Personalized tourism consumption scene construction method and system based on behavior map

Abstract

The invention relates to a personalized tourism consumption scene construction method and system based on a behavior map, and relates to the field of tourism consumption recommendation, and the method comprises the steps: collecting user behavior information; constructing a behavior graph based on the user behavior information; generating a fitting scene based on the behavior map; comparing the fitting scene with a scene of a preset scene library to obtain a matching degree; screening out the scene with the matching degree greater than a preset reference matching degree from the scenes in the scene library; and combining the screened scenes based on the behavior map to form a tourism consumption scene. The scene recommendation method and device have the effect of improving the scene recommendation precision.

Description

Technical Field

[0001] The present invention relates to the field of tourism consumption recommendations, and in particular to a method and system for constructing personalized tourism consumption scenarios based on behavior graphs. Background Art

[0002] Behavior mapping is a data visualization tool and modeling method used to understand and analyze individual or group behavior patterns. It graphically displays the relationships, sequences, and influencing factors between behaviors. Behavior mapping can be applied to scenario building.

[0003] With the development of the internet and big data technologies, travel recommendation systems are gradually evolving from traditional static recommendations (such as those based on popular attractions or fixed routes) to personalized recommendations. However, existing travel recommendation systems mostly rely on basic user information (such as age, gender, and past bookings) or simple behaviors (such as clicks and favorites). They lack in-depth analysis of user behavior patterns and fail to effectively integrate multi-source data, resulting in inaccurate recommendations. Summary of the Invention

[0004] In order to improve the accuracy of recommended scenarios, the present invention provides a method and system for constructing personalized tourism consumption scenarios based on behavior graphs.

[0005] In a first aspect, the present invention provides a method for constructing personalized tourism consumption scenarios based on behavior graphs, which adopts the following technical solutions: A method for constructing personalized tourism consumption scenarios based on behavioral graphs, including: S1: Collect user behavior information; S2: Constructing a behavior graph based on the user behavior information; S3: generating a fitting scenario based on the behavior map; S4: Compare the fitted scene with scenes in a preset scene library to obtain a matching degree; S5: Filtering out scenes whose matching degree is greater than a preset benchmark matching degree from the scenes in the scene library; S6: Combining the screened scenarios based on the behavior graph to form a tourism consumption scenario.

[0006] By adopting the above technical solution, the system collects users' deep behavioral pattern information and establishes a behavioral map. It matches the behavioral map with scenarios that are closer to the user's preferred behaviors, and combines multiple scenarios according to the behavioral map to form a unique personalized travel consumption scenario that suits the user. Compared with the existing static recommendation method, the recommendation structure is more accurate and can meet user needs.

[0007] Optional methods for collecting user behavior information include: S10: Searching the network based on the preset user information to determine whether the user exists on a public social platform; S11: If the user exists on the public social network platform, execute S12; otherwise, execute S14; S12: Collecting key user information from the user's online public social platform; S13: Match key question and answer questions based on the user key information; S14: combining the key questions and answers and the preset template questions and answers to form a questionnaire; S15: Obtaining user behavior information using the questionnaire and a preset behavior identification method.

[0008] Optional methods for collecting key user information include: S1000: Identifying historical travel locations from the user's public social networking platform; S1001: Identify the historical travel locations to determine whether there is a user evaluation record; S1002: Based on the existence of the evaluation record, extract good and bad keywords from the evaluation record; S1003: Classifying the historical travel locations based on the good and bad keywords to obtain travel location types, where the travel location types include grass planting locations and lightning protection locations; S1004: Identify all the grass planting locations to determine common grass planting points, and determine a grass planting location range based on the common grass planting points; S1005: Identify all the lightning protection locations to determine common lightning protection points, and determine a range of the lightning protection locations based on the common lightning protection points; S1006: Generate the user key information based on the grass planting location range and the lightning protection location range.

[0009] Optionally, the method for collecting key user information also includes: S1010: Identifying elements of people of the opposite sex and the frequency of occurrence of the same element of people of the opposite sex from the public social networking platform of the user; S1011: Extracting emotional keywords from the user's public social platform; S1012: Determine an emotional orientation based on the emotional keywords; S1013: Determine the emotional state of the person based on the frequency of occurrence of the same opposite-sex person element and the emotional orientation, where the emotional state includes a lovelorn state and a love state; S1014: Analyze the frequency of occurrence of the same opposite-sex person element in the user's online public social platform to determine the time node of the emotion change; S1015: Based on the lovelorn state, determining the location of the emotional change according to the emotional change time node; S10151: Analyze the location of the emotional change to determine the type element of the location of the broken heart, and generate the user key information based on the type element of the location of the broken heart.

[0010] Optionally, also include: S1016: Based on the relationship status, identifying the relationship location after the emotional change time point from the user's online public social platform; S10161: Analyze all the love locations to determine love location elements and love process elements; S10162: Generate the user key information based on the love location element and the love progress element.

[0011] Optionally, also include: S1017: Identify the emotion change time nodes from the user's online public social platform and determine the number of time nodes; S10171: When the number of the time nodes is not less than 2, determining the opposite-sex person element who has been in both the lovelorn state and the lovelorn state with the user based on the person's emotional state; S10172: extracting the corresponding love location element and the love process element based on the opposite-sex person element; S10173: Correct the love location element and the love process element to the love-breakup location type element.

[0012] Optional behavior identification methods include: S150: When the user fills in the questionnaire, collecting the user's facial micro-expression image and the questionnaire feedback information entered by the user in the questionnaire; S151: When the questionnaire feedback information corresponding to the key question is blank, outputting the preset estimated answer standard as the user behavior information; S152: When the questionnaire feedback information corresponding to the key question is not blank, continue to determine whether the questionnaire feedback information is consistent with the estimated answer standard; S153: When the questionnaire feedback information is inconsistent with the estimated answer standard, determining whether the facial micro-expression image is consistent with a preset unexpected micro-expression; S154: When the facial micro-expression image is consistent with the unexpected micro-expression, the user's answer is defined as abnormal, and the estimated answer standard is output as the user behavior information.

[0013] Optionally, the questionnaire includes a paper questionnaire and an electronic questionnaire; and the behavior identification method further includes: S1501: Based on the paper questionnaire, collect the questionnaire image information; S1502: Identify a preset writing area from the paper image information to determine the writing content; S1503: Determine whether the written content contains a preset erasure feature; S1504: When the written content contains the erasure feature, the written range is identified from the paper image information to determine the paper question and answer line where the erasure feature is located, and whether the paper question and answer line is related to the emotional direction; S1505: When the question and answer items on the paper are about sentiment, identifying the altered features from the paper image information to determine the estimated original content; S1506: Combining the estimated original content and the estimated answer standard to obtain user behavior information.

[0014] Optionally, the method of combining the screened scenarios to form a tourism consumption scenario includes: S60: Analyze the behavior graph to obtain the user's travel preferences; S61: Determine a travel route based on the user's travel preference plan; S62: Matching the user's travel preferences with the screened scenarios; S63: Arranging the filtered scenes in sequence based on the travel path to obtain a travel scene arrangement; S64: Form a tourism consumption scenario based on the itinerary scenario arrangement.

[0015] Secondly, this application provides a personalized tourism consumption scenario construction system based on behavioral graphs, which adopts the following technical solutions: A personalized tourism consumption scenario construction system based on behavioral graphs, including: Acquisition module, used to collect user behavior information; A memory for storing a program for a method for constructing personalized tourism consumption scenarios based on a behavior graph; The program in the processor and the memory can be loaded and executed by the processor to realize a method for constructing personalized tourism consumption scenarios based on behavior graphs.

[0016] In summary, this application includes at least one of the following beneficial technical effects: The system collects in-depth information about users' behavioral patterns and builds a behavioral graph. It then uses the behavioral graph to match scenarios that are more closely aligned with the user's preferences and behaviors. It then combines multiple scenarios according to the behavioral graph to create a unique, personalized travel consumption scenario that suits the user. Compared to existing static recommendation methods, the recommendation structure is more precise and can meet user needs. The system identifies users' deep personal behavioral information from their public social media platforms, including their evaluation of historical travel records and the travel locations involved in their romantic lives. This information serves as an important element in constructing personalized travel consumption scenarios, making recommendations more accurate. By obtaining the user's questionnaire completion status and micro-expressions when the user fills out the questionnaire, it is determined whether the user has filled out the questionnaire abnormally due to privacy issues, and thus whether the user's personal deep-level behavior information identified from the user's online public social platform is accurate, which facilitates the subsequent personalized tourism consumption scenario recommendations based on the obtained personal deep-level behavior information. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a method flow chart of a method for constructing personalized tourism consumption scenarios based on behavior graphs according to an embodiment of the present invention; Figure 2 This is the method flow of the method for collecting user behavior information in an embodiment of the present invention Figure 1 ; Figure 3 This is the method flow of the method for collecting user behavior information in an embodiment of the present invention Figure 2 ; Figure 4 This is the method flow of the method for collecting user behavior information in an embodiment of the present invention Figure 3 ; Figure 5 This is the method flow of the method for collecting user behavior information in an embodiment of the present invention Figure 4 ; Figure 6 This is the method flow of the method for collecting user behavior information in an embodiment of the present invention Figure 5 . DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0019] The embodiments of the present application disclose a method for constructing personalized tourism consumption scenarios based on behavior graphs.

[0020] Reference Figure 1 ,The method for constructing personalized tourism consumption scenarios based on ,behavior graph includes the following steps: Step S1: Collect user behavior information.

[0021] User behavior information refers to the user's deep-level personal preferences and personal status, which is different from the user's basic information (such as age, gender, historical orders) or simple behaviors (such as clicks, favorites), etc.

[0022] In this embodiment, user behavior information is analyzed and obtained from the user's personal network public social platform. The specific collection method is not described here in detail and will be described in subsequent embodiments.

[0023] Step S2: Constructing a behavior graph based on the user behavior information.

[0024] The behavior map has been described in the previous article. In this embodiment, the behavior map brings together various deep-level personal preference behaviors and personal status of users. The behavior map is formed by combining user behavior information. Therefore, after collecting user behavior information, all user behavior information is merged to obtain a behavior map.

[0025] Step S3: generating a fitting scenario based on the behavior map.

[0026] Fitting scenarios are virtual travel scenarios that reflect a user's deep-seated preferences and personal circumstances. Fitting scenarios are not physical landscapes or locations. Fitting scenarios are generated by matching the user's personalized behavior characteristics within the behavior graph. For example, if a user enjoys reading and drinking coffee, the fitting scenario would include locations such as bookstores and coffee shops.

[0027] Step S4: Compare the fitted scene with scenes in a preset scene library to obtain a matching degree.

[0028] The scene library is a database designed in advance by technical personnel that contains scenes of tourism consumption locations in various regions. It was collected in the early stage and will not be described in detail here.

[0029] The matching degree refers to the similarity between the virtual fitting scene and the actual location scene.

[0030] By comparing the fitting scene with the scenes in the scene library one by one, the matching degree of each scene in the scene library can be obtained. The more elements in the scene library that meet the fitting scene, the higher the matching degree.

[0031] Step S5: Filter out scenes whose matching degree is greater than a preset reference matching degree from the scenes in the scene library.

[0032] The benchmark matching degree is the matching degree standard set by the technical staff for the scenes in the scene library to meet the user's personalized requirements. When the matching degree between the scene in the scene library and the fitting scene is greater than the benchmark matching degree, it means that the scene can meet the user's personalized travel consumption requirements. No further details will be given here.

[0033] After determining the matching degree of all the scenes in the scene library, scenes that meet the benchmark matching degree requirements are screened out. The screened out scenes are scenes that meet the user's personalized travel consumption requirements.

[0034] Step S6: Combining the screened scenarios to form a tourism consumption scenario based on the behavior graph.

[0035] After obtaining the scenarios that meet the requirements, it is necessary to combine and arrange multiple scenarios so that the arranged scenario combination meets the characteristics of the user's behavior map. The final completed scenario combination is the required tourism consumption scenario. The specific combination method will not be described in detail here, but will be described in detail in the subsequent embodiments.

[0036] Reference Figure 2 ,The method for collecting user behavior information includes the following steps: Step S10: Search the network based on the preset user information to determine whether the user has a public social platform on the network.

[0037] User information refers to the user's basic personal information. When building a personalized travel consumption scenario for the user, technical personnel need to know the individual's basic information in advance, which will not be elaborated here.

[0038] An online public social platform refers to a social platform where users record and publish their daily personal updates on the Internet. The content is directly accessible to the public and is not a social platform for personal privacy.

[0039] Public online social platforms are usually registered using basic personal information, so whether a user exists on a public online social platform can be determined through the user's basic personal information.

[0040] Step S11: If the user exists on the public social network platform, execute S12; otherwise, execute S14.

[0041] There are two possibilities for whether the user has a public online social platform. The system performs different steps for different possibilities. If the user has a public online social platform, the system performs step S12. The method of step S12 is not described here and is described in detail in subsequent embodiments. If the user does not have a public online social platform, the system performs step S14. The method of step S14 is not described here and is described in detail in subsequent embodiments.

[0042] Step S12: collecting key user information from the user's public social network platform.

[0043] User key information refers to information that can reflect the user's deep personal preferences, behaviors and personal status. User key information includes the user's preferences for travel locations and travel location conditions related to the user's emotions.

[0044] Key user information can be obtained by analyzing public social platforms on the Internet. The specific analysis and collection methods are not described here in detail and will be introduced in subsequent embodiments.

[0045] Step S13: Match key question and answer questions based on the user key information.

[0046] In this embodiment, a questionnaire survey is conducted on users. Key questions and answers refer to questions in the questionnaire that contain key user information. By comparing the user's key information with a preset question and answer solution table, key questions and answers that match the user's key information are obtained from the question and answer solution table. The question and answer solution table is a table of question and answer solutions pre-summarized by technical personnel and contains various types of key user information. A detailed description of the table is omitted here.

[0047] Step S14: merging the key questions and answers and the preset template questions and answers to form a questionnaire.

[0048] Template question and answer questions refer to question and answer questions generated only based on personal basic information and general questionnaire content.

[0049] The questionnaire is used to investigate the user's personal situation and further clarify the user's personalized situation. The questionnaire can be answered in person through a paper questionnaire or online through an electronic questionnaire. In this embodiment, the questionnaire consists of key questions and template questions.

[0050] When the user has a public social platform on the Internet, key questions and answers can be formed based on the personal situation identified and analyzed in the user's public social platform on the Internet. At this time, the questionnaire includes key questions and answers and template questions and answers.

[0051] When the user does not have a public social platform on the Internet, there are no key questions and answers. In this case, the content of the questionnaire only includes template questions and answers.

[0052] Step S15: Obtaining user behavior information using the questionnaire and a preset behavior identification method.

[0053] The user behavior information has been explained in step S1 and will not be repeated here.

[0054] Behavioral identification methods are used to identify user behavior while completing a questionnaire. Depending on the questions in the questionnaire, users may exhibit different responses. These responses can reflect their actual responses. Users may complete the questionnaire normally or abnormally. By targeting different situations, we can ultimately obtain accurate user behavior information that reflects the user's true individuality.

[0055] Reference Figure 3 In this embodiment, the user's key information about preferences and dislikes of travel locations is collected. The method for collecting the user's key information includes the following steps: Step S1000: Identify historical travel locations from the user's public social networking platform.

[0056] Historical travel locations refer to places a user has publicly visited on social media platforms. Since historical travel locations are publicly available on social media platforms, they can be identified from the user's publicly available social media platforms. This identification can be done through image comparison or text extraction. For places a user hasn't publicly visited on social media platforms, they can be obtained through questionnaires.

[0057] Step S1001: Identify the historical travel locations to determine whether the user has any evaluation records.

[0058] Review records refer to information about users' reviews of historical travel locations. Most social media platforms allow users to write text. After a trip, users may write about their trip, which may include reviews of the location.

[0059] By recognizing the text written by the user, it is possible to identify whether there is an evaluation record of the historical travel location.

[0060] Step S1002: Based on the existence of the evaluation record, extract good and bad keywords from the evaluation record.

[0061] Good and bad keywords refer to the keywords used in users' positive and negative reviews of historical travel destinations. Good and bad keywords can be obtained by text recognition from review records.

[0062] Step S1003: Classify the historical travel locations based on the good and bad keywords to obtain travel location types, where the travel location types include grass planting locations and lightning protection locations.

[0063] A "planting location" is a location that users like and are happy to visit. A "shield location" is a location that users dislike and will not visit again.

[0064] The system can categorize the user's historical travel locations by identifying good and bad keywords, and determine the recommended places and safe places among the historical travel locations.

[0065] Step S1004: Identify all the grass planting sites to determine the same grass planting points, and determine the grass planting site range based on the same grass planting points.

[0066] The similarities in the recommended spots refer to the similarities in all the recommended spots that users like to visit. Users like the historical travel spots because of these similarities.

[0067] The scope of grass planting locations refers to the scope of locations with the same grass planting points. All tourist consumption scene locations with the same grass planting points are included in the scope of grass planting locations.

[0068] Step S1005: Identify all the lightning protection locations to determine the common lightning protection points, and determine the range of the lightning protection locations based on the common lightning protection points.

[0069] Similar to step S1004, the lightning avoidance similarity refers to the similarity of all lightning avoidance locations disliked by the user, and the user develops a dislike for the historical travel location due to the similarity.

[0070] The lightning protection location range refers to the location range with the same lightning protection point. All tourism consumption scenarios with the same lightning protection point are included in the lightning protection location range.

[0071] Step S1006: Generate the user key information based on the grass planting location range and the lightning protection location range.

[0072] After determining the range of the planting location and the range of the lightning protection location, the planting location range and the lightning protection location range are merged to obtain the user's key information about the user's likes and dislikes of travel locations.

[0073] Reference Figure 4 This embodiment aims to collect user key information about locations related to the user's personal emotional state. The method for collecting user key information includes the following steps: Step S1010: identifying opposite-sex person elements and the frequency of occurrence of the same opposite-sex person element from the user's public online social platform.

[0074] The opposite-sex person element refers to the opposite-sex person appearing on a user's social platform. The frequency of appearance refers to the number of times the same opposite-sex person appears in content on a user's social platform.

[0075] By collecting and identifying all people who have appeared on the user's online public social platform, the opposite-sex person element is determined. The frequency is then obtained by counting the number of times each opposite-sex person element appears.

[0076] Step S1011: extracting emotional keywords from the user's public social platform.

[0077] Emotional keywords refer to words that express a user's emotions in text posted on public social platforms. Emotional keywords can be directly extracted from the text posted by users on public social platforms.

[0078] Step S1012: Determine the emotional orientation based on the emotional keywords.

[0079] Emotional orientation refers to the trend of user emotions changing over a certain period of time, including gradually becoming happy or gradually becoming depressed.

[0080] The sentiment orientation is determined by analyzing the emotional keywords that appear over a period of time. If a user's posts over a period of time contain a large number of words related to a happy mood, the user's sentiment orientation is considered happy. If a user's posts over a period of time contain a large number of words related to a bad mood, the user's sentiment orientation is considered depressed.

[0081] Step S1013: Determine the emotional state of the person based on the frequency of occurrence of the same opposite-sex person element and the emotional orientation, where the emotional state includes a heartbroken state and a love state.

[0082] The emotional orientation of a heartbroken state is usually a depressed mood, while the emotional orientation of a love state is usually a happy mood.

[0083] By combining and analyzing the user's emotional orientation over a period of time with the frequency of occurrence of the same opposite-sex person element during that period of time, the user's personal emotional state can be determined.

[0084] If the frequency of occurrence of the same opposite-sex person element increases over a period of time, and the user's emotional orientation during this period is a happy mood, it means that the user's emotional state is a romantic state.

[0085] If the frequency of the same opposite-sex person element decreases over a period of time, and the user's emotional orientation during this period is depressed, it means that the user's emotional state is a state of broken love.

[0086] Step S1014: analyzing the frequency of occurrence of the same opposite-sex person element in the user's public online social platform to determine the time node of the emotion change.

[0087] The time node of emotional change refers to the time node when the emotions expressed by users on online social platforms change, that is, the time node when they enter a romantic state or a heartbroken state.

[0088] By analyzing the frequency of occurrence of the same opposite-sex person element, the moment when the frequency of occurrence of the same opposite-sex person element gradually decreases to 0 or the moment when the frequency of occurrence increases to the maximum is defined as the time node of emotional change.

[0089] Step S1015: Based on the lovelorn state, determine the location of the emotion change according to the emotion change time node.

[0090] The location of emotional change refers to the location where the user travels when he or she breaks up.

[0091] After determining the time point of the emotional change, we analyze the text posted by the user when they were heartbroken and identify the location of their travel at that time from the text. The user may have experienced an emotional change at the location of the emotional change due to something, such as a heartbreak.

[0092] Step S10151: Analyze the location of the emotional change to determine the type element of the location of the broken heart, and generate the user key information based on the type element of the location of the broken heart.

[0093] In this embodiment, since the user experiences a heartbreak at an emotionally changing location, it is defined that the emotionally changing location will stimulate the user, and scenarios involving emotionally changing locations are avoided when recommending travel consumption scenarios.

[0094] The "breakup location" type refers to the various visible elements of a location associated with a change in relationship. After identifying a location associated with a change in relationship, analysis is performed to identify the breakup location type elements present in that location. When recommending travel destinations, it's important to avoid including these elements in the scene. Therefore, these elements are key user information that needs to be collected.

[0095] Reference Figure 5 , the method for collecting key user information also includes the following steps: Step S1016: Based on the relationship status, identify the relationship location after the emotion change time point from the user's online public social platform.

[0096] The dating location refers to the travel locations displayed by users on online social platforms after they fall in love.

[0097] After determining the time point at which the user's emotions change, the text following the text published by the user at the time point of the emotions change is analyzed to extract the love locations where the user traveled during the love process.

[0098] Step S10161: Analyze all the love locations to determine love location elements and love process elements.

[0099] The dating location element refers to the scene elements contained in the places where the user travels during the dating period. The dating process element refers to the changes in the travel locations as the user's feelings gradually sublimate during the dating period.

[0100] Both the love location factor and the love process factor can be obtained by analyzing the love progress location.

[0101] Step S10162: Generate the user key information based on the love location element and the love progress element.

[0102] The places and scenes visited during the love process will be remembered by the user, so the love location elements and the love process elements can be used as the basis for recommending travel consumption scenes to the user, and user key information can be generated based on the love location elements and the love process elements.

[0103] Reference Figure 6 When a user is in a relationship or a breakup with the same person of the opposite sex, the method for handling the situation includes the following steps: Step S1017: Identify the emotion change time nodes from the user's public social platform and determine the number of time nodes.

[0104] Users may display multiple emotional states on public social platforms, experiencing both heartbreak and falling in love. When multiple emotional states are present, multiple time points will appear.

[0105] According to the method of step S1014, a plurality of emotion change time nodes displayed by the user on the public network social platform can be determined, and the number of the emotion change time nodes can be obtained by counting the emotion change time nodes.

[0106] Step S10171: When the number of the time nodes is not less than 2, the elements of the opposite-sex persons who have been in the love-break state and the love state at the same time as the user are determined based on the emotional state of the persons.

[0107] If the number of time nodes is less than 2, it means that the user has only experienced one love state or one heartbreak state.

[0108] If the number of time nodes is at least 2, it indicates that the user may have experienced both a romantic relationship and a breakup. Analyzing the emotional state of people on the user's public online social platforms can identify the same person of the opposite sex who experienced both a romantic relationship and a breakup with the user. The frequency of this person of the opposite sex first increases, then decreases.

[0109] Step S10172: extracting the corresponding love location element and the love process element based on the opposite-sex person element.

[0110] The same opposite-sex person element is identified from the user's online public social platform, the love location during the love period when the opposite-sex person element exists is determined, and the love location element and the love process element are retrieved from the love location.

[0111] Step S10173: Correct the love location element and the love process element to the love breakup location type element.

[0112] When a user falls in love and then breaks up with the same person of the opposite sex, the places they visited during the relationship will become a trigger for depression. Therefore, the love location elements and love process elements of the places they visited during the relationship will be changed to the breakup location type requirements.

[0113] The behavior identification method includes the following steps: Step S150: When the user fills in the questionnaire, the user's facial micro-expression image and the questionnaire feedback information input by the user in the questionnaire are collected.

[0114] The user's facial micro-expression image refers to the user's facial expressions while filling out the questionnaire. This facial micro-expression image is captured via a camera. When a user completes a paper questionnaire, the space where they fill in the questionnaire has a camera that can capture the user's facial micro-expressions in real time. When a user completes an electronic questionnaire, the system can, with the user's consent, use the system camera to capture the user's facial micro-expression image.

[0115] Questionnaire feedback information refers to the information filled out by the user in the questionnaire. The questionnaire feedback information is obtained by identifying and analyzing the questionnaire content through a camera.

[0116] Step S151: When the questionnaire feedback information corresponding to the key question and answer question is blank, the preset estimated answer standard is output as the user behavior information.

[0117] The estimated response standard refers to the user's estimated response situation based on the user's personal situation when the technical staff designs the questionnaire questions. The estimated response standard serves as a reference standard for the questionnaire questions and will not be elaborated here.

[0118] When users fill out a questionnaire, there may be three situations in the questionnaire feedback information they enter. The first is that it is blank; the second is that it is filled out normally; and the third is that it is filled out randomly and is contrary to the facts.

[0119] If the questionnaire feedback information corresponding to the key question is blank, it is determined that the user does not want to answer the key question. At this time, the estimated answer standard is used as the answer to the key question, that is, the estimated answer standard is determined to be user behavior information.

[0120] Step S152: When the questionnaire feedback information corresponding to the key question and answer question is not blank, continue to determine whether the questionnaire feedback information is consistent with the estimated answer standard.

[0121] If the questionnaire feedback information corresponding to the key question is not blank, it means that the user has answered the key question. At this time, the user's answer, that is, the questionnaire feedback information, is compared with the estimated answer standard to determine whether the user answered randomly.

[0122] Step S153: When the questionnaire feedback information is inconsistent with the estimated answer standard, it is determined whether the facial micro-expression image is consistent with a preset unexpected micro-expression.

[0123] Unexpected micro-expressions refer to the unexpected micro-expression characteristics that are pre-entered into the facial recognition system by technicians and will not be elaborated here.

[0124] If the user's responses are inconsistent with the estimated response standards, the system may have entered the responses carelessly or unintentionally. This is determined by identifying the user's facial expressions while entering the responses. Users may display unexpected facial expressions when presented with key questions, suggesting they may have filled out the responses carelessly.

[0125] Step S154: When the facial micro-expression image is consistent with the unexpected micro-expression, the user's answer is defined as abnormal, and the estimated answer standard is output as the user behavior information.

[0126] If the user's facial micro-expressions when filling out the questionnaire are consistent with unexpected micro-expressions, it is concluded that the user has made abnormal input when answering key questions. At this time, the user's questionnaire feedback information is not used as the final result, but the estimated answer standard is output as the final result, that is, the user behavior information.

[0127] If the user's facial micro-expression is inconsistent with the unexpected micro-expression when filling out the questionnaire, it is determined that the user's input is normal when answering the key questions. At this time, the user's input questionnaire feedback information is output as user behavior information.

[0128] The questionnaire includes a paper questionnaire and an electronic questionnaire; the behavior identification method also includes the following steps: Step S1501: Based on the paper questionnaire, collect paper questionnaire image information.

[0129] The questionnaire image information refers to the questionnaire image captured by a camera installed in the room.

[0130] Step S1502: Identify the preset writing range from the paper image information to determine the writing content.

[0131] The writing area refers to the area in the questionnaire where users need to write. The writing content refers to the content written by users in the questionnaire.

[0132] By recognizing the text within the writing range in the paper image information, the written content of the user can be identified.

[0133] Step S1503: Determine whether the written content contains a preset erasure feature.

[0134] The erasure feature refers to the situation where the user alters the text while writing it. In the image, the erasure feature is usually manifested as crossing out with a line segment or directly blackening it out.

[0135] By identifying the erasure features from the paper image information, it is possible to determine whether the written content contains erasure features.

[0136] Step S1504: When the written content contains the erasure feature, the writing range is identified from the paper image information to determine the paper question and answer line where the erasure feature is located, and it is determined whether the paper question and answer line is about the emotional direction.

[0137] If there are no alteration features, no behavior identification is performed.

[0138] If the written content contains erasure features, it means that the user may have made normal input errors or hesitated when answering. At this time, the location of the erasure features is identified to determine the possible situations when the user answered.

[0139] The question and answer row on the questionnaire refers to the location of the altered feature in the questionnaire. By identifying the writing range and the altered feature from the questionnaire image information, the question and answer row on the questionnaire where the altered feature is located can be determined.

[0140] After determining the question and answer line, it is determined whether the key question and answer question in the question and answer line is about the emotional direction. When answering questions about the emotional direction, users may hesitate to enter.

[0141] Step S1505: When the question and answer items on the paper are about the emotional direction, the altered features are identified from the paper image information to determine the estimated original content.

[0142] The estimated original content refers to the content of the altered feature before it is altered.

[0143] When the key question on the test paper is identified as sentiment-related, if an altered feature appears, it indicates that the user hesitated and entered the answer carelessly. The altered feature is then identified and the original estimated content is restored. This estimated original content can be used as a reference for the key question.

[0144] The method for restoring altered features is as follows: if the altered feature is directly blacked out, it cannot be used to estimate the original content and is therefore considered an invalid feature. However, if the altered feature is only crossed out by a line segment, the text and line segments are separated, the separated lines are removed, and the remaining text is recognized to determine the estimated original content.

[0145] Step S1506: combining the estimated original content and the estimated answer standard to obtain user behavior information.

[0146] Both the estimated original content and the estimated answer standards can be used as answers to key questions and answers. In this case, the estimated original content and the estimated answer standards are merged, and the merged result is used as user behavior information.

[0147] The method of combining the screened scenarios to form a tourism consumption scenario includes the following steps: Step S60: Analyze the behavior graph to obtain the user's travel preferences.

[0148] User itinerary preferences refer to the types of travel destinations that users prefer. Behavior graphs capture individual user behavior traits, and by analyzing these graphs, we can derive user itinerary preferences. For example, if a user enjoys coffee, coffee shops would be considered a user's itinerary preference.

[0149] Step S61: Determine a travel route based on the user's travel preference planning.

[0150] A travel route refers to the user's preferred travel route. For example, if a user likes to have a cup of coffee before going shopping, the travel route might include a coffee shop first and then a walking street.

[0151] Step S62: Match the user's travel preferences with the screened scenarios.

[0152] By matching the user's travel preferences with the filtered scenarios, it is possible to determine which scenarios match the user's travel preferences, thereby filtering out these scenarios.

[0153] Step S63: Arranging the filtered scenes in sequence based on the travel path to obtain a travel scene arrangement.

[0154] After obtaining all existing travel consumption scenarios that match the user's itinerary preferences, these travel consumption scenarios are arranged according to the user's itinerary path, thus obtaining an itinerary scenario arrangement. The itinerary scenario arrangement is the order of travel to the actual planned travel destinations.

[0155] Step S64: Arrange the itinerary scenarios to form a tourism consumption scenario.

[0156] After all the scenes are arranged, the overall tourism planning route formed after the arrangement is the tourism consumption scene.

[0157] Once the tourism consumption scenario is determined, the following steps are needed to correct it due to weather and environmental factors: Step S7: Based on the tourism consumption scenario, query and obtain the destination weather information from a preset national weather query platform.

[0158] The National Meteorological Query Platform is used to query local meteorological conditions. It is an online platform tool that can be directly used by the public and will not be described in detail here.

[0159] Destination weather information refers to the weather conditions at a specific location within a travel consumption scenario. By entering the destination location into the National Weather Query Platform, users can query and obtain this information. Destination weather information includes both rainy and sunny weather scenarios, and different weather conditions can affect users' travel plans.

[0160] Step S70: When the weather information of the destination indicates rainy days, the user's likes and dislikes for rainy days are determined based on the user's behavior graph.

[0161] If the weather information at the destination shows sunny, it will not affect the user's travel.

[0162] If the weather information at the destination shows rainy days, the user may not like rainy weather when traveling, which does not meet the user's personalized preferences. In this case, the travel consumption scenario needs to be corrected. The user's likes and dislikes for rainy days can be obtained by analyzing the user's behavior map.

[0163] Step S71: When the user's like or dislike type for rainy days is dislike, secondary favorite scenes are screened out from the scenes in the scene library, and secondary favorite scenes with a shorter distance are screened out based on the secondary favorite scenes and the destination tourism consumption scenes.

[0164] If the user's preference for rainy days is "like", then rainy days will not affect the user's travel experience.

[0165] If a user's preference for rainy days is "dislike," the user cannot proceed to the destination travel consumption scenario. In this case, a secondary preferred scenario is selected to replace the destination travel consumption scenario. A secondary preferred scenario is one that has a slightly lower user-specific match than the baseline.

[0166] Secondary favorite scenes can be obtained from the scenes in the scene library based on the matching degree. After a batch of secondary favorite scenes that meet the requirements are screened out, scenes that are closer to the destination tourism consumption scenes are screened out from the screened secondary favorite scenes to facilitate user travel.

[0167] Step S72: Replace the destination tourism consumption scenario with the secondary favorite scenario, and form a new tourism consumption scenario based on the secondary favorite scenario.

[0168] When the destination tourism consumption scene cannot be visited due to weather reasons, in order to meet the personalized needs of users, secondary favorite scenes are re-screened to replace the destination tourism consumption scene, thereby forming a new tourism consumption scene.

[0169] The method of correcting the tourism consumption scene due to weather and environmental reasons also includes the following steps: In this embodiment, if the user prefers a destination tourism consumption scenario, the destination will not be abandoned in the tourism consumption scenario.

[0170] Step S73: Determine whether the type of the secondary favorite scene is consistent with the type of the destination tourism consumption scene.

[0171] By judging whether the type of the secondary favorite scene is consistent with the type of the destination tourism consumption scene, it is determined whether the user's preference for the secondary favorite scene can replace the destination tourism consumption scene.

[0172] Step S74: When the two are inconsistent, query the raining duration of the destination tourism consumption scenario from the national meteorological query platform.

[0173] If the two are consistent, the user can go to the secondary favorite scene to meet personalized needs.

[0174] If the two are inconsistent, reasonable arrangements need to be made for the destination tourism consumption scenario. Rain duration refers to the duration of rainy weather in the destination tourism consumption scenario. The rain duration of the destination tourism consumption scenario can be obtained from the National Meteorological Query Platform.

[0175] Step S75: Determine the travel duration of the secondary favorite scene based on the rain duration.

[0176] In order to enable users to travel to the destination tourism consumption scene, the user's travel time in the secondary favorite scene is increased. When the user's travel time in the secondary favorite scene is consistent with the rain duration of the destination tourism consumption scene, when the user travels from the secondary favorite scene to the destination tourism consumption scene, the rain in the destination tourism consumption scene has stopped, thus meeting the user's personalized needs.

[0177] Step S76: Arrange the secondary favorite scenes according to the travel duration, and combine the secondary favorite scenes with the destination travel consumption scene to form a new travel consumption scene.

[0178] By rationally arranging the travel duration of the secondary favorite scenes, when users complete the travel of the secondary favorite scenes and go to the destination tourism consumption scene, the destination tourism consumption scene can be in sunny weather, meeting the user's personalized needs.

[0179] Based on the same inventive concept, an embodiment of the present invention provides a personalized tourism consumption scenario construction system based on behavior graphs, including: The acquisition module is used to collect user behavior information, user key information, user facial micro-expression images, questionnaire feedback information and questionnaire image information; A memory for storing a program for a method for constructing personalized tourism consumption scenarios based on a behavior graph; The program in the processor and the memory can be loaded and executed by the processor to realize a method for constructing personalized tourism consumption scenarios based on behavior graphs.

[0180] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for constructing personalized tourism consumption scenarios based on behavioral graphs, characterized in that: include: S1: Collect user behavior information; S2: Constructing a behavior graph based on the user behavior information; S3: generating a fitting scenario based on the behavior map; S4: Compare the fitted scene with scenes in a preset scene library to obtain a matching degree; S5: Filtering out scenes whose matching degree is greater than a preset benchmark matching degree from the scenes in the scene library; S6: Combining the screened scenarios based on the behavior graph to form a tourism consumption scenario.

2. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 1 is characterized in that: Methods for collecting user behavior information include: S10: Searching the network based on the preset user information to determine whether the user exists on a public social platform; S11: If the user exists on the public social network platform, execute S12; otherwise, execute S14; S12: Collecting key user information from the user's online public social platform; S13: Match key question and answer questions based on the user key information; S14: combining the key questions and answers and the preset template questions and answers to form a questionnaire; S15: Obtaining user behavior information using the questionnaire and a preset behavior identification method.

3. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 2 is characterized in that: Methods for collecting key user information include: S1000: Identifying historical travel locations from the user's public social networking platform; S1001: Identify the historical travel locations to determine whether there is a user evaluation record; S1002: Based on the existence of the evaluation record, extract good and bad keywords from the evaluation record; S1003: Classifying the historical travel locations based on the good and bad keywords to obtain travel location types, where the travel location types include grass planting locations and lightning protection locations; S1004: Identify all the grass planting locations to determine common grass planting points, and determine a grass planting location range based on the common grass planting points; S1005: Identify all the lightning protection locations to determine common lightning protection points, and determine a range of the lightning protection locations based on the common lightning protection points; S1006: Generate the user key information based on the grass planting location range and the lightning protection location range.

4. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 2 is characterized in that: Methods for collecting key user information also include: S1010: Identifying elements of people of the opposite sex and the frequency of occurrence of the same element of people of the opposite sex from the public social networking platform of the user; S1011: Extracting emotional keywords from the user's public social platform; S1012: Determine an emotional orientation based on the emotional keywords; S1013: Determine the emotional state of the person based on the frequency of occurrence of the same opposite-sex person element and the emotional orientation, where the emotional state includes a lovelorn state and a love state; S1014: Analyze the frequency of occurrence of the same opposite-sex person element in the user's online public social platform to determine the time node of the emotion change; S1015: Based on the lovelorn state, determining the location of the emotional change according to the emotional change time node; S10151: Analyze the location of the emotional change to determine the type element of the location of the broken heart, and generate the user key information based on the type element of the location of the broken heart.

5. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 4 is characterized in that: Also includes: S1016: Based on the relationship status, identifying the relationship location after the emotional change time point from the user's online public social platform; S10161: Analyze all the love locations to determine love location elements and love process elements; S10162: Generate the user key information based on the love location element and the love progress element.

6. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 5 is characterized in that: Also includes: S1017: Identify the emotion change time nodes from the user's online public social platform and determine the number of time nodes; S10171: When the number of the time nodes is not less than 2, determining the opposite-sex person element who has been in both the lovelorn state and the lovelorn state with the user based on the person's emotional state; S10172: extracting the corresponding love location element and the love process element based on the opposite-sex person element; S10173: Correct the love location element and the love process element to the love-breakup location type element.

7. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 2 is characterized in that: Behavioral identification methods include: S150: When the user fills in the questionnaire, collecting the user's facial micro-expression image and the questionnaire feedback information entered by the user in the questionnaire; S151: When the questionnaire feedback information corresponding to the key question is blank, outputting the preset estimated answer standard as the user behavior information; S152: When the questionnaire feedback information corresponding to the key question is not blank, continue to determine whether the questionnaire feedback information is consistent with the estimated answer standard; S153: When the questionnaire feedback information is inconsistent with the estimated answer standard, determining whether the facial micro-expression image is consistent with a preset unexpected micro-expression; S154: When the facial micro-expression image is consistent with the unexpected micro-expression, the user's answer is defined as abnormal, and the estimated answer standard is output as the user behavior information.

8. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 7 is characterized in that: The questionnaire includes a paper questionnaire and an electronic questionnaire; the behavior identification method also includes: S1501: Based on the paper questionnaire, collect the questionnaire image information; S1502: Identify a preset writing area from the paper image information to determine the writing content; S1503: Determine whether the written content contains a preset erasure feature; S1504: When the written content contains the erasure feature, the written range is identified from the paper image information to determine the paper question and answer line where the erasure feature is located, and whether the paper question and answer line is related to the emotional direction; S1505: When the question and answer items on the paper are about sentiment, identifying the altered features from the paper image information to determine the estimated original content; S1506: Combining the estimated original content and the estimated answer standard to obtain user behavior information.

9. The method for constructing personalized tourism consumption scenarios based on behavior graphs according to claim 1 is characterized in that: Methods for combining the selected scenarios to form tourism consumption scenarios include: S60: Analyze the behavior graph to obtain the user's travel preferences; S61: Determine a travel route based on the user's travel preference plan; S62: Matching the user's travel preferences with the screened scenarios; S63: Arranging the filtered scenes in sequence based on the travel path to obtain a travel scene arrangement; S64: Form a tourism consumption scenario based on the itinerary scenario arrangement.

10. A personalized tourism consumption scenario construction system based on behavior graph, characterized by: include: Acquisition module, used to collect user behavior information; A memory for storing a program for the method for constructing a personalized tourism consumption scenario based on a behavior graph according to any one of claims 1 to 9; The program in the processor and the memory can be loaded and executed by the processor to realize a method for constructing personalized tourism consumption scenarios based on behavior graphs.

Images