AI travel strategy automatic generation method and system based on large language model

Abstract

The invention belongs to the technical field of travel strategy making. The invention provides an AI travel strategy automatic generation method and system based on a large language model. The method comprises the following steps: calling a large language model to perform semantic understanding on tourism interest description information input by a user to obtain tourism interest characteristics; a plurality of tourist nodes are determined according to the tourist destination and the tourist interest characteristics, each tourist node corresponds to at least one tourist site of the same kind, and the tourist nodes are preliminarily sorted according to the number of the tourist sites corresponding to the tourist nodes; performing sorting adjustment on the preliminarily sorted tourist nodes according to the interestingness of the user on the tourist nodes; and obtaining touring sub-strategies corresponding to all the tourist nodes, and matching all the touring sub-strategies with all the tourist nodes after sorting adjustment one by one to obtain a target tourist strategy. According to the invention, the strategy alternative is set in each travel node of the travel strategy, so that the user can obtain good touring experience.

Description

Technical Field

[0001] The present invention relates to the technical field of travel guide production, and in particular to an AI travel guide automatic generation method and system based on a large language model. Background Art

[0002] Travel guides are guides that provide travelers with practical information such as destination information, itinerary planning, transportation and accommodation suggestions, dining recommendations, scenic spot introductions, cultural customs explanations, safety tips, etc. It helps travelers better plan and enjoy their journeys and avoid common travel pitfalls and inconveniences.

[0003] At present, some software can realize the automatic generation of travel guides, which can greatly improve the efficiency of users in making travel guides. However, the travel guides generated by the existing methods are basically fixed, and the original travel guides are difficult to implement due to various situations during the travel process. Tourists have to manually adjust the travel guides or simply give up the travel guides, which will undoubtedly significantly reduce the travel experience of tourists. The present invention aims to solve this technical problem. Summary of the invention

[0004] In this regard, the present invention provides an AI travel guide automatic generation method, system, electronic device, computer storage medium and computer program product based on a large language model to solve the above technical problems.

[0005] The present invention also provides an AI travel guide automatic generation method based on a large language model, the method comprising the following steps: Receiving tourist destination and tourist interest description information input by a user, calling a large language model to perform semantic understanding on the tourist interest description information to obtain tourist interest features; Determine a plurality of tourist nodes according to the tourist destination and the tourist interest characteristics, each tourist node corresponds to at least one tourist spot of the same type, and preliminarily sort each tourist node according to the number of tourist spots corresponding to each tourist node; Evaluate the interest of each of the tourist nodes according to the tourist interest characteristics, and adjust the ranking of each of the initially ranked tourist nodes according to the obtained interest; Acquire the tour sub-strategies corresponding to each of the tourist nodes, pair each of the tour sub-strategies with each of the tourist nodes after the sorting adjustment one by one, and obtain the target tourist strategy.

[0006] In some embodiments, the receiving of the tourist destination and tourist interest description information input by the user includes: receiving the tourist destination input by the user, evaluating and obtaining popularity information of the tourist destination, and generating prompt information according to the popularity information; wherein the prompt information includes a plurality of key prompt words, and the number of the key prompt words is related to the popularity information; The prompt information is output, and the travel interest description information input by the user according to the prompt information is received.

[0007] In some embodiments, the determining of a plurality of tourist nodes according to the tourist destination and the tourist interest characteristic information includes: According to the tourist destination and the tourist interest characteristic information, a plurality of tourist nodes are obtained by obtaining and screening from a plurality of designated network ports.

[0008] In some embodiments, the preliminary sorting of the tourist nodes according to the number of tourist spots corresponding to the tourist nodes includes: Determining whether the user has traveled to the travel destination; If yes, then sort the tourist nodes in positive order from sequence number 1 to sequence number n according to the number of tourist spots corresponding to each tourist node, that is, the tourist node with the largest number of corresponding tourist spots has the largest sequence number, and the tourist node with the smallest number of corresponding tourist spots has the smallest sequence number; If not, the tourist node with the highest popularity is selected and its serial number is determined as 1; for the remaining tourist nodes, they are sorted in positive order from serial number 2 to serial number n according to the number of tourist attractions corresponding to each tourist node, that is, the tourist node with the largest number of corresponding tourist attractions has the largest serial number, and the tourist node with the smallest number of corresponding tourist attractions has the smallest serial number; This completes the preliminary sorting.

[0009] In some embodiments, evaluating the interest of each of the tourist nodes according to the tourist interest characteristics, and adjusting the ranking of each of the initially ranked tourist nodes according to the obtained interest, includes: Acquire several tour sub-strategies associated with each of the tourism nodes, and construct each of the tour sub-strategies as a tourism node attachment; Based on the tourist interest features and the tourist node attachments, questions are asked to the general large model, and the answer content of the general large model is semantically analyzed using the large language model to obtain the interest level of each tourist node; The tourist nodes whose corresponding number of tourist spots is higher than the number threshold are selected as target tourist nodes, and the target tourist nodes which are preliminarily sorted are sorted in positive order according to the obtained interest levels, that is, the target tourist nodes are sorted from high to low according to the interest levels; At this point, the sorting adjustment is completed.

[0010] In some embodiments, after pairing each of the tour sub-strategies with each of the tourism nodes after the sorting adjustment one by one, the method further includes: Configuring an opening time for each of the tourism nodes; The reminder timing of each of the tourist nodes is determined according to the interest level; wherein the reminder timing is set earlier than the corresponding opening time, and the set time is positively correlated with the interest level.

[0011] The present invention also provides an AI travel strategy automatic generation system based on a large language model, the system comprising an interest feature extraction module, a sequence generation module, a sequence adjustment module, and a strategy generation module; The interest feature extraction module is used to receive the tourist destination and tourist interest description information input by the user, and call the large language model to perform semantic understanding on the tourist interest description information to obtain tourist interest features; The sequence generation module is used to determine a number of tourist nodes according to the tourist destination and the tourist interest characteristics, each tourist node corresponds to at least one tourist spot of the same type, and preliminarily sort each tourist node according to the number of tourist spots corresponding to each tourist node; The sequence adjustment module is used to evaluate the interest of each of the tourist nodes according to the tourist interest characteristics, and adjust the order of each of the initially ordered tourist nodes according to the obtained interest; The strategy generation module is used to obtain the tour sub-strategies corresponding to each of the tourist nodes, and to pair each of the tour sub-strategies with each of the tourist nodes after the sorting adjustment one by one to obtain the target tourist strategy.

[0012] The present invention also provides an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement any of the methods described above.

[0013] The present invention also provides a computer storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement any of the above methods.

[0014] The present invention also provides a computer program product, which includes computer code. When the computer code is executed by a processor of an electronic device, any of the above methods is implemented.

[0015] The solution of the present invention first determines each tourist node and its ranking according to the user input information, and then fine-tunes the ranking of the tourist nodes according to the user's interest in each tourist node, and finally matches the crawled tour sub-strategies with the corresponding tourist nodes one by one, thereby realizing the automatic generation of the tourist strategy.

[0016] At the same time, each tourist node in the travel guide generated by the present invention includes multiple similar tourist spots, realizing alternative guides for the same type of attractions. When some tourist spots cannot be visited or are not well visited for some reason, alternative tourist spots can be switched in time, and there is no need to manually generate travel guides, thereby ensuring that users have a good sightseeing experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flowchart of a method for automatically generating an AI travel guide based on a large language model disclosed in an embodiment of the present invention; Figure 2 is a schematic diagram of preliminary sorting of tourist nodes disclosed in an embodiment of the present invention; Figure 3 It is a structural diagram of an AI travel guide automatic generation system based on a large language model disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following is a description of the implementation of the present application by specific specific embodiments. People familiar with the technology can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0020] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0021] like Figure 1 As shown, the embodiment of the present invention discloses a method for automatically generating an AI travel guide based on a large language model, and the method comprises the following steps: Receiving tourist destination and tourist interest description information input by a user, calling a large language model to perform semantic understanding on the tourist interest description information to obtain tourist interest features; Determine a plurality of tourist nodes according to the tourist destination and the tourist interest characteristics, each tourist node corresponds to at least one tourist spot of the same type, and preliminarily sort each tourist node according to the number of tourist spots corresponding to each tourist node; Evaluate the interest of each of the tourist nodes according to the tourist interest characteristics, and adjust the ranking of each of the initially ranked tourist nodes according to the obtained interest; Acquire the tour sub-strategies corresponding to each of the tourist nodes, pair each of the tour sub-strategies with each of the tourist nodes after the sorting adjustment one by one, and obtain the target tourist strategy.

[0022] In the above scheme of the present invention, the user inputs a tourist destination and a description of tourist interests. The description of tourist interests is some personal interest information of the user for the tourist destination, such as "I found on a certain APP that Zhoushan has many archipelagos, beautiful sea views, and can pray and watch the sunrise. Please make a travel guide for me." The large language model can perform semantic understanding on the tourist interest description information, thereby obtaining the tourist interest characteristics of the user. According to the tourist destination and tourist interest characteristic information, several tourist nodes can be determined, wherein each tourist node corresponds to at least one similar tourist spot, such as XX temple and YY temple, XX beach and YY beach. After the tourist nodes are determined, these tourist nodes need to be sorted to determine the user's tour order. Specifically, each tourist node is preliminarily sorted according to the number of tourist attractions corresponding to each tourist node. For example, tourist nodes with more alternative tourist attractions can be ranked in the front sequence to improve the anti-change ability of the tourist node; then, the interest of each tourist node is evaluated according to the tourist interest characteristics, and the ranking of each tourist node preliminarily sorted is adjusted accordingly; finally, the tour sub-strategies corresponding to each tourist node are crawled and screened out through the Internet, and each tour sub-strategy is matched one by one with each tourist node after the adjustment of the ranking, so as to generate the corresponding target tourist strategy.

[0023] The above-mentioned travel guide automatic generation solution of the present invention can be implemented in various forms, such as H5 web pages, PC web pages, mini programs or APPs, etc., and the present invention does not make specific limitations.

[0024] Since large language models are generally large in scale, direct integration will result in an excessively large module size, so the present invention uses the large language model by calling. The large language model in the present invention is preferably constructed based on a Transformer deep neural network.

[0025] In some embodiments, the receiving of the tourist destination and tourist interest description information input by the user includes: receiving the tourist destination input by the user, evaluating and obtaining popularity information of the tourist destination, and generating prompt information according to the popularity information; wherein the prompt information includes a plurality of key prompt words, and the number of the key prompt words is related to the popularity information; The prompt information is output, and the travel interest description information input by the user according to the prompt information is received.

[0026] In this embodiment, in order to improve the accuracy of obtaining tourist interest features, the present invention prompts the user to input tourist interest description information of the specified content through prompt information. Specifically, after receiving the tourist destination input by the user, the popularity information of the tourist destination is first evaluated, for example, it can be evaluated by network search volume, discussion volume, etc.; then, the number of key prompt words is determined according to the heat information obtained by the evaluation, and the corresponding prompt information is generated according to these key prompt words; finally, the prompt information is output to the user, and the user can enter the corresponding tourist interest description information under the guidance of the prompt information. Among them, the key prompt words can be, for example, "exploration", "food", "history", "art", "shopping", etc. These key prompt words can be part of the prompt word library. Specifically, the prompt word library determines the priority for each key prompt word according to the praise of the travel guide, and the key prompt words with high priority are ranked first, so that several key prompt words with high priority can be screened out according to the above-determined number.

[0027] In addition, the more popular a tourist destination is, the more places it can visit. At this time, more key prompts should be configured to obtain more accurate tourist interest description information. On the contrary, the less popular a tourist destination is, the fewer places it can visit are. In addition, because users generally do not know enough about tourist destinations that are not very popular, it is difficult to answer certain key prompts. By reducing the number of key prompts appropriately, the difficulty of users entering tourist interest description information can be reduced. Obviously, the number of key prompts should be positively correlated with the popularity information.

[0028] It should be noted that the prompt information can be generated multiple times to allow the user to input additional information about the travel interest description. Moreover, the user can also initiate active questions, that is, actively input the travel interest description information without prompt information, to assist the large language model to fully understand the user's travel intention.

[0029] In some embodiments, the determining of a plurality of tourist nodes according to the tourist destination and the tourist interest characteristic information includes: According to the tourist destination and the tourist interest characteristic information, a plurality of tourist nodes are obtained by obtaining and screening from a plurality of designated network ports.

[0030] In this embodiment, the designated network port is a tourism-related website, public account, APP, etc. The designated network port can avoid information search in the entire network to reduce the amount of invalid data and increase the speed of determining the tourism node. Finally, according to the tourism destination and tourism interest feature information, several tourism nodes are obtained and screened from several designated network ports.

[0031] In some embodiments, the preliminary sorting of the tourist nodes according to the number of tourist spots corresponding to the tourist nodes includes: Determining whether the user has traveled to the travel destination; If yes, then sort the tourist nodes in positive order from sequence number 1 to sequence number n according to the number of tourist spots corresponding to each tourist node, that is, the tourist node with the largest number of corresponding tourist spots has the largest sequence number, and the tourist node with the smallest number of corresponding tourist spots has the smallest sequence number; If not, the tourist node with the highest popularity is selected and its serial number is determined as 1; for the remaining tourist nodes, they are sorted in positive order from serial number 2 to serial number n according to the number of tourist attractions corresponding to each tourist node, that is, the tourist node with the largest number of corresponding tourist attractions has the largest serial number, and the tourist node with the smallest number of corresponding tourist attractions has the smallest serial number; This completes the preliminary sorting.

[0032] In this embodiment, when sorting the determined tourist nodes, the present invention takes into account whether the user has traveled to the tourist destination, thereby achieving a more reasonable sorting.

[0033] Specifically, if it is determined that the user has traveled to this tourist destination, it means that the user has most likely visited some of the most popular attractions in the tourist destination. If such attractions are placed in the front tour sequence number, it does not meet the actual expectations of the user. Moreover, such an arrangement may also make it difficult for the user to visit new attractions on this trip. For example, sudden factors make it impossible for the user's subsequent tour itinerary to be realized, and the new attractions ranked later will not be able to be visited. In this regard, the present invention arranges a natural sorting from high to low according to the number of tourist locations corresponding to each tourist node. Such a setting can make tourist nodes with alternative options more forward, and the corresponding anti-change ability is stronger. When the corresponding attractions are temporarily closed or too congested, the user can promptly select other similar alternative attractions in the guide, thereby ensuring that the travel experience is at a high level.

[0034] If it is determined that the user has not traveled to the tourist destination, it means that the user has probably not visited some of the most popular attractions in the tourist destination. At this time, the most popular tourist node is ranked first, and the remaining tourist nodes are also ranked from high to low according to the number of corresponding tourist locations, such as Figure 2 Such a setting can ensure that the "new tourists" complete the tour of the most popular attractions in the tourist destination, and on this basis can also visit some other non-hot attractions, so as to obtain a good tour experience.

[0035] Among them, based on the historical record data of the user on the corresponding travel website / APP or even the travel guide automatic generation website / APP of the present invention, it can be known whether the user has traveled to the current travel destination. The user can also attach relevant information of whether he has traveled to the current travel destination to the travel interest description information, so that it can also be known whether the user has traveled to the current travel destination. The present invention does not limit the specific implementation method.

[0036] Among them, when screening the most popular tourist nodes, you can first screen out the most popular tourist spots in each tourist node, then compare the most popular tourist spots in each tourist node horizontally, and then screen out the most popular tourist spots in the horizontal comparison, and determine the tourist node to which it belongs as the most popular tourist node.

[0037] In some embodiments, evaluating the interest of each of the tourist nodes according to the tourist interest characteristics, and adjusting the ranking of each of the initially ranked tourist nodes according to the obtained interest, includes: Acquire several tour sub-strategies associated with each of the tourism nodes, and construct each of the tour sub-strategies as a tourism node attachment; Based on the tourist interest features and the tourist node attachments, questions are asked to the general large model, and the answer content of the general large model is semantically analyzed using the large language model to obtain the interest level of each tourist node; The tourist nodes whose corresponding number of tourist spots is higher than the number threshold are selected as target tourist nodes, and the target tourist nodes which are preliminarily sorted are sorted in positive order according to the obtained interest levels, that is, the target tourist nodes are sorted from high to low according to the interest levels; At this point, the sorting adjustment is completed.

[0038] In this embodiment, on the basis of the aforementioned preliminary sorting based on the number of tourist spots included, the present invention further adjusts the sorting of the preliminary sorting results according to the user's interest in each tourist node. Specifically, firstly, each tourist node in the preliminary sorting queue whose number of tourist spots is higher than the number threshold is selected as the target tourist node, and then these target tourist nodes are sorted from high to low according to the size of the interest, and the sorting adjustment is completed.

[0039] Among them, the determination of the interest of each tourist node is realized by a general large model, which refers to large models such as ChatGPT, Wenxin Yiyan, Tongyi Qianwen, etc. Specifically, several tour sub-strategies associated with each tourist node are first obtained, each tour sub-strategy can contain the existing tour guides of each tourist location in the tourist node, and all tour sub-strategies are integrated and processed to construct a tourist node attachment. Then the above-mentioned tourist interest characteristics and the tourist node attachment are sent to the general large model to ask questions. For example, the question sentence is "Attachment 1 is the guide information of some tourist nodes, and Attachment 2 is the user's "tourism interest characteristics". Please analyze the user's interest analysis of each tourist node based on Attachment 1 and Attachment 2". Of course, "tourism interest characteristics" can also be directly asked to the general large model. Finally, after the general large model outputs the answer content of the user's interest in each tourist node, the aforementioned large language model is used to perform semantic analysis on the answer content, so as to obtain the user's interest in each tourist node, and the interest can be expressed in the form of a value or a level.

[0040] It should be noted that the general large model is used to analyze the tourist node attachments containing guides for multiple tourist spots of the corresponding tourist node, that is, the user's overall interest in the tourist node is determined by synchronously analyzing multiple similar tourist spots.

[0041] In addition, the tourist spots corresponding to the tourist nodes can be sorted from high to low according to their corresponding popularity. Of course, it can also be set by the user.

[0042] In some embodiments, after pairing each of the tour sub-strategies with each of the tourism nodes after the sorting adjustment one by one, the method further includes: Configuring an opening time for each of the tourism nodes; The reminder timing of each of the tourist nodes is determined according to the interest level; wherein the reminder timing is set earlier than the corresponding opening time, and the set time is positively correlated with the interest level.

[0043] In this embodiment, after each tour sub-strategy is matched with each tourist node after the sorting adjustment, it is also necessary to configure the opening time for each tourist node. For example, the opening time when the user arrives at the corresponding tourist location is predicted based on the user's transportation mode (self-driving, taxi, public transportation, cycling, walking, etc.), the distribution location of each tourist location, traffic conditions, etc., and the opening time is the predicted time when the user enters the tourist location.

[0044] On this basis, the present invention further determines the reminder timing for each tourist node according to the interest level obtained above. When the reminder timing arrives, the terminal (such as a mobile phone) or server storing the tour guide will obtain the real-time information of the next tourist location. If the next tourist location cannot be visited for some reason, the user can be notified in time to switch to the alternative tourist location in the tourist node, so that the user can ensure the completion rate of the journey by visiting other similar tourist locations, thereby obtaining a good travel experience.

[0045] Among them, the reminder timing is earlier than the corresponding opening time setting duration, and is positively correlated with the interest level. When the user's interest in the next tourist node is higher, the set duration is set to be longer, that is, the above-mentioned "reminder" is performed earlier, so as to ensure the user's travel satisfaction with the scenic spots that he is interested in; on the contrary, when the user's interest in the next tourist node is lower, the set duration is set to be shorter, that is, the degree of advance "reminder" is appropriately reduced or the above-mentioned reminder timing is not adjusted, so as to reduce the inconvenience of adjusting the tour route in the later stage of the tour, because adjusting the tour route in the later stage of the tour may cause the time-consuming journey to cause the user to be unable to arrive at the alternative tourist location in time even if the alternative tourist location is changed.

[0046] like Figure 3 As shown, an embodiment of the present invention further discloses an AI travel guide automatic generation system based on a large language model, the system comprising an interest feature extraction module, a sequence generation module, a sequence adjustment module, and a guide generation module; The interest feature extraction module is used to receive the tourist destination and tourist interest description information input by the user, and call the large language model to perform semantic understanding on the tourist interest description information to obtain tourist interest features; The sequence generation module is used to determine a number of tourist nodes according to the tourist destination and the tourist interest characteristics, each tourist node corresponds to at least one tourist spot of the same type, and preliminarily sort each tourist node according to the number of tourist spots corresponding to each tourist node; The sequence adjustment module is used to evaluate the interest of each of the tourist nodes according to the tourist interest characteristics, and adjust the order of each of the initially ordered tourist nodes according to the obtained interest; The strategy generation module is used to obtain the tour sub-strategies corresponding to each of the tourist nodes, and to pair each of the tour sub-strategies with each of the tourist nodes after the sorting adjustment one by one to obtain the target tourist strategy.

[0047] An embodiment of the present invention further discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method described in the above embodiment.

[0048] An embodiment of the present invention further discloses a computer storage medium, wherein the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method described in the above embodiment.

[0049] An embodiment of the present invention further discloses a computer program product, which includes computer code. When the computer code is executed by a processor of an electronic device, the method described in the above embodiment is implemented.

[0050] The various component embodiments of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) can be used in practice to implement some or all functions of some or all components in the device according to an embodiment of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., computer program and computer program product) for executing part or all of the methods described herein. Such a program implementing the present invention can be stored on a computer-readable medium, or can have the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0051] References herein to "one embodiment," "embodiment," or "one or more embodiments" mean that a particular feature, structure, or characteristic described in conjunction with the embodiment is included in at least one embodiment of the present invention. In addition, please note that examples of the term "in one embodiment" do not necessarily all refer to the same embodiment.

[0052] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0053] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

[0054] In addition, it should be noted that the language used in this specification is selected primarily for readability and teaching purposes, rather than for explaining or limiting the subject matter of the present invention. Therefore, many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the appended claims. The disclosure of the present invention is illustrative rather than restrictive with respect to the scope of the present invention, and the scope of the present invention is defined by the appended claims.

Claims

1. An AI travel guide automatic generation method based on a large language model, characterized by: The method comprises the following steps: Receiving tourist destination and tourist interest description information input by a user, calling a large language model to perform semantic understanding on the tourist interest description information to obtain tourist interest features; Determine a plurality of tourist nodes according to the tourist destination and the tourist interest characteristics, each tourist node corresponds to at least one tourist spot of the same type, and preliminarily sort each tourist node according to the number of tourist spots corresponding to each tourist node; Evaluate the interest of each of the tourist nodes according to the tourist interest characteristics, and adjust the ranking of each of the initially ranked tourist nodes according to the obtained interest; Acquire the tour sub-strategies corresponding to each of the tourist nodes, pair each of the tour sub-strategies with each of the tourist nodes after the sorting adjustment one by one, and obtain the target tourist strategy.

2. According to claim 1, the method for automatically generating AI travel guides based on a large language model is characterized by: Receive the tourist destination and tourist interest description information entered by the user, including: receiving the tourist destination input by the user, evaluating and obtaining popularity information of the tourist destination, and generating prompt information according to the popularity information; wherein the prompt information includes a plurality of key prompt words, and the number of the key prompt words is related to the popularity information; The prompt information is output, and the travel interest description information input by the user according to the prompt information is received.

3. According to claim 2, the method for automatically generating AI travel guides based on a large language model is characterized by: According to the tourist destination and the tourist interest characteristic information, a number of tourist nodes are determined, including: According to the tourist destination and the tourist interest characteristic information, a plurality of tourist nodes are obtained by obtaining and screening from a plurality of designated network ports.

4. According to claim 3, the method for automatically generating AI travel guides based on a large language model is characterized by: Preliminary sorting of each of the tourist nodes is performed according to the number of tourist spots corresponding to each of the tourist nodes, including: Determining whether the user has traveled to the travel destination; If yes, then sort the tourist nodes in positive order from sequence number 1 to sequence number n according to the number of tourist spots corresponding to each tourist node, that is, the tourist node with the largest number of corresponding tourist spots has the largest sequence number, and the tourist node with the smallest number of corresponding tourist spots has the smallest sequence number; If not, the tourist node with the highest popularity is selected and its serial number is determined as 1; for the remaining tourist nodes, they are sorted in positive order from serial number 2 to serial number n according to the number of tourist attractions corresponding to each tourist node, that is, the tourist node with the largest number of corresponding tourist attractions has the largest serial number, and the tourist node with the smallest number of corresponding tourist attractions has the smallest serial number; This completes the preliminary sorting.

5. According to claim 4, the method for automatically generating AI travel guides based on a large language model is characterized by: The interest level of each of the tourist nodes is evaluated according to the tourist interest characteristics, and the ranking of each of the tourist nodes that are initially ranked is adjusted according to the obtained interest level, including: Acquire several tour sub-strategies associated with each of the tourism nodes, and construct each of the tour sub-strategies as a tourism node attachment; Based on the tourist interest features and the tourist node attachments, questions are asked to the general large model, and the answer content of the general large model is semantically analyzed using the large language model to obtain the interest level of each tourist node; The tourist nodes whose corresponding number of tourist spots is higher than the number threshold are selected as target tourist nodes, and the target tourist nodes which are preliminarily sorted are sorted in positive order according to the obtained interest levels, that is, the target tourist nodes are sorted from high to low according to the interest levels; At this point, the sorting adjustment is completed.

6. The method for automatically generating an AI travel guide based on a large language model according to claim 5 is characterized by: After pairing each of the tour sub-strategies with each of the tourism nodes after the sorting adjustment one by one, the method further includes: Configuring an opening time for each of the tourism nodes; The reminder timing of each of the tourist nodes is determined according to the interest level; wherein the reminder timing is set earlier than the corresponding opening time, and the set time is positively correlated with the interest level.

7. An AI travel guide automatic generation system based on a large language model, the system comprising an interest feature extraction module, a sequence generation module, a sequence adjustment module, and a guide generation module; characterized in that: The interest feature extraction module is used to receive the tourist destination and tourist interest description information input by the user, and call the large language model to perform semantic understanding on the tourist interest description information to obtain tourist interest features; The sequence generation module is used to determine a number of tourist nodes according to the tourist destination and the tourist interest characteristics, each tourist node corresponds to at least one tourist spot of the same type, and preliminarily sort each tourist node according to the number of tourist spots corresponding to each tourist node; The sequence adjustment module is used to evaluate the interest of each of the tourist nodes according to the tourist interest characteristics, and adjust the order of each of the initially ordered tourist nodes according to the obtained interest; The strategy generation module is used to obtain the tour sub-strategies corresponding to each of the tourist nodes, and to pair each of the tour sub-strategies with each of the tourist nodes after the sorting adjustment one by one to obtain the target tourist strategy.

8. An electronic device, comprising: At least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method according to any one of claims 1 to 6.

9. A computer storage medium storing a computer program, characterized in that: The computer program is executed by a processor to implement the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that: The computer program product includes computer codes, and when the computer codes are executed by a processor of an electronic device, the method according to any one of claims 1 to 6 is implemented.

Images