Search

CN-122022940-A - Tourism recommendation platform based on knowledge graph

CN122022940ACN 122022940 ACN122022940 ACN 122022940ACN-122022940-A

Abstract

The invention relates to the technical field of travel recommendation platforms and discloses a travel recommendation platform based on a knowledge graph, which comprises a multi-role cognitive portrait construction module, a cognitive enhancement type dynamic knowledge graph construction module, a multi-role demand fusion and conflict detection module, a cognitive suitability travel generation and optimization module, an experience feedback and cognitive model iteration module and a experience feedback and cognitive model iteration module, wherein the multi-role cognitive portrait construction module is used for constructing a structured multi-role cognitive portrait, the cognitive enhancement type dynamic knowledge graph construction module is used for marking specific requirements of travel resource entities on cognitive abilities of users, the multi-role demand fusion and conflict detection module is used for identifying potential conflicts and generating a coordination strategy, the cognitive suitability travel generation and optimization module is used for generating a travel scheme which gives consideration to cognitive demands of all members, and the experience feedback and cognitive model iteration module is used for calibrating cognitive portraits and resource cognitive attributes. According to the scheme, the dynamic knowledge graph integrating the user cognitive portraits and the resource cognitive attributes is constructed, so that the fine modeling of the multi-role group travel demands is realized, the elastic journey considering the cognitive suitability, the space-time constraint and the budget is generated, the feedback closed-loop continuous optimization model and the resource labeling are relied on, and the participation degree and the journey feasibility of the heterogeneous group are improved.

Inventors

  • LIU HAO
  • JIANG LIHUI
  • WANG QINTING
  • WANG ZAORONG

Assignees

  • 浙江深大智能科技有限公司

Dates

Publication Date
20260512
Application Date
20251217

Claims (10)

  1. 1. The travel recommendation platform based on the knowledge graph is characterized by comprising the following modules: the multi-role cognitive portrayal construction module is used for converting the information of the staff in the same row input by the user into a structured multi-role cognitive portrayal so as to provide basis for subsequent personalized recommendation; The cognition enhancement type dynamic knowledge graph construction module is used for labeling specific requirements of each travel resource entity on cognition capability of a user, obtaining resource cognition attributes and constructing a knowledge graph with a resource view and a cognition demand view; the multi-role demand fusion and conflict detection module is used for integrating user preference, multi-role cognitive portraits and cognitive enhancement dynamic knowledge maps, identifying potential conflicts in resource selection in a group and generating a structured coordination strategy; the cognitive suitability travel generation and optimization module is used for generating a travel scheme meeting the constraints of budget, time and space and considering the cognitive demands of all members based on the candidate travel resource set after conflict coordination; And the experience feedback and cognition model iteration module is used for collecting experience feedback of the user after actual traveling, calibrating cognition portrait and resource cognition attribute and realizing continuous learning and optimization of the platform.
  2. 2. The knowledge-graph-based travel recommendation platform of claim 1, wherein the multi-persona cognitive representation construction module comprises: the character semantic analysis unit is used for filling in the identity description of each peer item by item in the platform input interface by a user; the cognitive characteristic reasoning unit is used for calling a cognitive characteristic reasoning model trained based on a large-scale travel user behavior data set and predicting key cognitive indexes; And the cognitive portrayal storage and indexing unit is used for storing the finally output multi-role cognitive portrayal in a user session context in a structured object mode.
  3. 3. The travel recommendation platform based on knowledge graph according to claim 2, wherein the identity description in the character semantic parsing unit comprises character type, age interval, educational background, interest trend, capability restriction statement.
  4. 4. The knowledge-graph-based travel recommendation platform of claim 3, wherein in the cognitive characteristic reasoning unit of the multi-role cognitive portrayal construction module, the cognitive index comprises: The character understanding capability reflects the obtaining efficiency of the user to the written information such as the display board, the notice board and the like; The visual and auditory perception capability is used for evaluating the receiving quality of the user on image details and voice navigation content; Attention maintenance capability, representing the effective participation duration of the user in a single activity; Abstract concept understanding capability, a measure of the user's degree of confidence in non-specific content.
  5. 5. The knowledge graph-based travel recommendation platform of claim 4, wherein in the multi-role cognitive portrayal construction module, if a user provides an explicit capability restriction statement, the platform preferentially uses the capability restriction statement to override a model default inference result and marks the model default inference result as a high confidence feature.
  6. 6. The knowledge-based travel recommendation platform of claim 1, wherein the cognition enhancement dynamic knowledge construction module comprises: The platform integrates the multi-source data, extracts the basic information of tourist sites and projects, and stores all the data into a map database after cleaning, deduplication and entity alignment; the cognition attribute automatic labeling unit supplements cognition related attributes for each entity; and the dynamic updating and manual checking unit is used for regularly grabbing the latest comments and operation notices by adopting a daily increment updating mechanism, automatically adjusting the cognitive attribute value and opening a checking interface of the travel mechanism.
  7. 7. The knowledge-graph-based travel recommendation platform of claim 1, wherein the multi-role demand fusion and conflict detection module comprises: the individual suitability scoring unit is used for traversing the candidate travel resource set and calculating the matching degree of each resource and each peer; the group experience balance evaluation unit is used for determining the overall acceptance of the resource by the score of the least suitable member; and the conflict classification and coordination strategy generation unit classifies the conflict according to the conflict source, and provides a solution for each type of conflict by a platform built-in strategy library.
  8. 8. The knowledge-based travel recommendation platform as recited in claim 7, wherein said conflict root classification in said conflict classification and coordination strategy generation unit comprises: cognitive exclusion type conflicts in that the cognitive abilities of some members are insufficient to meet the basic interactive requirements of the travel resource; The physiological-rhythm misplacement type conflict is that the physical tolerance, the activity rhythm and the attention duration of the members are different, so that the schedule is difficult to consider; Interest splitting type conflicts-members have fundamental preference differences for travel topics, activity types, and are difficult to satisfy through a single resource.
  9. 9. The knowledge-based travel recommendation platform of claim 1, wherein the cognitive adaptation trip generation and optimization module comprises: The main travel skeleton generating unit is used for searching a feasible path in the cognition enhancement type dynamic knowledge graph by utilizing a space-time constraint satisfaction problem solving algorithm; The cognitive compensation micro-experience filling unit is used for inserting the micro-experience units into gaps of a trunk stroke, and the design of the micro-experience units follows the principle of cognitive load alternation; and the elastic alternative generating unit generates at least one set of alternatives for each main pushing stroke.
  10. 10. The knowledge-based travel recommendation platform of claim 1, wherein said experience feedback and cognition model iteration module comprises: The platform pushes a lightweight feedback questionnaire to a user after the stroke is finished; the system compares the feedback data with the original cognitive portraits and the resource attributes to identify the prediction deviation; and the cold start and template optimization unit is used for providing a plurality of typical cognitive templates for a new user.

Description

Tourism recommendation platform based on knowledge graph Technical Field The invention relates to the technical field of travel recommendation platforms, in particular to a travel recommendation platform based on a knowledge graph. Background The existing travel recommendation platform based on the knowledge graph generally integrates multi-source information such as destination related scenic spots, hotels, restaurants, traffic and play items in a structured manner, constructs a static or semi-dynamic knowledge graph, performs entity retrieval and sorting by combining user input preferences such as budgets, travel dates, people numbers, interest labels and the like, finally generates scenic spot lists or simple travel suggestions, and generally depends on predefined entity attributes such as ticket prices, open time, scores and user historical behavior data, realizes recommendation in a collaborative filtering, graph embedding or rule matching mode and the like, and has the core aim of providing reasonable travel options under limited resource constraints. However, the prior art fails to model individual cognitive ability differences in a user group in fine granularity, specifically, a user portrait in the prior system generally only comprises surface features such as age, gender and the like, quantitative descriptions of cognitive dimensions such as word understanding ability, visual/auditory perception ability, attention maintenance duration, abstract concept understanding ability and the like are lacked, meanwhile, a travel resource entity is not marked with metadata related to cognitive interaction requirements, therefore, even if the system recognizes that 'family goes out and contains old people and children', whether the old people or children have the ability of understanding specific display contents cannot be judged, and the recommendation result is seriously mismatched with actual experience, and the direct consequence is that part of members cannot effectively participate due to too high cognitive threshold, so that the overall travel completion rate and satisfaction are reduced, and the system is prevented from being applicable in high-potential scenes such as economic and special educational travel. Disclosure of Invention The invention aims to provide a travel recommendation platform based on a knowledge graph, which integrates individual cognitive characteristics as key dimensions into the whole process of knowledge graph construction and recommendation decision making, and realizes the refinement and inclusive trip planning oriented to heterogeneous groups through multi-role modeling, cognitive demand mapping and experience balanced optimization. In order to achieve the above purpose, the present invention adopts the following technical scheme: A travel recommendation platform based on a knowledge graph comprises a multi-role cognition portrait construction module, a cognition enhancement dynamic knowledge graph construction module, a multi-role demand fusion and conflict detection module, a cognition suitability travel generation and optimization module and an experience feedback and cognition model iteration module, wherein the multi-role cognition portrait construction module is used for converting information of a peer person input by a user into a structured multi-role cognition portrait, the cognition enhancement dynamic knowledge graph construction module is used for labeling specific requirements of each travel resource entity on the cognition capability of the user, the multi-role demand fusion and conflict detection module is used for identifying potential conflicts in resource selection in a group and generating a structured coordination strategy, the cognition suitability travel generation and optimization module is used for generating a travel scheme meeting budgeting, time and space constraints and considering cognition demands of each member based on a candidate travel resource set after conflict coordination, and the experience feedback and cognition model iteration module is used for collecting experience feedback after the user actually travels and calibrating cognition portraits and resource cognition attributes. The multi-role cognitive portrait construction module comprises a role semantic analysis unit, a cognitive feature reasoning unit, a cognitive portrait storage and indexing unit and a multi-role cognitive portrait storage and indexing unit, wherein the role semantic analysis unit is used for filling in identity description of each peer item by item on a platform input interface, the cognitive feature reasoning unit is used for calling a cognitive feature reasoning model formed by training based on a large-scale text travel user behavior data set, and the cognitive portrait storage and indexing unit is used for storing finally output multi-role cognitive portraits in a user session context in a structured object mode. As a preferable technical scheme of the inven