Road book processing method and related device

Abstract

The invention provides a road book processing method and a related device, and relates to the field of road book processing. According to the method and the device, when the road book is generated, the road book generation strategy corresponding to the attribute information is determined according to the attribute information of the data input by the user, so that the selected road book generation strategy is matched with the user demand, and the accuracy and the reliability of subsequent road book generation are improved. Besides, after the first road book is generated based on the user demand information and the user portrait, the first road book is subsequently adjusted based on the road book adjustment parameters to obtain the second road book, so that the road book can be adaptively adjusted according to the user preference, weather, traffic and other contents, and the reliability and personalization of the generated road book are improved. After the second road book is obtained, the second road book can be subjected to content optimization and sorting operation, and the content quality of the road book and the user experience feeling are guaranteed. According to the invention, on the basis of intelligently generating the road book, the user experience is improved.

Description

Technical Field

[0001] The present invention relates to the field of road book processing, and more specifically, to a road book processing method and related devices. Background Art

[0002] With the increasing frequency of tourism and business travel, road books, as an important tool for itinerary planning, have a direct impact on users' travel experience due to their quality and degree of personalization.

[0003] Traditional road books often rely on manual writing and have a low level of intelligence. So how to intelligently generate road books is a technical problem that technicians in this field need to solve urgently. Summary of the invention

[0004] In view of this, the present invention provides a road book processing method and related devices to solve the problem of urgently needing to intelligently generate road books.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A road book processing method, comprising:

[0007] Determine user demand information corresponding to user input data of a target user, and obtain a user profile of the target user;

[0008] Determine a road book generation strategy corresponding to the attribute information of the user input data;

[0009] Using the route book generation strategy, generate a first route book corresponding to the user demand information and the user portrait;

[0010] Acquiring a road book adjustment parameter, so as to adjust the first road book based on the road book adjustment parameter to obtain a second road book;

[0011] The second route book is subjected to content optimization and sorting operations to obtain a target route book.

[0012] Optionally, determining user demand information corresponding to the user input data of the target user includes:

[0013] Get user input data;

[0014] If the user input data is voice, perform a voice recognition operation on the voice to obtain a voice recognition result, and perform an intent recognition and / or entity recognition operation on the voice recognition result to obtain user demand information;

[0015] If the user input data is text, intent recognition and / or entity recognition operations are performed on the text to obtain user demand information.

[0016] Optionally, the attribute information of the user input data includes format and data integrity;

[0017] Determining a road book generation strategy corresponding to the attribute information of the user input data includes:

[0018] Determining whether the format of the user input data is a preset template format;

[0019] If so, use the template method as the road book generation strategy;

[0020] If not, calculating the data integrity of the user input data;

[0021] If the data completeness is greater than a preset completeness threshold, a search-based generation method is used as a road book generation strategy;

[0022] If the data integrity is not greater than a preset integrity threshold, the model generation method is used as a road book generation strategy.

[0023] Optionally, using the route book generation strategy to generate a first route book corresponding to the user demand information and the user portrait includes:

[0024] If the road book generation strategy is a model generation method, the user demand information and the user portrait are converted into a vector, and a road book generation model is called to process the vector to obtain a first road book;

[0025] If the road book generation strategy is a search-based generation method, the first road book is retrieved from a pre-configured road book database using the user demand information and the user portrait;

[0026] If the road book generation strategy is a template method, the user demand information and the user portrait are used to retrieve the road book information online, and the road book information is filled into a preset road book template to obtain a first road book.

[0027] Optionally, the road book adjustment parameter includes the behavior preference information of the target user, and the first road book is adjusted based on the road book adjustment parameter to obtain the second road book, including:

[0028] Based on the user interaction operation with the target user, obtaining the behavior preference information of the target user;

[0029] The content in the first path book that is not related to the behavior preference information is adjusted to obtain a second path book.

[0030] Optionally, the road book adjustment parameter includes weather information or traffic information; and adjusting the first road book based on the road book adjustment parameter to obtain a second road book includes:

[0031] Determining weather information and traffic information for each location in the first route book;

[0032] If there is a target location among the locations where the weather information indicates that the weather is preset to be bad or the traffic information indicates that the traffic is congested, outputting a prompt message;

[0033] If a first modification instruction that does not include a designated itinerary is received, determining that the itinerary of the target location corresponds to a replacement itinerary, updating the itinerary of the target location in the first route book to the replacement itinerary, and obtaining a second route book;

[0034] If a second modification instruction including a designated itinerary is received, the itinerary of the target location in the first route book is updated to the designated itinerary in the second modification instruction to obtain a second route book.

[0035] Optionally, the road book adjustment parameter includes: user configuration information obtained by a user operating a road book configuration option or a road book configuration slider; and adjusting the first road book based on the road book adjustment parameter to obtain a second road book includes:

[0036] According to the user configuration information, the content in the first path book is adaptively adjusted to obtain the second path book; the adaptive adjustment includes at least one of content addition, content deletion and content modification.

[0037] Optionally, performing content optimization and sorting operations on the second route book to obtain a target route book includes:

[0038] Performing content optimization operation on the content of the second-path book to obtain a third-path book; the content optimization operation at least includes a grammar checking optimization operation and a semantic understanding optimization operation;

[0039] According to preset sorting rules, each itinerary in the third route book is sorted to obtain a target route book; the preset sorting rules include at least one or more of ratings, popularity, user behavior preference information, and user tags.

[0040] A road book processing device, comprising:

[0041] A data determination module, used to determine user demand information corresponding to user input data of a target user, and to obtain a user portrait of the target user;

[0042] A strategy determination module, used to determine a road book generation strategy corresponding to the attribute information of the user input data;

[0043] A route book generating module, configured to generate a first route book corresponding to the user demand information and the user portrait using the route book generating strategy;

[0044] A road book adjustment module, used for obtaining a road book adjustment parameter, and adjusting the first road book based on the road book adjustment parameter to obtain a second road book;

[0045] The road book optimization module is used to perform content optimization and sorting operations on the second road book to obtain a target road book.

[0046] An electronic device comprises at least one processor and a memory connected to the processor, wherein:

[0047] The memory is used to store computer programs;

[0048] The processor is used to execute the computer program so that the electronic device can implement the above-mentioned road book processing method.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention provides a road book processing method and related devices. In the present invention, when generating a road book, the corresponding road book generation strategy will be determined according to the attribute information of the user input data, so that the selected road book generation strategy matches the user demand, and the accuracy and reliability of the subsequent road book generation are improved. In addition, after the first road book is generated based on the user demand information and the user portrait, the first road book will be adjusted based on the road book adjustment parameters to obtain a second road book, so that the road book can be adaptively adjusted according to the user preference, weather, traffic and other contents, and the reliability and personalization of the generated road book can be improved. After the second road book is obtained, the second road book can be optimized and sorted to ensure the content quality of the road book and the user experience. That is, the present invention improves the user experience on the basis of intelligently generating the road book. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0052] Figure 1 A flowchart of a road book processing method provided by an embodiment of the present invention;

[0053] Figure 2 A flowchart of a method for adjusting road book content provided by an embodiment of the present invention;

[0054] Figure 3 A flowchart of another method for adjusting road book content provided by an embodiment of the present invention;

[0055] Figure 4 A schematic diagram of the structure of a road book processing device provided by an embodiment of the present invention;

[0056] Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] With the increasing frequency of tourism and business travel, road books, as an important tool for itinerary planning, have a direct impact on users' travel experience due to their quality and degree of personalization.

[0059] Traditional road books often rely on manual writing and have a low level of intelligence. So how to intelligently generate road books is a technical problem that technicians in this field need to solve urgently.

[0060] In order to generate road books intelligently, simple rule-based automatic generation can be used, but this method is difficult to meet users' requirements for content richness, personalization and real-time performance. Therefore, a solution is needed to improve the content richness, personalization and real-time generation of road books.

[0061] To this end, the present invention introduces a model to generate a road book. Since the model is trained by a large number of training samples, the training samples will include a large amount of content. Therefore, the road book generated by the model will have a higher content richness and higher real-time performance. In addition, multimodal data such as pictures, videos, and voices can be inserted into the road book to improve the content richness. In addition, the road book content can be intelligently adjusted according to user behavior preferences to meet user personalized needs.

[0062] Based on the above content, an embodiment of the present invention provides a road book processing method, and the execution subject can be a road book generating device such as a vehicle controller and a mobile phone. The application scenario of the present invention can be, for example, when a user is driving a vehicle, the user can input data, and the road book generating device can intelligently generate a road book, and then obtain user behavior preference information based on user interaction, collect weather and traffic information through networking, and the user obtains user configuration information through slider or option configuration, and adaptively adjusts the road book based on this information.

[0063] It should be noted that since user preference information is not collected initially and weather and traffic information will change in real time during the user's travel, the initially generated road book may contain inappropriate parts and will be adjusted later based on user needs, traffic, weather and other information.

[0064] In addition, user preferences (which can be reflected in user portraits) and weather information for the next few days can be collected at the beginning. When the road book is generated, this information is taken into account to obtain a road book that meets user preferences and weather information. The road book can be adaptively adjusted based on traffic information later. In addition, where the user thinks the content of the road book is inappropriate, it can also be adaptively adjusted according to user needs. For example, if the user thinks the itinerary in the road book is not suitable, the content of the road book can be adjusted according to the user's requirements. Alternatively, during the tour, the recorded human-computer dialogue content can be filled into the road book, or pictures or videos of the beautiful scenery ahead can be filled into the road book to achieve the richness of the content of the road book, so that the road book has multimodal data.

[0065] Reference Figure 1 , a road book processing method may include:

[0066] S11. Determine user demand information corresponding to user input data of a target user, and obtain a user portrait of the target user.

[0067] In actual applications, if the user is driving a vehicle or not driving a vehicle, the user input data can be input by text input or voice input. For example, if the user inputs "Please arrange a three-day travel schedule to XXX" by voice, the road book generating device can obtain the user input data.

[0068] After receiving the user input data, the road book generating device starts to analyze the user demand information.

[0069] Generally, the data processed by the road book generation device is text. Therefore, if the user input data is text, user demand analysis can be performed on it. When analyzing user demand, intent recognition and / or entity recognition can be used. Among them, intent recognition can use the intent detection technology in NLP (Natural Language Processing), and entity recognition can use the entity extraction technology in NLP.

[0070] When performing intent recognition, key information such as destination, time, budget, etc. can be accurately extracted from the user input data. This key information is the user demand information.

[0071] Similarly, when performing entity recognition, key information such as destination, time, budget, etc. is also identified.

[0072] If intent recognition and entity recognition are used at the same time, the results of the two recognitions can be fused to obtain the final accurate and reliable user demand information.

[0073] If the user input data is speech, then ASR (Speech Recognition Technology) can be used to perform speech recognition operation on the speech to obtain speech recognition results. Speech recognition technology can be speech-to-text or DeepSpeech, etc. Then, the speech recognition results are subjected to intent recognition and / or entity recognition operations to obtain user demand information. The implementation process thereof refers to the corresponding description above.

[0074] In this embodiment, multiple input methods of user text and voice input are supported to meet the needs of different users.

[0075] When generating a road book, in addition to obtaining user demand information corresponding to user input data, the road book can also be generated based on the user portrait of the target user. In this case, the user portrait of the target user needs to be obtained.

[0076] The process of constructing the user portrait in this embodiment may be as follows:

[0077] When the user initially enters the route book generation interface of the web version or mobile software APP (application) version, there will be click options, which include content that the user is interested in, such as mountain climbing, water playing, escape room, ice sculpture, amusement park, action world, etc. If the user clicks on certain options, it means that the user is more interested in these contents, and a user portrait can be generated based on this.

[0078] In addition, users of different ages like different entertainment content, such as young people like roller coasters and the elderly like fishing. Similarly, different genders and occupations will also lead to different content. Therefore, you can also collect information such as the user's age, gender, occupation, etc. and add these contents to the user portrait.

[0079] You can also use user profile technology to build user profiles based on user historical behavior, preferences and other information, so as to incorporate personalized elements into the content when adjusting the road book.

[0080] In addition, users can also input their own preferences. For example, if a user inputs “I like climbing mountains, especially high mountains, and I like the sea”, etc., a user profile can also be built based on this.

[0081] In addition, if no data is entered, the content of the user portrait can be empty, and the user portrait can be continuously improved based on the interaction with the user.

[0082] S12: Determine a road book generation strategy corresponding to the attribute information of the user input data.

[0083] When users input data, some users like to use a regular format, such as template input. For example, the user input data is:

[0084] 1. Starting point: Beijing

[0085] 2: Destination: Shanghai

[0086] 3. Two attractions to visit every day: XX, YY;

[0087] 4. Accommodation: Five-star

[0088] …

[0089] If it is determined that the format of the user input data is a preset template format, as shown above, it means that the user is a user with high standardization and also wants to output the output road book content in a standardized format. Therefore, the template method can be used as a road book generation strategy. The template method in this embodiment refers to generating a road book according to the format of the template.

[0090] If it is determined that the format of the user input data is not the preset template format, it means that the user has low requirements for the content format, and the road book can be generated by using a search-based generation method or a model generation method.

[0091] Among them, the search formula generation method retrieves from a pre-configured road book database to obtain a road book. The model generation method uses a large model to generate a road book. The model generation method tends to generate freely and interacts with the user many times. Therefore, the requirements for user input data are relatively low, and the number of words can be very small, and the data integrity is very low. However, the search formula generation method requires a certain amount of search terms to be able to search, so it requires a higher data integrity. Then, the data integrity of the user input data can be further calculated, and the corresponding road book generation strategy is selected based on the data integrity.

[0092] In this embodiment, when calculating data completeness, the following data items may be pre-configured, such as departure point, destination, tourist attractions, accommodation, meals, etc., and then the user demand information extracted from the user input data is matched with the above data items, and the ratio of the successfully matched data items to all data items is used as the data completeness. The higher the data completeness, the more detailed the data input by the user.

[0093] If the data integrity is high, such as greater than a preset integrity threshold, the search-based generation method is used as the road book generation strategy. If the data integrity is low, such as not greater than a preset integrity threshold, the model generation method is used as the road book generation strategy.

[0094] S12: Use the route book generation strategy to generate a first route book corresponding to the user demand information and the user portrait.

[0095] In this embodiment, the first route book is an initial version of the route book, which can be adjusted later according to actual conditions.

[0096] In practical applications, there are many ways to implement the generation process of the first path book, which are introduced below:

[0097] 1. Generate the first route book using the model generation method. In this case, the route book generation strategy is the model generation method.

[0098] Specifically, in this embodiment, the preferred model may be LLM (Large Language Model), because it has powerful language understanding and generation capabilities and can provide a new solution for the generation of intelligent road books.

[0099] LLMs suitable for applications in the tourism field include: GPT (Generative Pre-Trained) series, specifically GPT-3, GPT-4, etc., or BERT (Bidirectional EncoderRepresentations from Transformers, a pre-trained language model based on Transformer architecture) series, such as RoBERTa (Robustly Optimized BERT Pretraining Approach, a pre-trained language model based on BERT), ERNIE (Enhanced Representation through Knowledge Integration, a pre-trained language model based on knowledge enhancement), etc., or other models specifically optimized for tourism scenarios.

[0100] The LLM model in this embodiment can be trained using data in the tourism field so that it can generate road books in the tourism field.

[0101] In practical applications, during training, reliable and comprehensive data sources can include: official tourism websites, tourism apps, social media platforms, etc., to ensure the accuracy and real-time nature of the data.

[0102] After determining the data source, you can configure some road book configuration parameters, such as location, travel days, location tags (attractions, planetarium, dinosaur museum, etc.), and then search the data source for road book content and tourist attraction introduction information corresponding to the road book configuration parameters. In addition, the road book configuration parameters can be sent to the data source so that the data source provides matching data.

[0103] After obtaining the data, the road book configuration parameters and the collected data are matched and sent to the model for training. When the loss function is less than a preset threshold or the number of training times reaches the maximum number, the training can be stopped to obtain the LLM model in this embodiment.

[0104] It should be noted that in order to improve processing efficiency, the LLM model can be trained in advance so that the model can be quickly used to generate the road book when the road book is generated later.

[0105] In another implementation of the present invention, in order to ensure the accuracy and reliability of the collected data, a data cleaning operation can be performed after the data is collected to remove duplicate, invalid or erroneous data entries. In addition, during data cleaning, the text data can also be denoised, such as removing stop words, punctuation marks, etc., to avoid the influence of irrelevant words on subsequent training.

[0106] After data cleaning is completed, the data is vectorized. Specifically, the cleaned text data is converted into vector form through word embeddings or sentence embeddings technology for subsequent LLM model processing. Commonly used word embedding models include Word2Vec (word to vector), GloVe (Global Vectors for Word Representation), BERT, etc.

[0107] In practical applications, if the data obtained after data cleaning is words, they are converted into vectors using word embedding. If the data obtained after data cleaning is sentences, they are converted into vectors using sentence embedding.

[0108] The converted vector will be used as the input of the model to carry out the model training process. After the model training is completed, the pre-trained LLM model is loaded into the system to prepare for reasoning and generation tasks. That is, the present invention realizes the intelligent and personalized generation of road book content by introducing LLM technology.

[0109] When using the model, word embedding or sentence embedding can be used to convert the user demand information and the user portrait into a vector, and the road book generation model is called to process the vector to obtain the first road book.

[0110] 2. Retrieve and generate the first route book, that is, the route book generation strategy at this time is the retrieval generation method.

[0111] Specifically, all the data collected above can be stored in a road book database, and then the user demand information and the user portrait can be used to retrieve relevant information from the road book database and recombined through a search formula generation method to obtain a first road book.

[0112] This method is suitable for scenarios where the data stored in the road book database is relatively comprehensive. Only then can the corresponding road book content be searched in the road book database. If no results are found, the above model can be used to generate the first road book.

[0113] 3. Use the template method to generate the first route book, that is, the route book generation strategy at this time is the template method.

[0114] In this embodiment, in order to achieve standardization of the generated road book, a road book template may be pre-constructed, and the road book template may be as follows:

[0115] 1. Starting point;

[0116] 2: Destination;

[0117] 3. Two attractions to visit every day;

[0118] 4. Accommodation;

[0119] …

[0120] The road book template can be obtained in advance based on manual experience or by using a model to conduct statistical analysis on common tourism data.

[0121] In this embodiment, in order to obtain as comprehensive data as possible, an online search method can be used to search for road book content. In specific implementation, the user demand information and the user portrait are used to retrieve the road book information online, and then the road book information is filled into a preset road book template (i.e., the above-mentioned road book template) to obtain the first road book.

[0122] When filling in data, you can use the above-mentioned intent recognition and / or entity recognition method to perform keyword extraction operations from the road book information obtained from the online search, and match the extracted keywords with the labels in the road book template (i.e. the above-mentioned starting point, destination, two attractions to visit every day, accommodation, etc.), so as to fill the matched content into the corresponding labels and obtain the first road book.

[0123] As can be seen from the above, when implementing the generation of the first road book, the present invention can adopt the above-mentioned template method (filling content based on a preset template), search-based generation (retrieval of relevant information from a database and recombination) or complete generation (free generation of content by the LLM model). In specific implementation, the appropriate generation strategy can be selected for content generation according to user needs and input information. If one of the methods cannot generate a road book, the other method can be switched to generate the road book.

[0124] S13: Acquire a road book adjustment parameter, and adjust the first road book based on the road book adjustment parameter to obtain a second road book.

[0125] In actual applications, although the first route book is generated, there may be a situation where the first route book does not meet the user's needs. For example, the user likes to climb mountains, but the attractions recommended in the first route book are mostly water play. In this case, the recommended first route book is not suitable for the user. At this time, the content of the route book can be adjusted according to the user's behavioral preference information.

[0126] In addition, after the first route book is generated, there may be bad weather, traffic congestion, etc. At this time, playing according to the content in the first route book will cause users to encounter extreme weather or be stuck in traffic for a long time, which will lead to a bad user experience. Therefore, the content of the road book can be adjusted according to weather information or traffic information.

[0127] In addition, after the first route book is generated, it may not match the user's needs. For example, the user likes a more casual itinerary, but the itinerary arranged in the first route book is tight, which does not match the user's needs. Or the user likes northern cuisine, but the recommended restaurants are all southern cuisine restaurants, which does not match the user's needs. In this case, the user can realize user configuration by operating the slider or option, and then adjust the route book according to the user configuration information.

[0128] At this time, the road book adjustment parameters may include at least one or more of the target user's behavior preference information, weather information or traffic information, and user configuration information obtained by the user operating the road book configuration options or the road book configuration slider.

[0129] After adjusting the first route book according to the route book adjustment parameters, the second route book can be obtained.

[0130] S14: performing content optimization and sorting operations on the second route book to obtain a target route book.

[0131] After obtaining the second book, in order to improve the quality of the second book, the second book can be optimized to obtain the third book. The content optimization operation can be subdivided into a grammar checking optimization operation and a semantic understanding optimization operation.

[0132] The grammar checking optimization operation is to check whether there are grammatical errors in the second book, and if so, correct the grammatical errors to make the content of the second book smooth and semantically coherent.

[0133] Semantic understanding optimization operation refers to the process of interpreting and understanding the meaning expressed by language symbols (including words, phrases, sentences, etc.). After understanding the semantics in the second route book, the content of the sentence can be adjusted to make it highly accurate, smooth and fluent, making the sentence more in line with user habits. In addition, the beauty of the sentence can be adjusted to improve the quality of the route book.

[0134] In addition, you can also select other optimization operations based on actual needs to optimize the content and ensure the accuracy and fluency of the content.

[0135] After the content optimization is completed, the itineraries in the third route book may be sorted according to a preset sorting rule to obtain a target route book.

[0136] The preset sorting rules include at least one or more of ratings, popularity, user behavior preference information, and user tags.

[0137] Among them, the ratings and popularity can be crawled from the web page, and the user behavior preference information and user tags can be obtained by user input.

[0138] Sorting algorithms (such as PageRank, TF-IDF (term frequency-inverse document frequency), etc.) are used to sort the recommended items such as attractions and restaurants in the generated third-party book according to the above preset sorting rules, and give priority to displaying content that users may be more interested in or content that the public thinks is good, thereby improving user experience.

[0139] In this embodiment, when generating a road book, the corresponding road book generation strategy will be determined based on the attribute information of the user input data, so that the selected road book generation strategy matches the user's needs, thereby improving the accuracy and reliability of subsequent road book generation. In addition, after the first road book is generated based on the user's demand information and the user's portrait, the first road book will be adjusted based on the road book adjustment parameters to obtain a second road book, so that the road book can be adaptively adjusted according to user preferences, weather, traffic and other contents, thereby improving the reliability and personalization of the generated road book. After obtaining the second road book, the second road book can be optimized and sorted to ensure the content quality of the road book and the user experience. That is, the present invention improves the user experience on the basis of intelligently generating the road book.

[0140] In addition, the embodiment of the present invention uses the generation capability of LLM to make the road book more abundant and detailed, meeting the diverse travel needs of users. With the continuous training and optimization of the LLM model and the real-time update of basic data, the road book content can be kept up to date.

[0141] In addition, by automatically generating road books, the time cost of manual writing is reduced and the overall efficiency is improved.

[0142] In addition, the embodiment of the present invention generates a road book according to the personalized needs of the user, thereby improving the user's satisfaction and sense of belonging.

[0143] The above embodiment mentioned that "based on the road book adjustment parameters, the first road book is adjusted to obtain the second road book". In the specific implementation, there are multiple implementation methods, and the specific implementation is now introduced.

[0144] 1. The road book adjustment parameters include the behavior preference information of the target user.

[0145] At this time, refer to Figure 2 , based on the road book adjustment parameter, adjusting the first road book to obtain a second road book, including:

[0146] S21. Acquire behavior preference information of the target user based on the user interaction operation with the target user.

[0147] In actual applications, pushing a road book to a user can be achieved through user interaction. For example, a user says in a chat interface: "Recommend me a three-day itinerary for traveling in Beijing." Then, a road book is generated using the above method and recommended to the user. After the user opens it and finds that the itinerary is not suitable, for example, if he likes mountain climbing, the recommended attractions are water attractions. Then, the user says in a voice voice "Please recommend a mountain climbing itinerary." This is the user's behavioral preference information.

[0148] In addition to inputting user behavior preference information via voice, users can also input it in text form.

[0149] It should be noted that when users enter preference information, they can enter it multiple times, such as "Please recommend a hiking itinerary" for the first time, "Please delete the second attraction" for the second time, and "Change the accommodation location to a five-star hotel" for the third time...

[0150] S22: Adjust the contents in the first path book that are not related to the behavior preference information to obtain a second path book.

[0151] Specifically, an intent recognition operation or an entity recognition operation can be used to filter out content in the first book that is not related to the behavior preference information, and then adjust the content to the user's preferred content, such as adjusting water play to mountain climbing, and adjusting accommodation to five-star accommodation, thereby obtaining a second book.

[0152] It should be noted that after obtaining the second route book, the user may still be dissatisfied with some aspects. At this time, the user can continue to input preference information, and then adjust the second route book again based on user needs. After multiple continuous adjustments, the final route book is the route book that meets user needs.

[0153] In actual applications, during the interaction with the user, user preference information will be recorded in real time, so that the user preference can be directly considered when the first route book is generated later. The generated first route book will be able to meet the user's needs with a high probability, and the adjustment content of the subsequent first route book will be less. After continuous optimization, the route book generated for the user later can meet the user's personalized needs and recommend the user's preferred itinerary.

[0154] It should be noted that user behavior preference information can be displayed through user portraits. By using user profile technology, user portraits can be built based on users' historical behavior, preferences and other information, so that personalized elements can be incorporated into the content when the road book is adjusted.

[0155] 2. The road book adjustment parameters include weather information or traffic information.

[0156] Reference Figure 3 , based on the road book adjustment parameter, adjusting the first road book to obtain a second road book, including:

[0157] S31, determining weather information and traffic information of each location in the first route book.

[0158] In this embodiment, during the user's travel, the weather and traffic at the tourist spots change in real time. Therefore, the initially generated road book may contain inappropriate parts, and it will be adjusted later based on user needs, traffic, weather and other information.

[0159] Weather information can be obtained through weather forecasts, and traffic information can be obtained based on maps.

[0160] S32: If there is a target location among the locations whose weather information is preset bad weather or whose traffic information is preset traffic congestion, determine that the itinerary of the target location corresponds to a replacement itinerary.

[0161] Specifically, if the weather information of a certain location is not good, such as preset bad weather, the preset bad weather may be heavy rain, heavy snow, heavy wind, sandstorm, etc.

[0162] If the weather is bad, it will affect the trip. For example, if there is heavy rain while hiking, it will make it impossible to hike. At this time, the itinerary needs to be adjusted.

[0163] In addition, with regard to traffic congestion, if there is traffic congestion at the location, users may waste most of their time on the road, and have little time to visit the attractions, and the user experience is not very good.

[0164] Therefore, the itinerary needs to be adjusted for target locations with bad weather or traffic congestion.

[0165] When adjusting the itinerary, priority is given to places that are close to the location and have similar attractions. For example, if a user is hiking in a certain district in Beijing and the weather in that district is bad, we can recommend another district in Beijing where the weather is better and there is no congestion for hiking. The recommended other attraction and the content of the tour at that attraction are the replacement itinerary.

[0166] S33: Update the itinerary of the target location in the first route book to a replacement itinerary to obtain a second route book.

[0167] Specifically, the replacement itinerary is directly used to replace the itinerary of the target location described in the first itinerary book to obtain the second itinerary book.

[0168] In another implementation of the present invention, the above operation is to directly adjust the road book when the weather is bad or the traffic is congested. In addition, when it is determined that the weather is bad or the traffic is congested, a prompt message can be output to remind the user that this trip may not be suitable for travel, and the user decides whether to adjust the road book content. The prompt message can be such as "The weather at XX scenic spot in the road book is bad, do you need to adjust the road book content?"

[0169] If a first modification instruction that does not include a designated itinerary is received, such as "confirm adjustment", steps S32 and S33 may be executed to automatically perform an adjustment operation on the road book content.

[0170] If a second modification instruction including a specified itinerary is received, such as the user's voice output "adjust XX attraction to YY attraction", then the itinerary of the target location described in the first route book is directly updated to the specified itinerary in the second modification instruction according to the user's needs, that is, the XX attraction is adjusted to the YY attraction, and the second route book is obtained.

[0171] It should be noted that, when the road book is initially created, user portraits and weather information for the next few days can be collected. When the road book is generated, all this information that can be considered can be input into the model to obtain a road book that meets user preferences and weather information. When users subsequently travel in actual situations, when there is traffic congestion or bad weather, they can use the large model to adaptively adjust the road book based on traffic and weather information, thereby realizing the function of continuously and dynamically adjusting the road book based on weather and traffic information in real time, thereby improving the feasibility of the road book and meeting the user's real-time adjustment needs.

[0172] 3. The road book adjustment parameters include: user configuration information obtained by the user operating the road book configuration options or the road book configuration slider.

[0173] In the chat interface, configurable options or sliders are provided to allow users to adjust content details according to their needs, such as adjusting the number of attractions, types of dining, etc.

[0174] At this time, based on the road book adjustment parameter, the first road book is adjusted to obtain a second road book, including:

[0175] According to the user configuration information, the content of the first path book is adaptively adjusted to obtain a second path book.

[0176] The adaptive adjustment includes at least one of content addition, content deletion and content modification.

[0177] If the user selects two attractions through the slider, the number of attractions in the first route book can be reduced or increased to two. The same is true when the user adjusts the dining type.

[0178] In addition, if the user chooses not to stay overnight, the accommodation part in the first trip book can be deleted.

[0179] Alternatively, if the user configures the itinerary to include XX attraction through options, the user can add the itinerary of XX attraction from the first route book.

[0180] In actual applications, the options can be implemented in various ways, such as through a drop-down menu or a user-defined method. Similarly, the implementation method of the slider can also be set accordingly.

[0181] In this embodiment, after the road book is initially generated, the road book can be adaptively adjusted based on the user's behavior preferences to improve the matching degree of the road book content to the user, meet the user's personalized needs, and improve the user experience.

[0182] In another implementation of the present invention, an interactive interface or application program may be designed to allow the user to annotate, modify or add custom content while reading the road book.

[0183] When customizing and adding content, in addition to conventional text, the road book can also support multimodal data such as pictures and videos. For example, pictures or short video introductions of scenic spots can be embedded in the road book to enhance the user's reading experience. Another example is that during the travel process, the recorded human-computer dialogue content is filled into the road book, or pictures or videos of the beautiful scenery ahead are filled into the road book to achieve the richness of the road book content and make the road book have multimodal data. That is, this embodiment improves the richness of the road book content by adding multimodal data such as pictures, videos, and voices to the road book.

[0184] Based on the above-mentioned embodiment of the road book processing method, another embodiment of the present invention provides a road book processing device, referring to Figure 4 , which may include:

[0185] The data determination module 11 is used to determine the user demand information corresponding to the user input data of the target user, and to obtain the user portrait of the target user;

[0186] A strategy determination module 12, used to determine a road book generation strategy corresponding to the attribute information of the user input data;

[0187] A route book generating module 13, configured to generate a first route book corresponding to the user demand information and the user portrait using the route book generating strategy;

[0188] A road book adjustment module 14 is used to obtain a road book adjustment parameter, so as to adjust the first road book based on the road book adjustment parameter to obtain a second road book;

[0189] The road book optimization module 15 is used to perform content optimization and sorting operations on the second road book to obtain a target road book.

[0190] In one implementation, the information determination module 11 includes:

[0191] The data acquisition submodule is used to obtain user input data;

[0192] A first processing submodule is used for, if the user input data is voice, performing a voice recognition operation on the voice to obtain a voice recognition result, and performing an intent recognition and / or entity recognition operation on the voice recognition result to obtain user demand information;

[0193] The second processing submodule is used to perform intent recognition and / or entity recognition operations on the text to obtain user demand information if the user input data is text.

[0194] In one implementation, the policy determination module 12 includes:

[0195] A format determination submodule, used to determine whether the format of the user input data is a preset template format;

[0196] The first strategy determination submodule is used to use the template method as a road book generation strategy if yes;

[0197] The integrity calculation submodule is used to calculate the data integrity of the user input data if no;

[0198] The second strategy determination submodule is used to use the search formula generation method as the road book generation strategy if the data completeness is greater than a preset completeness threshold; if the data completeness is not greater than the preset completeness threshold, use the model generation method as the road book generation strategy.

[0199] In one implementation, the road book generating module 13 includes:

[0200] A first generation submodule is used for converting the user demand information and the user portrait into a vector if the road book generation strategy is a model generation method, and calling a road book generation model to process the vector to obtain a first road book;

[0201] A second generation submodule is used for retrieving a first road book from a pre-configured road book database using the user demand information and the user portrait if the road book generation strategy is a search-based generation method;

[0202] The third generation submodule is used to, if the road book generation strategy is the template method, use the user demand information and the user portrait to retrieve the road book information online, fill the road book information into the preset road book template, and obtain the first road book.

[0203] In one implementation, the road book adjustment parameter includes the behavior preference information of the target user, and the road book adjustment module 14 includes:

[0204] The first adjustment submodule is used to obtain the behavior preference information of the target user based on the user interaction operation with the target user, and adjust the content in the first path book that is not related to the behavior preference information to obtain the second path book.

[0205] In one implementation, the road book adjustment parameter includes weather information or traffic information; the road book adjustment module 14 includes:

[0206] A data determination submodule, used to determine weather information and traffic information of each location in the first route book;

[0207] A prompt information output submodule, for outputting prompt information if there is a target location in each location where the weather information is preset bad weather or the traffic information is preset congested traffic;

[0208] The second adjustment submodule is used for, if a first modification instruction that does not include a designated itinerary is received, determining that the itinerary of the target location corresponds to a replacement itinerary, updating the itinerary of the target location in the first route book to the replacement itinerary, and obtaining a second route book.

[0209] The third adjustment submodule is used to update the itinerary of the target location in the first route book to the specified itinerary in the second modification instruction to obtain a second route book if a second modification instruction including a specified itinerary is received.

[0210] In one implementation, the road book adjustment parameters include: user configuration information obtained by the user operating the road book configuration option or the road book configuration slider; the road book adjustment module 14 also includes:

[0211] The fourth adjustment submodule is used to adaptively adjust the content in the first path book according to the user configuration information to obtain the second path book; the adaptive adjustment includes at least one of content addition, content deletion and content modification.

[0212] In one implementation, the road book optimization module 15 includes:

[0213] A content optimization submodule, configured to perform content optimization operations on the content of the second-path book to obtain a third-path book; the content optimization operations at least include grammar checking optimization operations and semantic understanding optimization operations;

[0214] The sorting submodule is used to sort the itineraries in the third route book according to preset sorting rules to obtain a target route book; the preset sorting rules include at least one or more of ratings, popularity, user behavior preference information, and user tags.

[0215] In this embodiment, when generating a road book, the corresponding road book generation strategy will be determined based on the attribute information of the user input data, so that the selected road book generation strategy matches the user's needs, thereby improving the accuracy and reliability of subsequent road book generation. In addition, after the first road book is generated based on the user's demand information and the user's portrait, the first road book will be adjusted based on the road book adjustment parameters to obtain a second road book, so that the road book can be adaptively adjusted according to user preferences, weather, traffic and other contents, thereby improving the reliability and personalization of the generated road book. After obtaining the second road book, the second road book can be optimized and sorted to ensure the content quality of the road book and the user experience. That is, the present invention improves the user experience on the basis of intelligently generating the road book.

[0216] It should be noted that, for the working process of each module and sub-module in this embodiment, please refer to the corresponding description in the above embodiment, which will not be repeated here.

[0217] The present application also provides an electronic device in an embodiment. Figure 5 As shown, it shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiment of the present application. The electronic device in the embodiment of the present application may include but is not limited to fixed terminals such as vehicle controllers, mobile phones, notebook computers, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, etc. Figure 5 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0218] like Figure 5 As shown, the electronic device may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 602 or the program loaded from the storage device 608 to the random access memory (RAM) 603 to implement the above-mentioned road book processing method. When the electronic device is powered on, the RAM 603 also stores various programs and data required for the operation of the electronic device. The processing device 601, the ROM 602 and the RAM 603 are connected to each other through the bus 604. The input / output (I / O) interface 605 is also connected to the bus 604.

[0219] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a memory card, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5 An electronic device having various devices is shown, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0220] An embodiment of the present application also provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements any road book processing method provided in the embodiment of the present application.

[0221] A computer-readable storage medium is also provided in an embodiment of the present application. The storage medium carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement any road book processing method provided in the embodiment of the present application.

[0222] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A road book processing method, characterized in that: include: Determine user demand information corresponding to user input data of a target user, and obtain a user profile of the target user; Determine a road book generation strategy corresponding to the attribute information of the user input data; Using the route book generation strategy, generate a first route book corresponding to the user demand information and the user portrait; Acquiring a road book adjustment parameter, so as to adjust the first road book based on the road book adjustment parameter to obtain a second road book; The second route book is subjected to content optimization and sorting operations to obtain a target route book.

2. The road book processing method according to claim 1, characterized in that: Determine the user demand information corresponding to the user input data of the target user, including: Get user input data; If the user input data is voice, perform a voice recognition operation on the voice to obtain a voice recognition result, and perform an intent recognition and / or entity recognition operation on the voice recognition result to obtain user demand information; If the user input data is text, intent recognition and / or entity recognition operations are performed on the text to obtain user demand information.

3. The road book processing method according to claim 1, characterized in that: The attribute information of the user input data includes format and data integrity; Determining a road book generation strategy corresponding to the attribute information of the user input data includes: Determining whether the format of the user input data is a preset template format; If so, use the template method as the road book generation strategy; If not, calculating the data integrity of the user input data; If the data completeness is greater than a preset completeness threshold, a search-based generation method is used as a road book generation strategy; If the data integrity is not greater than a preset integrity threshold, the model generation method is used as a road book generation strategy.

4. The road book processing method according to claim 1, characterized in that: Using the road book generation strategy to generate a first road book corresponding to the user demand information and the user portrait includes: If the road book generation strategy is a model generation method, the user demand information and the user portrait are converted into a vector, and a road book generation model is called to process the vector to obtain a first road book; If the road book generation strategy is a search-based generation method, the first road book is retrieved from a pre-configured road book database using the user demand information and the user portrait; If the road book generation strategy is a template method, the user demand information and the user portrait are used to retrieve road book information online, and the road book information is filled into a preset road book template to obtain a first road book.

5. The road book processing method according to claim 1, characterized in that: The road book adjustment parameter includes the behavior preference information of the target user, and the first road book is adjusted based on the road book adjustment parameter to obtain a second road book, including: Based on the user interaction operation with the target user, obtaining the behavior preference information of the target user; The content in the first path book that is not related to the behavior preference information is adjusted to obtain a second path book.

6. The road book processing method according to claim 1, characterized in that: The road book adjustment parameters include weather information or traffic information; The first road book is adjusted based on the road book adjustment parameter to obtain a second road book, including: Determining weather information and traffic information for each location in the first route book; If there is a target location among the locations where the weather information indicates that the weather is bad or the traffic information indicates that the traffic is congested, outputting a prompt message; If a first modification instruction that does not include a designated itinerary is received, determining that the itinerary of the target location corresponds to a replacement itinerary, updating the itinerary of the target location in the first route book to the replacement itinerary, and obtaining a second route book; If a second modification instruction including a designated itinerary is received, the itinerary of the target location in the first route book is updated to the designated itinerary in the second modification instruction to obtain a second route book.

7. The road book processing method according to claim 1, characterized in that: The road book adjustment parameters include: user configuration information obtained by the user operating the road book configuration options or the road book configuration slider; adjusting the first road book based on the road book adjustment parameters to obtain the second road book, including: According to the user configuration information, the content in the first path book is adaptively adjusted to obtain the second path book; the adaptive adjustment includes at least one of content addition, content deletion and content modification.

8. The road book processing method according to claim 1, characterized in that: The second route book is optimized and sorted to obtain a target route book, including: Performing content optimization operation on the content of the second-path book to obtain a third-path book; the content optimization operation at least includes a grammar checking optimization operation and a semantic understanding optimization operation; According to preset sorting rules, each itinerary in the third route book is sorted to obtain a target route book; the preset sorting rules include at least one or more of ratings, popularity, user behavior preference information, and user tags.

9. A road book processing device, characterized in that: include: A data determination module, used to determine user demand information corresponding to user input data of a target user, and to obtain a user portrait of the target user; A strategy determination module, used to determine a road book generation strategy corresponding to the attribute information of the user input data; A route book generating module, configured to generate a first route book corresponding to the user demand information and the user portrait using the route book generating strategy; A road book adjustment module, used for obtaining a road book adjustment parameter, and adjusting the first road book based on the road book adjustment parameter to obtain a second road book; The road book optimization module is used to perform content optimization and sorting operations on the second road book to obtain a target road book.

10. An electronic device, characterized in that: The method comprises at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program so that the electronic device can implement the road book processing method as described in any one of claims 1 to 8.

Images