] If the user uploads an audio file and the parsing reveals the entities "moon" and "grandmother" that are not included in the dictionary, the desensitization module will add a category suffix to the end of the string and output items of the categories "moon" and "grandmother".

[0067] All encapsulated intervention plan data packets are JSON formatted strings, encrypted using the TLS 1.3 protocol at the transport layer, with the server's private key stored in a hardware security module, and the terminal certificate verified bidirectionally. Each data packet is appended with a server digital signature using the ECDSA with P-256 algorithm; the receiving end can only parse the packet after verifying the signature.

[0068] The original photos, audio, and text descriptions uploaded by the user terminal are immediately deleted from the temporary working directory after 102 parsing is completed. Only the de-identified identifiers and necessary metadata are retained for subsequent steps. The original digital representation of the nostalgic guide is not persistently stored on the server to minimize the risk of user privacy leakage.

[0069] According to a second embodiment of the present invention, the present invention claims protection for a depression intervention program generation system based on nostalgia guidance, comprising: One or more processors; A memory storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the aforementioned method for generating a depression intervention program based on nostalgia guidance.

[0070] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0071] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0072] The specific embodiments of the invention have been described in detail above, but they are only examples, and this application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. A method for generating a depression intervention program based on nostalgia guidance, characterized in that, Includes the following steps: 101: Obtain the digital representation of the nostalgic guide uploaded by the user's interactive terminal, wherein the digital representation of the nostalgic guide includes image files, audio files, or text descriptions; 102: Analyze the digital representation of the nostalgic guide and extract the entity object identifiers and scene context identifiers contained therein; 103: Based on the entity object identifier and the scene context identifier, match the corresponding personal historical period label on the preset life time period division axis; 104: Based on the personal historical period tags, retrieve the set of positive emotional arousal keywords associated with that period from the emotion association mapping table; 105: Based on the aforementioned set of positive emotion arousal keywords, construct an intervention cue sequence that includes narrative cues, sensory stimulation cues, and behavioral suggestion cues; 106: Sort the intervention cue sequence according to the emotional arousal gradient to generate a nostalgia intervention step arrangement with a time progression relationship; 107: The nostalgia intervention steps are arranged and packaged into a structured intervention plan data package and pushed to the user interaction terminal; 108: In response to the solution execution feedback signal returned by the user interaction terminal, adjust the order of unexecuted clues in the intervention clue sequence or replace the clue content, rearrange the adjusted intervention clue sequence and update the intervention solution data package; The specific steps for matching the corresponding personal historical period tags on the preset life time division axis in step 103 are as follows: 1031: Combine entity object identifiers with scene context identifiers into composite search criteria; 1032: Query the constructed nostalgia mapping index table using composite search criteria. This index table contains an item-scene combination field and the corresponding era field. 1033: When a composite search condition matches the item-scene combination field in the index table, extract the corresponding era stage field value as a candidate period label; 1034: If 1033 does not find any record, then query the entity object identifier and the scene context identifier as single search conditions, and take the intersection of the era stage field values ​​that they hit. If the intersection is not empty, then take the intersection as the candidate period label. 1035: Map the candidate period labels to the life period division axis. The life period division axis is pre-divided into five segments according to age range: childhood, adolescence, youth, middle age and old age. Determine the age segment to which the candidate period label belongs based on the year range indicated by the candidate period label, and output the name of the age segment as the personal historical period label.

2. The method for generating a depression intervention program based on nostalgia guidance according to claim 1, characterized in that, The image files of the digital representation of the nostalgic guide in 101 are obtained through the following methods: Receive photos of physical vintage items taken by a user interaction terminal, extract the shooting timestamp and geolocation tag of the photos, and append the shooting timestamp and geolocation tag to the metadata field of the image file.

3. The method for generating a depression intervention program based on nostalgia guidance according to claim 1, characterized in that, The specific steps for parsing the digital representation of the nostalgic guide and extracting entity object identifiers and scene context identifiers in step 102 are as follows: 1021: If the digital representation of the nostalgic guide is an image file, then the image file is segmented to identify the foreground subject outline and background environment outline in the image; 1022: Compare the outline of the foreground subject with the outline features of a pre-set library of nostalgic items, and extract the names of items with a matching degree exceeding a threshold as entity object identifiers; 1023: Compare the background environment outline with the features of the pre-set scene type library to identify the architectural style, natural landscape or interior layout type in the environmental elements and generate scene context identifiers; 1024: If the nostalgic guide digital representation is an audio file, then the audio file is transcribed into speech, and noun phrases are extracted from the transcribed text as entity object identifiers, and noun phrases representing locations are extracted as scene context identifiers. 1025: If the nostalgic guide's digital representation is a text description, then perform syntactic dependency analysis on the text description, extract the subject headword as the entity object identifier, and extract the place words in the adverbial or complement as the scene context identifier.

4. The method for generating a depression intervention program based on nostalgia guidance according to claim 1, characterized in that, The 104 emotion association mapping tables are constructed and invoked in the following ways: Pre-configure at least three positive emotional evoking keywords for each individual's historical period tag. These keywords are adjective phrases that describe feelings of pleasure, pride, emotion, or tranquility. Based on the personal historical period tags output by 103, perform a hash lookup in the emotion association mapping table to extract all positive emotion arousal keywords corresponding to the tag, and sort them in descending order according to the frequency of user responses triggered by the keywords in historical interaction data. Take the top five keywords as the set of positive emotion arousal keywords.

5. The method for generating a depression intervention program based on nostalgia guidance according to claim 1, characterized in that, The specific steps for constructing the intervention cue sequence in step 105 are as follows: 1051: Select one keyword from the set of positive emotional arousal keywords as the core emotional tone of the current cue; 1052: Based on the core emotional tone, nostalgic narrative short articles that match the personal historical period tags and are emotionally oriented with the keyword are retrieved from the narrative template library to form narrative prompts; 1053: Based on the core emotional tone, retrieve image or audio samples associated with the entity object identifier or scene context identifier from the multimedia material library to form a sensory stimulation cue; 1054: Based on the core emotional tone, extract simple physical activities or creative expression task descriptions that match the personal historical period labels from the behavior suggestion rule base to form behavior suggestion prompts; 1055: Combine narrative cues, sensory stimulus cues, and behavioral suggestion cues in a fixed order to form an intervention cue unit. Repeat steps 1051 to 1054 until all keywords in the positive emotional arousal keyword set have been processed, resulting in multiple intervention cue units. The set forms an intervention cue sequence.

6. The method for generating a depression intervention program based on nostalgia guidance according to claim 5, characterized in that, The specific steps for sorting according to the emotional arousal gradient in the 106 are as follows: Each intervention cue unit is assigned an arousal intensity coefficient, which is determined based on the level of the positive emotional arousal keywords used in that unit in a preset intensity level table. The intervention cue units in the intervention cue sequence are arranged from low to high according to the arousal intensity coefficient, forming a progressive structure in which emotional arousal transitions from gentle to intense. The arranged intervention clue units are assigned sequential numbers, and these sequential numbers serve as the basis for the temporal progression of the nostalgia intervention steps.

7. The method for generating a depression intervention program based on nostalgia guidance according to claim 6, characterized in that, The specific steps for adjusting the intervention cue sequence in step 108 in response to the scheme execution feedback signal returned by the user interaction terminal are as follows: 1081: Receive the solution execution feedback signal returned by the user interaction terminal. The feedback signal includes the user's satisfaction rating and emotion self-rating scale score for the completed intervention cue unit. 1082: Compare the satisfaction score and the score of the emotion self-rating scale with the preset satisfaction threshold and emotion improvement threshold respectively to generate an execution effect level label, wherein the execution effect level label includes high approval, average approval and low approval; 1083: If the effectiveness level label is highly approved, then the original order and content of the unexecuted intervention cue units remain unchanged; If the performance level label is "generally acceptable", then identify the positive emotional arousal keywords used by the intervention cue unit corresponding to the feedback signal, find other units that use the same keywords in the unexecuted intervention cue units, and adjust the sorting position of the other unit one position forward. If the performance level label is low approval, identify the positive emotion arousal keyword used by the intervention cue unit corresponding to the feedback signal, delete the keyword from the set of positive emotion arousal keywords, retrieve the unused positive emotion arousal keywords from the emotion association mapping table according to 104, and regenerate the subsequent intervention cue units according to 105 to 106 and replace the original unexecuted units. 1084: Integrate the adjusted sequence of intervention clue units with the replaced unit content to form an adjusted sequence of intervention clues.

8. The method for generating a depression intervention program based on nostalgia guidance according to claim 7, characterized in that, After rearranging the adjusted intervention cue sequences and updating the intervention protocol data package in step 108, the following steps are also included: Recalculate the mean and standard deviation of the arousal intensity coefficients of all intervention cue units in the updated intervention protocol data package, and write the mean and standard deviation into the metadata field of the intervention protocol data package; Generate a version change record, which includes the timestamp of this adjustment, the number of leads before the adjustment, the number of leads after the adjustment, and the execution effect level label field. Append this version change record to the end of the intervention plan data package. The personal history period tags, positive emotion arousal keyword sets, intervention clue sequences, and intervention plan data packages used in 101 to 108 are all transmitted between the user interaction terminal and the server in encrypted JSON data format. During the transmission process, the entity object identifiers and scene context identifiers are desensitized by replacing the specific names in the identifiers with the generic names of the categories to which those names belong.

9. A depression intervention program generation system based on nostalgia guidance, characterized in that, include: One or more processors; A memory having stored one or more programs that, when executed by one or more processors, cause the one or more processors to implement a method for generating a depression intervention program based on nostalgia guidance, as described in any one of claims 1 to 8.

Images