Information providing method, device and equipment

Abstract

The embodiment of the invention provides an information providing method, device and equipment. According to the scheme, the method comprises the steps of obtaining first information which is published on an external content platform and used for describing an offline service provided at a target offline merchant, then carrying out structured processing on the first information to generate second information containing time information, place information and service introduction information of the offline service, and then sending the second information to the external content platform, and based on at least part of contents in the first information and the second information, performing association matching on the target offline merchant and interest points in an interest point database, determining a target interest point corresponding to the target offline merchant, and then providing the second information, the target interest point is displayed in a user interface corresponding to the target interest point.

Description

TECHNICAL FIELD

[0001] One or more embodiments of the present specification relate to the technical field of data processing and artificial intelligence, in particular to an information providing method. One or more embodiments of the present specification also relate to an information providing device and a computing device. BACKGROUND

[0002] With the development of Internet technology, a large number of online platforms have emerged in the local life field to provide users with surrounding merchant information, preferential activities, new product information, and other content. Such local information service platforms aim to connect consumers (C-end) and offline entities (B-end) to improve the efficiency of users discovering and experiencing local services.

[0003] In the current Internet local service field, the prosperity of online platforms highly depends on the active entry and content maintenance of offline merchants (B-end), forming a traditional mode of "first B-end, then C-end". This mode requires merchants to manually upload and continuously update their store information, product information, activity information, and other information. The long chain and high cost of the startup process of such online service platforms result in a severe "cold start" dilemma for emerging platforms or new vertical businesses in the early stage, i.e., lack of content, difficulty in attracting users, and serious constraints on the rapid startup and scale development of the business, making it difficult to quickly and cost-effectively meet users' information acquisition needs of "when, where, and what activities".

[0004] Therefore, there is a need to solve the problem of cold start difficulty of online local information service platforms. SUMMARY

[0005] In view of the above, one or more embodiments of the present specification provide an information providing method and device to solve the problem of cold start difficulty of online local information service platforms.

[0006] According to a first aspect of one or more embodiments of the present specification, an information providing method is provided, comprising:

[0007] obtaining first information published on an external content platform; the first information is used to describe offline services provided at a target offline merchant;

[0008] performing structured processing on the first information to generate second information; the second information contains time information, location information, and service introduction information of the offline services;

[0009] based on at least part of the content in the first information and the second information, associating and matching the target offline merchant with a point of interest in a point of interest database to determine a target point of interest corresponding to the target offline merchant;

[0010] The second information is provided for display in the user interface corresponding to the target point of interest.

[0011] According to a second aspect of one or more embodiments of this specification, an information providing apparatus is provided, comprising:

[0012] The first information acquisition module is used to acquire first information published on external content platforms; the first information is used to describe the offline services provided by the target offline merchants.

[0013] The second information generation module is used to perform structured processing on the first information to generate the second information; the second information includes the time information, location information, and service description information of the offline service.

[0014] The point of interest determination module is used to associate and match the target offline merchant with the points of interest in the point of interest database based on at least part of the first information and the second information, and determine the target point of interest corresponding to the target offline merchant;

[0015] The second information providing module is used to provide the second information for display in the user interface corresponding to the target point of interest.

[0016] According to a third aspect of one or more embodiments of this specification, a computing device is provided, including a memory, a processor, and computer instructions stored in the memory and executable on the processor, wherein the processor, when executing the computer instructions, implements the steps of the information providing method.

[0017] One embodiment of this specification can achieve at least the following beneficial effects: by acquiring first information published by a target offline merchant on an external content platform to describe the offline services provided by the target offline merchant, and then performing structured processing on the first information to generate second information containing time information, location information, and service description information of the offline services, then, based on at least part of the content in the first and second information, associating and matching the target offline merchant with points of interest in a point of interest database to determine the target point of interest corresponding to the target offline merchant, and finally providing the second information to be displayed in the user interface corresponding to the target point of interest, thereby realizing the automated cold start of local information service content by automatically acquiring and intelligently processing information from an external content platform and associating it with physical points of interest. On the one hand, it significantly reduces the platform's dependence on merchants' active onboarding; on the other hand, the cold start process of local information services reduces the operational burden on merchants; and furthermore, it can provide users with timely and efficient one-stop domain information, providing convenience for domain information audiences. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A schematic diagram illustrating an application scenario of an information provision method provided in an embodiment of this specification;

[0020] Figure 2 A flowchart illustrating an information provision method provided in an embodiment of this specification;

[0021] Figure 3 This is a flowchart illustrating an information provision method in a practical application scenario provided by an embodiment of this specification.

[0022] Figure 4 A schematic diagram of an information details page provided in an embodiment of this specification;

[0023] Figure 5 The embodiments provided in this specification correspond to Figure 2 A schematic diagram of the structure of an information providing device;

[0024] Figure 6 This is a structural block diagram of a computing device provided as an embodiment of this specification. Detailed Implementation

[0025] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0026] This specification uses specific terms to describe embodiments thereof. Terms such as "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described herein, as well as the features of those different embodiments or examples, without contradiction.

[0027] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “an,” “an,” “the,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification includes any or all possible combinations of one or more associated listed items.

[0028] The terms “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in the process, method, product, or apparatus that includes said elements is not excluded.

[0029] Although the terms "first," "second," etc., may be used to describe various information in one or more embodiments of this specification, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, "first" may also be referred to as "second," and similarly, "second" may also be referred to as "first," without departing from the scope of one or more embodiments of this specification. Ordinal numbers such as "first," "second," etc., do not necessarily indicate order; often they are used to facilitate the distinction of objects. For example, "first server" and "second server" usually refer to two servers. To distinguish these two servers, they are described as "first server" and "second server." Of course, sometimes these two servers may be the same server.

[0030] Depending on the context, the word "if" as used here can be interpreted as "when," "when," or "in response to determination."

[0031] In this specification, unless explicitly stated otherwise, "receiving and sending data" does not necessarily mean direct receiving and sending; it can also mean indirect receiving and sending. For example, A receiving data sent by B can be understood as A directly receiving the data sent by B, or it can be understood as A indirectly receiving the data sent by B through other entities such as C. Similarly, B sending data to A can be understood as B sending the data directly to A, or it can be understood as B indirectly sending the data to A through other entities such as C. Here, C can be one entity, or it can be two or more entities.

[0032] In this specification, unless explicitly stated otherwise, the relationships between structures can be direct or indirect. For example, when describing "A is connected to B," unless it is explicitly stated that A and B are directly connected, it should be understood that A can be directly connected to B or indirectly connected to B. Similarly, when describing "A is on top of B," unless it is explicitly stated that A is directly above B (AB is adjacent and A is above B), it should be understood that A can be directly above B or indirectly above B (AB is separated by other elements, and A is above B). And so on.

[0033] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties. The collection, use and processing of related data shall comply with the relevant laws, regulations and standards of the relevant regions, and corresponding operation entry points shall be provided for users to choose to authorize or refuse.

[0034] The following explains the terms and concepts used in one or more embodiments of this specification.

[0035] Vertical domains: These represent fields that have developed unique language patterns. While the distinctive language of these domains may seem like "encrypted communication" to outsiders, it serves a crucial function within their communities. This specific language acts as a code for insiders to identify and confirm fellow enthusiasts, strengthening community identity and cohesion, while also implicitly establishing boundaries between different groups. Within a vertical domain, each technical term or jargon typically refers to a highly precise and complex concept within that domain, enabling insiders to quickly understand and efficiently communicate complex information.

[0036] An example of a vertical field can include the ACG field. ACG is relative to the real world (referred to as the "three-dimensional world"), with the three-dimensional world referring to the real world and the ACG world referring to a two-dimensional fictional world. ACG terms refer to some specialized terms that often appear in ACGN (Animation, Comic, Game, Novel) works, and are commonly used to describe fictional worlds, character attributes, and interaction behaviors, helping users integrate into the ACG community and understand the connotations of works. The language features of the ACG field are as follows: a large number of terms are derived from Japanese translations, specific memes in works, transliterations, abbreviations, and exaggerated emotional expressions. Typical ACG terms can include tsundere, kichiku, high energy ahead, FFF group,三无少女 (mute, cool, and stoic girl), zannen, and collecting anime merchandise. Among them, the ACG term "otaku who collect anime merchandise" is used to represent enthusiasts who collect anime peripheral products (such as badges). In the ACG field, there are offline merchants such as "goods stores" (peripheral specialty stores), ACG-themed shopping malls, IP-themed pop-up stores, anime carnivals, IP peripheral retail stores, themed restaurants, and immersive entertainment experience stores.

[0037] Another example of a vertical field can include the game / e-sports field. The language features of the game / e-sports field are as follows: emphasis on tactical actions, character / equipment attributes, version balance, many English abbreviations, or visualization expressions. Typical term examples in the game / e-sports field include movement, DPS, tank, nerf, buff, AOE, cooldown, ward,吃鸡 (PUBG), and speedrun. For example, "nerf" specifically refers to the official weakening of a certain attribute. In the game / e-sports field, there are offline merchants such as e-sports-themed hotels, offline viewing events, game-themed blocks, and game IP offline experience stores.

[0038] Another example of a vertical field can include board games / tabletop role-playing games. The language features of the board games / tabletop role-playing games field are as follows: integrating game mechanism terms, character status descriptions, and specific rule abbreviations to create an immersive co-imaginative space. Typical term examples in the board games / tabletop role-playing games field include DKP, rolling dice, SAN value, character creation, game master (KP), expansion and starting a game, and dice luck. In the board games / tabletop role-playing games field, there are offline merchants such as themed board game stores, murder mystery game halls, and card battle centers.

[0039] Large Language Models (LLMs) are deep learning models trained on massive amounts of text data, enabling them to generate natural language text or understand the meaning of language text. LLMs can provide in-depth knowledge and language production on a wide range of topics through training on large datasets. Through large-scale unsupervised training, LLMs learn patterns and structures of natural language, mimicking human language cognition and generation processes to some extent. LLMs employ a similar Transformer architecture and pre-training objectives as smaller models, with the main differences being increased model size, training data, and computational resources. Compared to traditional Natural Language Processing (NLP) models, LLMs better understand and generate natural text, while also exhibiting some logical thinking and reasoning abilities. LLMs possess in-context learning capabilities, enabling them to learn complex patterns in language and perform a wide range of tasks, including text summarization, translation, sentiment analysis, multi-turn dialogue, and more. For example, LLM can include the GPT series, T5 (Text to Text Transfer Transformer) model, PaLM model, BERT, LLaMA (Large Language Model MetaAI) model, Tongyi Qianwen model, Bailing model, etc.

[0040] In the current local information service ecosystem, the launch and prosperity of online service platforms highly depend on the proactive onboarding and cooperation of offline merchants. The standard business process in the traditional model typically involves the platform first guiding merchants to complete onboarding and signing agreements. Merchants then manually upload and continuously maintain their store information, products, services, promotional activities, and other content. This process can be summarized as a "B first, then C" model, meaning that a sufficient number of high-quality merchant resources must be accumulated before attracting and serving consumer users.

[0041] However, this traditional model has significant pain points—the startup process for online service platforms is lengthy and difficult to cold start. For an emerging platform or a new business segment that wants to expand, communicating with, signing contracts with, and guiding a large number of merchants to upload content from scratch is a time-consuming, labor-intensive, and costly process. This results in a lack of content in the early stages of the platform, which severely restricts the rapid launch and scaling of the service platform's business.

[0042] Furthermore, the traditional online service platform launch and operation model places the entire burden of content production on the merchants. However, many small and medium-sized merchants, especially in specific vertical sectors (such as specialty retail and cultural consumption), may lack the willingness or ability to operate online, resulting in untimely information updates and inconsistent content quality. This makes it difficult for the platform to guarantee the accuracy and richness of information, thus affecting the user experience.

[0043] Furthermore, traditional local information services suffer from information silos and aggregation challenges. Specifically, in the local information service sector, a large amount of valuable information is actually scattered across numerous third-party social platforms, content communities, and social groups. This information exists in unstructured forms (such as social media posts and community announcements). Traditional online service platform startup and operation models cannot effectively understand and utilize this scattered "intelligence," resulting in a waste of information resources and making it difficult for consumers to obtain comprehensive and timely local information services within a single platform. This is especially true in scenarios involving specific vertical fields, where information fragmentation is even more severe.

[0044] Specifically, in the context of local information services, consumers need to conveniently discover "when, where, and what activities or services are available." Due to the aforementioned pain points, traditional business models struggle to meet this need quickly and cost-effectively, especially in vertical sectors with frequent information updates and distinctive activities, where their inadequacy is even more pronounced.

[0045] Therefore, there is an urgent need for a technical solution that can break through the limitations of the traditional "B-end first" model, enabling the automatic generation of high-quality local service content without the active participation of merchants, and accurately matching it with offline entities. This would effectively reduce the cold start threshold of the platform, solve key problems such as the difficulty of platform cold start, poor information aggregation, and low degree of automation in content production, and accelerate the construction of the local service ecosystem.

[0046] Based on the solutions implemented in this specification, firstly, original information from multiple sources (such as text and image posts on social media) is actively acquired. Then, the unstructured original information is automatically converted into standardized, displayable content cards and accurately associated with offline Points of Interest (POIs). Thus, local information services can be provided to users conveniently and efficiently without the need for merchants to actively participate.

[0047] The technical solutions provided in the various embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0048] Figure 1 This is a schematic diagram illustrating an application scenario of an information provision method provided in an embodiment of this specification.

[0049] like Figure 1As shown, offline merchants can publish first information on external content platform 101; server 102 can obtain the first information published by offline merchants from external content platform 101, and then perform structured processing on the first information to obtain second information, and based on at least part of the content in the first information and the second information, associate and match the target offline merchant with the points of interest in the point of interest database to determine the target point of interest corresponding to the target offline merchant. After that, server 102 can provide the second information to user terminal 103 so that the second information can be displayed in the user interface corresponding to the target point of interest on user terminal 103.

[0050] In this context, "external" in "external content platform 101" refers to content distribution channels that are not controlled by server 102, in contrast to server 102. Although in Figure 1 Although not shown in the diagram, it is understood that offline merchants can publish initial information through applications on their own terminals. This initial information can then be sent via the server corresponding to the application to other terminals that have that application or a corresponding information receiving application installed. From a hardware perspective, Figure 1 The external content platform 101 may include a terminal for receiving the first information, and may also include a server for the application used by the merchant to publish the first information. From a software perspective, Figure 1 The external content platform 101 may include applications used by merchants to publish first information and applications used to receive first information.

[0051] In such Figure 1 In the application scenario shown, server 102 can connect to one or more external content platforms 101 and link to one or more user terminals 103 via LAN connection, WAN connection, Internet connection or other types of data network. Figure 1 The server 102 may include, but is not limited to, any device, equipment, platform, or equipment cluster with computing and processing capabilities. Figure 1 The user terminal 103 may include, but is not limited to, smartphones, tablets, laptops, PDAs, personal computers, smart home devices, and in-vehicle devices.

[0052] although Figure 1 The information provision method shown can be executed on server 102. However, in practical applications, when the user terminal 103 meets the operating conditions of the information provision method, at least some steps of the information provision method in this application embodiment can be performed on user terminal 103.

[0053] This application provides an information providing method and also relates to an information providing device and a computing device, which will be described in detail in the following embodiments.

[0054] Figure 2 This is a flowchart illustrating an information provision method provided in an embodiment of this specification.

[0055] From a programming perspective, the entity executing the process can be a program hosted on an application server or application terminal. It can be understood that this method can be executed by any device, equipment, platform, or cluster of devices with computing and processing capabilities.

[0056] like Figure 2 As shown, the process may include the following steps:

[0057] Step 202: Obtain the first information published on the external content platform; the first information is used to describe the offline services provided at the target offline merchant.

[0058] Information, in this context, refers to information that can be accessed by users and generates value within a certain timeframe. Information typically exhibits timeliness and regionality.

[0059] The information provision method can be applied to a first device. In practical applications, the first device may include a server or a user terminal. The external content platform may be a content publishing channel that is not controlled by the first device. In practical applications, the external content platform may include, but is not limited to, social media, content communities, and merchant private domain communities. The first information may be independently published by the target offline merchant or its affiliates on the external content platform. The first device can obtain first information related to the target offline merchant from at least one external content platform.

[0060] In one or more embodiments of this specification, obtaining the first information published on an external content platform may specifically include: collecting the first information published by the content publishing account of the target offline merchant from the external content platform; and / or, collecting the first information matching the specified vertical domain keywords from the external content platform; the vertical domain keywords include intellectual property keywords.

[0061] Optionally, based on known offline merchant entities, their content publishing accounts on the external content platform can be identified, and the content published by those content publishing accounts can be collected as the first information.

[0062] Optionally, based on a pre-built vertical domain knowledge base (such as an intellectual property object knowledge base), keywords related to a specified vertical domain (such as intellectual property keywords) can be extracted, and relevant information can be collected on the external content platform based on the vertical domain keywords as the first information.

[0063] In one or more embodiments of this specification, the first information may specifically include information from a vertical field, such as the anime / manga / game (ACG) field. For the ACG field, the intellectual property keywords may include intellectual property objects, such as information about intellectual property works or virtual character information.

[0064] In one or more embodiments of this specification, the first information may include at least one of information title, information content, and first multimedia information; wherein the type of the first multimedia information includes at least one of static image, dynamic image, and video.

[0065] Furthermore, the first piece of information can be specifically analyzed and identified from basic information such as information title, information content, and first multimedia information (such as images, GIFs, videos, etc.) to obtain information such as time information, location information, service introduction information, service type tags, rights point tags, and intellectual property tags.

[0066] Optionally, the first information may also include derivative information such as original information tags. The original information tags can be used as a basis for subsequently determining information such as service introduction information, service type tags, rights point tags, and intellectual property tags for the offline services.

[0067] Optionally, the first information may also include publisher-related information such as content publishing account identifiers (e.g., account ID, account name, etc.). This publisher-related information can be used to identify the publisher of the first information, and based on a pre-maintained publisher information knowledge base, it can, on the one hand, identify the offline merchant publishing the first information, and on the other hand, identify the intellectual property tags or keywords involved in the first information.

[0068] Optionally, the first information may also include information related to the information release channel, such as the information release time, which can be used as a basis for subsequently determining the content quality factor and push priority of the second information.

[0069] Step 204: Perform structured processing on the first information to generate second information; the second information includes the time information, location information, and service description information of the offline service.

[0070] Specifically, the first information can be structured to generate second information containing standardized fields; the standardized fields can include at least a time field, a location field, and a service description field.

[0071] The location information can be a text identifier used for identification and reading, serving as location description text to identify the address and merchant providing the service (or where the activity takes place) in textual form; that is, telling the user which store / business district the activity is in at which location. The location information can be extracted from the original primary information through semantic parsing. For example, the location information could be the text content "XX Bookstore, XX Dimension Station, Hangzhou City".

[0072] In one or more embodiments of this specification, the step of structuring the first information to generate the second information may specifically include: inputting the first information into a first large language model to obtain the time information and the location information extracted by the first large language model from the first information; or, extracting the time information and the location information from the first information based on preset structured information extraction rules.

[0073] In an optional embodiment, time and location information can be extracted from the first information using a rule-based method, which is more efficient. In another optional embodiment, time and location information can be extracted using a first large language model, which is more accurate and has a wider range of applications.

[0074] The first large language model can be a general-purpose large language model. Alternatively, the first large language model can be a large language model obtained by supervised fine-tuning and / or reinforcement learning on the basis of a general-purpose large language model using information corpus, which has the ability to extract time and location information from information describing offline services.

[0075] Specifically, the first large language model can be a multimodal large language model, used to jointly understand the information title, information content, and first multimedia information, thereby extracting the required information. In the embodiments of this specification, other large language models (such as the second and third large language models) subsequently used to process the first information can also be multimodal large language models, used to jointly understand the information title, information content, and first multimedia information.

[0076] Furthermore, in the embodiments of this specification, the process of interacting with the first large language model and other large language models (such as the second and third large language models) can use prompts. Prompts are instructions or questions entered by the user when interacting with a large language model, used to guide the model to generate expected output. They can include at least one of task descriptions, format specifications, example guidance, and constraints. Prompts can include at least one of thought chain prompts for requiring distributed reasoning by the model, role-playing prompts for specifying the model's identity, and dynamic prompts for adjusting subsequent instructions based on the model's output. For example, in the process of inputting the first information into the first large language model to obtain the time and location information extracted by the first large language model from the first information, the first information can be filled into a preset prompt template and input into the first large language model. This preset prompt template can include task description information instructing the first large language model to extract time and location information from the first information, and can also include format specifications, example guidance, constraints, etc.

[0077] In one or more embodiments of this specification, the second information further includes second multimedia information; the second multimedia information includes at least a portion of the information in the first multimedia information.

[0078] The types of the first multimedia information and the second multimedia information may include, but are not limited to, static images, animated images, and videos.

[0079] Specifically, in the process of generating second information based on the first information, firstly, the first multimedia information can be transferred from an external content platform to a local or internal storage system (such as object storage service), and standardization processing can be performed during this process. For example, format processing can be performed, converting images or videos of different formats to the platform's specified standard format to ensure compatibility. Another example is resizing, generating versions with different resolutions based on different display scenarios (such as thumbnails on list pages and large images on detail pages). Furthermore, watermark processing can be performed, removing watermarks from external platforms in the original information and selectively adding the platform's logo.

[0080] Then, at least a portion of the information in the first multimedia information can be used as the second multimedia information. Optionally, the clear, relevant, and high-quality portions of the first multimedia information can be directly used as the second multimedia information. Optionally, when the first multimedia information contains multiple images or videos, one or more of the clearest and best-composed images are selected as the second multimedia information using an image quality assessment algorithm to ensure the quality of the final displayed content.

[0081] In one or more embodiments of this specification, the service introduction information specifically includes title information and body information; the step of structuring the first information to generate second information may specifically include: inputting the first information into a second language model to obtain the title information and body information generated by the second language model based on the first information; wherein, the second language model is trained based on first training data in the vertical domain to which the first information belongs; the first training data includes sample information and sample tags for the sample information; the sample tags include title tags and body tags corresponding to the sample information.

[0082] The title tags and body tags can be used to provide a standardized description of the offline services described in the sample information, conforming to the display requirements of the user interface. In practical applications, the title tags and body tags can optionally be manually annotated; alternatively, they can also be extracted from the sample information by a large language model with a large number of parameters, serving as a teacher model.

[0083] Furthermore, during the training process of the second language model, sample information can be input into the second language model to be trained / training, and the predicted title and predicted body text output by the second language model according to the prompt words are obtained after understanding, summarizing / rewriting the sample information. Then, based on the difference between the predicted title and the title tag and the difference between the predicted body text and the body text tag, the prediction loss value can be calculated, and the model parameters of the second language model can be adjusted until the preset model training termination condition is reached (e.g., the loss value reaches the corresponding threshold or the number of parameter adjustments reaches the corresponding threshold), and the trained second language model can be obtained.

[0084] Furthermore, the first information can be input into a second language model to obtain the time information, location information, title information, and body text information generated by the second language model based on the first information; wherein, the second language model is trained based on first training data in the vertical domain to which the first information belongs; the first training data includes sample information and sample tags for the sample information; the sample tags include time tags, location tags, title tags, and body text tags corresponding to the sample information.

[0085] Based on at least some embodiments of this specification, considering that accurately extracting the title and body text from the first information requires a second language model with an accurate understanding of the first information, in practical applications, the second language model can be trained on corpus from the vertical domain to which the first information belongs. Since different vertical domains have different language expressions and habits, training the second language model using corpus from the vertical domain to which the first information belongs enables the second language model to extract the title and body text more accurately.

[0086] In one or more embodiments of this specification, before structuring the first information to generate the second information, the process may further include: determining whether the first information matches a preset filtering rule; the preset filtering rule includes at least one of the following: the content length is less than a preset length threshold, the content is unrelated to the target vertical domain, or the service location is not within the target service area. Accordingly, structuring the first information to generate the second information may specifically include: structuring the first information that does not match the preset filtering rule to generate the second information.

[0087] Based on at least some embodiments of this specification, by introducing pre-filtering rules based on content length, domain relevance, and service area, information is screened before structured processing, resulting in a triple benefit of improved data processing efficiency, guaranteed quality of generated content, and enhanced system adaptability. This mechanism effectively reduces the computational load of invalid data on subsequent large language models and controls the quality of input data from the source, thereby improving the overall system performance and the reliability of model output results.

[0088] Step 206: Based on at least part of the first information and the second information, associate and match the target offline merchant with the points of interest in the point of interest database to determine the target point of interest corresponding to the target offline merchant.

[0089] Points of interest (POIs) are a concept in the fields of electronic maps, navigation, and local life services, and can be used to represent entities with specific names, categories, and geographic coordinates.

[0090] In embodiments of this specification, the type of the target point of interest may include a business district or a store.

[0091] In this context, "points of interest" (POIs) within a commercial district can refer to a comprehensive geographical area encompassing multiple business entities. These PPIs can serve as a platform for disseminating information about activities within the entire area, or as a location marker when it's impossible to pinpoint a single store. Examples of PPIs within commercial districts could include Hangzhou XX, Beijing Xidan XX City, and Shanghai XX Port.

[0092] Store-type points of interest can refer to a single, specific offline merchant or service outlet, offering high precision. Examples of store-type points of interest could include XX Bookstore (XX Store), XX Anime Theme Store (Xidan XX City Store), XX Planet Pop-up Store (Shanghai XX Port), etc.

[0093] In the embodiments described in this specification, by defining two levels of Points of Interest (POIs) – "business district" and "store" – granularity and flexibility in associating information with geographic location are achieved. Information such as anime conventions and themed events covering the entire business district can be directly associated with the business district's POI, while information such as new product launches and promotions at specific stores is associated with the more precise store POI. This hierarchical mechanism ensures that geographic information of different ranges is accurately and reasonably represented, thereby providing consumers with clear and practical location guidance in electronic maps and user interfaces, significantly improving the usability of local service information and user experience.

[0094] Typically, each POI in a Point of Interest (POI) database can contain basic identifier information, geographic location information (i.e., geographic coordinates), location information (also known as address information), city information, province information, category information, attribute information, account information, etc., but is not limited to these. Basic identifier information can be used to represent the identity of the POI, specifically including POI ID and POI name. Geographic location information can represent the geographic location of the POI, specifically including latitude and longitude information, geocoded address, etc. Location information can specifically represent the detailed address of the POI, specifically implemented as structured address description text, which can include merchant address and merchant name, for example, Hangzhou XX Dimension Station XX Bookstore. Attribute information can be used to represent the characteristics of the POI. For example, taking a merchant as an example, location attribute information can include business status (e.g., open, closed, under renovation, temporarily closed), business hours (e.g., daily business hours and special business hours on holidays and weekends), associated brand (if it is a chain store, it can be associated with its headquarters or brand owner), price range (e.g., average spending per person), etc., but is not limited to these. Account information can be used to represent the online accounts used by operators of points of interest to publish information.

[0095] Traditional POI databases may simply categorize "XX Bookstore" as "bookstore". However, based on the solutions implemented in this specification, through data integration and rules, vertical tags such as "anime" and "grain store" can be added. Furthermore, it can be linked with IP knowledge bases, thereby enhancing and reconstructing the vertical semantics of general POI databases.

[0096] In one or more embodiments of this specification, a specific method for determining target points of interest corresponding to the target offline merchant is further provided.

[0097] In an optional embodiment, the point of interest database includes point of interest address information for each point of interest; step 206, which involves associating and matching the target offline merchant with the points of interest in the point of interest database based on at least a portion of the first information and the second information, to determine the target point of interest corresponding to the target offline merchant, may specifically include: matching the location information with the point of interest address information in the point of interest database; if the match is successful, the matched point of interest is determined as the target point of interest corresponding to the target offline merchant.

[0098] In an optional embodiment, the point of interest database includes point of interest account information for each point of interest; the first information includes a content publishing account identifier; step 206, which involves associating and matching the target offline merchant with the points of interest in the point of interest database based on at least a portion of the first information and the second information to determine the target point of interest corresponding to the target offline merchant, may specifically include: matching the content publishing account identifier with the point of interest account information in the point of interest database; if the match is successful, the matched point of interest is determined as the target point of interest corresponding to the target offline merchant.

[0099] Optionally, the first information may originate from the content publishing account of the target offline merchant. In this case, when the first information is obtained, the association between the first information and the content publishing account can already be determined. Therefore, the target interest point can be queried from the interest point database based on the affiliation between the content publishing account and the target offline merchant.

[0100] Optionally, the first information may originate from the keywords of the vertical domain. In this case, when obtaining the first information based on the solution of the embodiments of this specification, the first information may include the content publishing account identifier. Thus, the target interest point corresponding to the target offline merchant who published the first information can also be queried from the interest point database based on the content publishing account.

[0101] Based on at least some embodiments of this specification, location information, content publishing account identifiers, and other explicit information can be matched with corresponding fields (such as address fields and account fields) of merchant information in the point of interest database, so as to successfully recall the target point of interest after a successful match.

[0102] In an optional embodiment, step 206, which involves associating and matching the target offline merchant with points of interest in the point of interest database based on at least a portion of the first information and the second information, to determine the target point of interest corresponding to the target offline merchant, may specifically include: retrieving a set of candidate points of interest from the point of interest database based on the query vector corresponding to the first information; filtering the set of candidate points of interest; and determining the target point of interest from the set of candidate points of interest.

[0103] Specifically, the first information (e.g., information title, information content, etc.) can be input into a vector embedding model to generate a query vector; the similarity between the query vector and the interest point vectors pre-stored in the interest point database can be calculated; and the candidate interest point set can be recalled from the interest point database based on the similarity.

[0104] Based on the solutions in the embodiments of this specification, when location information is ambiguous or missing, the semantics of the first information itself can be used to recall the target point of interest from the point of interest database through vector matching.

[0105] Based on the scheme described in this specification, the original information of the first information and the structured data of the second information are used to collaboratively associate points of interest, which greatly improves the accuracy and coverage of point of interest matching.

[0106] Step 208: Provide the second information to be displayed in the user interface corresponding to the target point of interest.

[0107] Optionally, if the executing entity is a server, the server may provide the second information to the user terminal so that the second information can be displayed in the user terminal in the user interface corresponding to the target point of interest.

[0108] Optionally, if the executing entity is a user terminal, the second information can be displayed in the user terminal in a user interface corresponding to the target point of interest.

[0109] Furthermore, when displayed in the user interface of a user terminal, the second information and the interactive location element corresponding to the point of interest can be displayed in association in the user interface of the user terminal; the interactive location element can be configured to visualize the geographical location of the target offline merchant on an electronic map after being triggered.

[0110] While one or more embodiments of this specification provide method steps as described in the embodiments or flowcharts, it is understood that the order of steps listed in the embodiments or flowcharts is merely one possible execution order among many steps and does not represent the only possible execution order. The order of some steps may be adjusted according to actual needs, or some steps may be omitted. When the claims involve method steps, changes in the order of such steps, or parallel execution between steps, are also within the scope of protection of the claims.

[0111] Figure 2 The method described herein involves acquiring first information from a target offline merchant published on an external content platform, describing the offline services offered by the merchant. This first information is then structured to generate second information containing time, location, and service descriptions of the offline services. Based on at least a portion of the first and second information, the target offline merchant is matched with points of interest in a point-of-interest database to determine the target point of interest corresponding to the merchant. Finally, the second information is provided for display in the user interface corresponding to the target point of interest. This method achieves automated cold start for local information services by automatically acquiring and intelligently processing information from external content platforms. On one hand, it significantly reduces the platform's reliance on merchants' active participation; on the other hand, it reduces the operational burden on merchants during the local information service cold start process; and thirdly, it provides users with timely and efficient one-stop information on vertical fields, offering convenience to users in these fields.

[0112] based on Figure 2 In addition to the method described herein, this specification also provides some improved implementation methods, which will be described below.

[0113] In one or more embodiments of this specification, the information provision method may further include: inputting the first information into a first large language model to obtain a benefit point label of the offline service determined by the first large language model based on the first information; the benefit point label is used to represent the benefits that the offline service can provide.

[0114] In this context, obtaining the benefit point tag for the offline service essentially establishes a connection between the benefit point tag and the second information; in other words, the second information carries the benefit point tag. Typically, one benefit point tag can be used to label one or more information items. Therefore, multiple information items can carry the same benefit point tag, allowing for the classification or filtering of information through the benefit point tag.

[0115] In practical applications, one or more benefit tags can be identified. These benefit tags can also be called highlight tags or selling point tags. As an example, the benefit tags may include free, free gifts, benefits, limited edition, and other privileges.

[0116] In practical applications, the primary language model used to extract rights point labels can be a general language model, or it can be a language model obtained by supervising fine-tuning and / or reinforcement learning on the basis of a general language model using information corpus, which is capable of determining rights point labels for information describing offline services.

[0117] Furthermore, the benefit point tag can be used to display in the information details page of the second information; and / or, the benefit point tag can be used as the basis for filtering target information carrying the benefit point tag in the information collection page; the target information includes the second information.

[0118] The information collection page may display multiple pieces of information in a list format, matrix format, map format, or other formats, but is not limited to these.

[0119] In practical applications, the rights and benefits tags can serve as information points for users to quickly understand the offline services described in the second information when viewing the second information.

[0120] For the user terminal, the rights and benefits tags can be displayed on the information details page of the second information.

[0121] In practical applications, the rights and benefits tags can also serve as filtering conditions for users to route to the second information, so that users can quickly find the second information and learn about the offline services.

[0122] Optionally, for the user terminal, a single selection control corresponding to each of the multiple benefit point labels can be displayed on the information collection page containing the second information; then, in response to the selection operation of the target benefit point label based on the single selection control, the information collection page can be updated to display target information carrying the target benefit point label, such as the second information.

[0123] Optionally, a rights point label selection control (such as a drop-down control) can be displayed on the information collection page containing the second information; then, in response to the user's triggering operation on the rights point label selection control, a rights point label set (such as a list) can be displayed; subsequently, in response to the user's triggering operation on a target rights point label in the rights point label set, the information collection page can be updated to display target information carrying the target rights point label, such as the second information.

[0124] Based on at least some embodiments of this specification, by introducing a first major language model to automatically extract and associate benefit point tags, an intelligent conversion from raw information to structured tags is achieved. This enables the system to accurately mark and classify the selling points of massive amounts of information, thereby supporting users to efficiently filter and quickly locate information based on benefit points in the front-end interface. This significantly improves information retrieval efficiency and the accuracy of content distribution, and optimizes the user experience of obtaining core service information.

[0125] In one or more embodiments of this specification, the information provision method may further include: inputting the first information into a second language model to obtain a service type label generated by the second language model based on the first information for tagging the second information; the service type label is used to represent the service form of the offline service; the second language model is trained based on first training data of the vertical domain to which the first information belongs; the first training data includes sample information and sample labels for the sample information; the sample labels include a title label, a body label, and a service type sample label corresponding to the sample information.

[0126] The process of obtaining a service type tag to label the second information essentially establishes an association between the service type tag and the second information; in other words, the second information carries the service type tag. Typically, a service type tag can be used to label one or more pieces of information. Therefore, multiple pieces of information can carry the same service type tag, allowing for the classification or filtering of information using the service type tag.

[0127] In practical applications, one or more service type tags can be identified.

[0128] In practical applications, the service type can be related to a specific vertical domain. The way service types are categorized can differ for different vertical domains. For example, for the anime / manga / anime / manga (ACG) vertical domain, the service type tags could include "New Tani Release," "Good Tani Restock," "Pop-up Event," "Main Event," "Anime Convention," etc.

[0129] Optionally, the service type sample labels may be manually labeled; or alternatively, the service type sample labels may be determined by another large language model with a large number of parameters, serving as a teacher model, based on the sample information.

[0130] Furthermore, during the training process of the second language model, sample information can be input into the second language model to be trained / training, and the predicted service type output by the second language model according to the prompt words is obtained after understanding the sample information. Then, based on the difference between the predicted service type and the service type sample label, the predicted loss value can be calculated, and the model parameters of the second language model can be adjusted until the preset model training termination condition is reached (e.g., the loss value reaches the corresponding threshold or the number of parameter adjustments reaches the corresponding threshold), and the trained second language model can be obtained.

[0131] Based on at least some embodiments of this specification, considering the differences in language expression and habits in different vertical fields, in practical applications, the second language model can be trained based on the corpus in the vertical field to which the first information belongs. By adjusting the model parameters based on the prediction error of the service type during the training process, the trained model can accurately determine the service type label for the first information in that vertical field.

[0132] Furthermore, the service type tag can be used to display on the information details page of the second information; and / or, the service type tag can be used as a basis for filtering target information carrying the service type tag on the information collection page; the target information includes the second information.

[0133] In practical applications, the service type tag can serve as a quick way for users to understand the offline services described in the second information when viewing the second information.

[0134] For user terminals, the service type label can be displayed on the information details page of the second information.

[0135] In practical applications, the service type tag can also be used as a filtering condition for the process of routing users to the second information, so that users can quickly find the second information and learn about the offline service.

[0136] Optionally, for a user terminal, a single-selection control corresponding to each of the multiple service type tags can be displayed on the information collection page containing the second information; then, in response to the selection operation of the target service type tag based on the single-selection control, the information collection page can be updated to display the target information carrying the target service type tag, such as the second information.

[0137] Optionally, a service type tag selection control (such as a drop-down control) may be displayed on the information collection page containing the second information; then, in response to a user's triggering operation on the service type tag selection control, a service type tag set (such as a list) may be displayed; subsequently, in response to a user's triggering operation on a target service type tag in the service type tag set, the information collection page may be updated to display target information carrying the target service type tag, such as the second information.

[0138] Based on at least some embodiments of this specification, by employing a second language model specifically trained with vertical domain data, the problem of bias in understanding terminology in specific domains by general models is effectively solved, achieving accurate identification and labeling of service type tags. By applying the generated service type tags to the display of information details pages and the filtering of information collection pages, the system can support users to efficiently search and navigate by service type, greatly improving the efficiency and relevance of information acquisition, and ultimately optimizing the user's service discovery experience in vertical domains.

[0139] In one or more embodiments of this specification, the information provision method may further include: inputting the first information into a third language model to obtain an intellectual property tag generated by the third language model based on the first information for marking the second information; the intellectual property tag is used to represent the intellectual property object involved in the offline service; the intellectual property object includes at least one of intellectual property works and virtual characters; the third language model is trained based on second training data in the vertical domain to which the first information belongs; the second training data includes at least one of text data and image-text data; the text data includes at least one of conceptual text and relational text, the conceptual text is used to explain a specified intellectual property object, and the relational text is used to explain the relationship between multiple intellectual property objects; the image-text data includes a specified image and descriptive information about the intellectual property object in the specified image.

[0140] As an example, conceptual text can be used to indicate what a certain intellectual property object is, what a certain intellectual property object is, etc., such as "Arknights is xxxxxx, and its alias is yyyy", "xxxxxx is Arknights", etc.

[0141] As an example, relational text can be used to represent the relationship between an intellectual property work and a virtual character, such as "xx IP is xxxxx, which includes the characters aa, bb, and cc", "the character aa is xxxx, which belongs to xx IP", "the aliases of the character aa include aa1, aa2, and aa3", etc.

[0142] As an example, image-text data can include two cases: images containing specific virtual characters and images without specific virtual characters. For images containing specific virtual characters, the corresponding description information could be such as "The IP address in this anime image is xxxx, and it contains the characters aa and bb"; for images without specific virtual characters, the corresponding description information could be such as "The IP address in this anime image is xxxx, and it does not contain any specific characters".

[0143] The acquisition of the intellectual property label used to mark the second information can be considered as establishing an association between the intellectual property label and the second information, or in other words, the second information carries the intellectual property label. Typically, one intellectual property label can be used to mark one or more pieces of information; therefore, multiple pieces of information can carry the same intellectual property label, thereby allowing for the classification or filtering of information through the intellectual property label.

[0144] In practical applications, one or more intellectual property tags can be identified. These tags can be used to describe one or more intellectual property objects. Typically, these intellectual property objects are related to each other. For example, for a given piece of information, two intellectual property tags can be identified: "Journey to the West" and "Sun Wukong," where "Sun Wukong" is a fictional character in the intellectual property work "Journey to the West."

[0145] In practical applications, the intellectual property tags can be related to specific vertical fields. The classification method and corresponding tag sets of intellectual property tags can differ for different vertical fields. For example, in the ACG (Anime, Comics, and Games) vertical field, the intellectual property tags can include tags for intellectual property works and tags for virtual characters. Examples of intellectual property work tags could be works like *Detective Conan* or *Pokémon*, while examples of virtual character tags could be Conan (from *Detective Conan*) or Pikachu (from *Pokémon*).

[0146] Furthermore, in the process of determining intellectual property labels, Retrieval-augmented Generation (RAG) technology can be used to query the intellectual property object knowledge base in order to identify the intellectual property object involved in the information in the first piece of information and obtain the corresponding intellectual property label. As an example, the intellectual property object knowledge base may contain multiple intellectual property works, virtual characters contained in each intellectual property work, and the names, aliases, nicknames, etc. of each virtual character, and is not limited to these.

[0147] RAG (Reference-Based Language) is a method for optimizing the output of large language models, enabling them to reference authoritative knowledge bases beyond the training data before generating responses. RAG comprises three main processes: retrieval, augmentation, and generation. Retrieval: Based on the user's query, relevant information is retrieved from an external knowledge base. Specifically, the user's query is converted into a vector using an embedding model for comparison with relevant knowledge stored in a vector database. Similarity search identifies the top K most relevant data points. Augmentation: The user's query and the retrieved relevant knowledge are embedded together into a pre-defined prompt template. Generation: The retrieved and augmented prompt is input into the large language model to generate the desired output. Through this process, RAG models can be used in various natural language processing tasks, such as question-answering systems, document generation and automatic summarization, intelligent assistants and virtual agents, information retrieval, and knowledge graph filling.

[0148] Furthermore, the intellectual property object knowledge base may include information constructed by experts, as well as information obtained from the property rights holders associated with the intellectual property object (e.g., official websites). In practical applications, the knowledge in the intellectual property object knowledge base can be updated or supplemented based on feedback from model prediction results. It can also be updated or supplemented based on the development of intellectual property objects in that specific vertical field.

[0149] Furthermore, the intellectual property label can be used to display on the information details page of the second information; and / or, the intellectual property label can be used as a basis for filtering target information carrying the intellectual property label on the information collection page; the target information includes the second information.

[0150] In practical applications, the intellectual property label can serve as a point of information for users to quickly understand the offline services described in the second information when viewing the second information.

[0151] For user terminals, the intellectual property label can be displayed on the information details page of the second information.

[0152] In practical applications, the intellectual property label can also serve as a filtering condition for users to route to the second information, so that users can quickly find the second information and learn about the offline service.

[0153] Optionally, for the user terminal, an intellectual property label selection control (such as a drop-down control) can be displayed on the information collection page containing the second information; then, in response to the user's triggering operation on the intellectual property label selection control, an intellectual property label collection (such as a list) can be displayed; subsequently, in response to the user's triggering operation on a target intellectual property label in the intellectual property label collection, the information collection page can be updated to display target information carrying the target intellectual property label, such as the second information.

[0154] Optionally, in the information collection page containing the second information, a single selection control corresponding to each of the multiple intellectual property labels can be displayed; then, in response to the selection operation of the target intellectual property label based on the single selection control, the information collection page can be updated to display target information carrying the target intellectual property label, such as the second information.

[0155] Based on the embodiments of this specification, a second language model is trained using knowledge from the vertical domain, and then the service type tag is determined using the second language model. Similarly, a third language model is trained using knowledge from the vertical domain, and then the intellectual property tag is determined using the third language model. Therefore, the service type tag and the intellectual property tag can be considered as vertical attribute tags related to the vertical domain, or in other words, vertical attribute tags based on the domain knowledge of the vertical domain. In practical applications, the corresponding set of vertical attribute tags is usually different for different vertical domains.

[0156] Based on at least some embodiments of this specification, by introducing a third language model specifically trained with vertical domain knowledge and combining it with a dynamically updated intellectual property object knowledge base queried using retrieval enhancement generation technology, the system effectively solves the identification ambiguity problems caused by non-standard expressions such as aliases and nicknames in specific domains, achieving accurate identification and labeling of intellectual property works and virtual characters. By establishing a fine-grained intellectual property tagging system and applying it to front-end filtering and display, the system significantly improves the structuring and searchability of information content, enabling users to quickly locate relevant service information for specific works or characters, ultimately achieving a deeper understanding of vertical domain content and a more precise search experience.

[0157] In one or more embodiments of this specification, during the stage of displaying the second information to the user, the information provision method may further include: obtaining the target user's preference for different intellectual property objects; determining the matching degree between the second information and the target user based on the preference degree and the intellectual property tag; and determining the push priority of the second information in the information collection page for the target user based on the matching degree.

[0158] Specifically, the higher the matching degree, the higher the push priority of the second information, and the more prominent the second information will be displayed in the information collection page compared to other information.

[0159] In practical applications, for a target user, the push priority of multiple information items can be calculated simultaneously. Then, by sorting the push priorities of different information items, the items with higher push priorities are displayed in a relatively earlier position or in a relatively prominent form.

[0160] In an optional embodiment, a time decay factor for the second information relative to the target user can also be calculated; then, the push priority of the second information in the information collection page for the target user can be determined based on the matching degree and the time decay factor.

[0161] Specifically, the current time when the target user accesses the information collection page can be obtained, and then a time decay factor can be calculated based on the current time and the time information of the offline service.

[0162] Optionally, taking the push priority being positively correlated with the matching degree and negatively correlated with the time decay factor as an example, then: if the time information is later than the current time, it indicates that the offline service has not expired. In this case, the time decay factor can be positively correlated with the time interval between the time information and the current time; that is, the shorter the time interval, the smaller the time decay factor and the higher the push priority; conversely, the longer the time interval, the larger the time decay factor and the lower the push priority. In a special case, if the time information is earlier than the current time, it indicates that the offline service has expired. In this case, the time decay factor can be set to the maximum to minimize the push priority.

[0163] In an optional embodiment, a spatial decay factor for the second information to the target user can also be calculated; then, the push priority of the second information in the information collection page for the target user can be determined based on the matching degree and the spatial decay factor.

[0164] Specifically, the current location of the target user when accessing the information collection page can be obtained, and then a spatial decay factor can be calculated based on the current location and the location information of the offline service.

[0165] Optionally, taking the push priority as positively correlated with the matching degree and negatively correlated with the spatial decay factor as an example, then: the spatial decay factor can be positively correlated with the distance interval between the location information and the current location, that is, the shorter the distance interval, the smaller the spatial decay factor and the greater the push priority, and vice versa, the longer the spatial interval, the greater the spatial decay factor and the lower the push priority.

[0166] In an optional embodiment, a content quality factor for the second information can also be calculated; then, the priority of pushing the second information in the information collection page for the target user can be determined based on the matching degree and the content quality factor.

[0167] Specifically, based on at least a portion of the content in the first and second pieces of information, a large language model can be used to determine the content quality factor of the second piece of information. This content quality factor reflects the quality of the content in the second piece of information; a larger content quality factor indicates higher quality content, and a smaller content quality factor indicates lower quality content. Typically, the push priority is positively correlated with the matching degree and also positively correlated with the content quality factor.

[0168] In practical applications, factors affecting content quality factors may include, but are not limited to, the quality of the news title, news content, and first multimedia information contained in the first news item, the authority of the content publishing account of the first news item, the authority of the news publishing channel of the first news item, and whether the information types contained in the second news item (including time information, location information, service introduction information, various tags, etc.) are complete, etc.

[0169] In an optional embodiment, the priority of pushing the second information in the information collection page for the target user can be determined based on one or more of the matching degree, the time decay factor, the spatial decay factor, and the content quality factor.

[0170] As an example, assuming the priority of pushing the second information in the information collection page targeting the target user is determined based on the matching degree, the time decay factor, the spatial decay factor, and the content quality factor, and the push priority is positively correlated with the matching degree and the content quality factor, and negatively correlated with the time decay factor and the spatial decay factor, then, for example, the push priority can be determined based on the following formula: Push Priority = Matching Degree × (Content Quality Factor / (Time Decay Factor × Spatial Decay Factor)). In practical applications, weight parameters can be added to this formula. The calculation method for push priority is not limited to the methods listed here.

[0171] Based on at least some embodiments of this specification, a multi-dimensional ranking model is constructed by integrating user preference, spatiotemporal decay factor, and content quality factor, achieving a leap from "one-way distribution" to "intelligent adaptation" in information push strategy. This technical approach achieves precise matching based on intellectual property tags and user interests, dynamically calibrates the timeliness and geographical relevance of information through spatiotemporal factors, and introduces content quality factors to ensure information value. This allows the push results to simultaneously meet the triple requirements of user interest, spatiotemporal context, and content quality, thereby significantly improving the accuracy of content distribution and user participation efficiency in complex information environments.

[0172] In one or more embodiments of this specification, the information provision method may further include: obtaining user historical interaction information of the target point of interest; the user historical interaction information includes at least one of historical access information and historical evaluation information for the target point of interest; and determining the overall recommendation tag of the target offline merchant based on the user historical interaction information.

[0173] The overall recommendation tag can reflect information such as the target user group, the user group attracted, or the service quality of the target offline merchant. Therefore, by providing this auxiliary information, users can better understand the target offline merchant and decide whether to visit it offline. For example, the overall recommendation tag could be something like "x people have visited".

[0174] Furthermore, when a vertical category includes multiple sub-categories, determining the overall recommendation tag for the target offline merchant based on the user's historical interaction information may specifically include: filtering a portion of historical interaction information corresponding to the target sub-category from the user's historical interaction information; and determining personalized recommendation tags associated with the target offline merchant and the sub-category based on the portion of historical interaction information. As an example, the personalized recommendation tag could be something like "xx group x people visited".

[0175] In practical applications, on the user terminal side, optionally, if the sub-domain associated with the target user includes the target sub-domain, the personalized recommendation tag can be displayed to the target user when showing the target offline merchants.

[0176] Furthermore, the information provision method may further include: obtaining individual historical interaction information of a target user; the individual historical interaction information includes at least one of the target user's individual access information and individual evaluation information for various offline merchants; determining individual recommendation tags for the target offline merchants based on the user's historical interaction information and the individual historical interaction information; the individual recommendation tags are used to be displayed in association with the target offline merchants in the user interface shown to the target user.

[0177] The individual recommendation tags can reflect the historical overlap or personalized matching degree between the target user and the target offline merchant. Therefore, by providing this auxiliary information, users can better understand the target offline merchant and decide whether to visit it offline. For example, the overall recommendation tags could be "You've been there before," "You visited a week ago," etc.

[0178] Based on at least some embodiments of this specification, a multi-level recommendation tagging system from group to individual is constructed, enabling the precise delivery of merchant recommendation information from general to personalized. On the one hand, "overall recommendation tags" reflecting the overall popularity of merchants can be generated based on historical interaction data. On the other hand, more targeted "personalized recommendation tags" can be formed through vertical sub-domain filtering. Furthermore, exclusive "individual recommendation tags" can be generated by combining user's personal behavior data. This hierarchical and progressive tagging mechanism allows recommendation information to reflect the public recognition of merchants, match the preferences of users' interest groups, and evoke users' personal historical memories. Thus, in the decision-making process, it simultaneously satisfies users' triple needs for group consensus, circle recognition, and personal relevance, significantly improving the credibility and decision-making reference value of recommendation information.

[0179] The various technical features in the above embodiments can be combined arbitrarily, as long as there is no conflict or contradiction between the combinations of features. However, due to space limitations, they have not been described one by one. Therefore, the arbitrary combination of various technical features in the above embodiments is also within the scope of this specification.

[0180] Based on the above description, the embodiments of this specification provide a flowchart of an information provision method in a practical application scenario, as shown below. Figure 3 As shown.

[0181] like Figure 3 As shown, the first stage involves acquiring primary information from multiple sources.

[0182] Specifically, first-hand information can be obtained from multiple data sources (i.e., multiple external content publishing platforms and multiple information dissemination channels). Furthermore, the process of obtaining first-hand information can employ a targeted collection strategy; that is, first-hand information can be collected in a targeted manner. On the one hand, one can start from offline stores (such as offline grain stores), first identify online accounts, and then collect the content published by them from online platforms or communities. On the other hand, relevant information can be collected online using intellectual property keywords as clues. By adopting a targeted collection strategy, the validity and relevance of the collected information can be ensured, efficiency can be improved, noise data can be reduced, and invalid information can be prevented from entering the processing chain.

[0183] In the second stage, a large language model with the ability to understand the language of vertical domains is used to determine the second information and related tags based on the first information.

[0184] In practical applications, the content of the first information used to input into the large language model may include, but is not limited to, information page links, information titles, information content, original information multimedia information, original information tags, content publishing account ID, content publishing account name, information publishing channel, and information publishing time.

[0185] The content of the second information determined by the large language model may include, but is not limited to, time information, location information, title information, and body information.

[0186] The association tags for the second information determined by the large language model may include, but are not limited to, intellectual property tags, service type tags, rights point tags, and information recommendation tags.

[0187] In addition, in the process of determining the second information, the second multimedia information in the second information can be determined based on the first multimedia information in the first information.

[0188] In the third stage, based on the point of interest database, the target point of interest corresponding to the target offline merchants associated with the first piece of information can be determined.

[0189] The specific types of points of interest can include stores or business districts. The information for each point of interest in the database can include, but is not limited to, basic identification information, geographic location information (i.e., geographic coordinate information), location information (also known as address information), city information, province information, category information, attribute information, account information, etc.

[0190] Therefore, in practical applications, at least some information from the first and second information can be used to recall points of interest associated with the current first / second information from the point of interest database.

[0191] In the fourth stage, after the target point of interest is determined, the second information can be displayed on the user terminal side in the user interface corresponding to the target point of interest.

[0192] In practical applications, the second information (which may include time information, location information, title information, and body information, etc.) and its associated tags can be displayed in the first area of ​​the user interface; interactive location elements of target points of interest associated with the second information can be displayed in the second area of ​​the user interface. The interactive location elements can be configured to visualize the geographical location of the target offline merchant on an electronic map after being triggered.

[0193] Based on the solutions implemented in this specification, the original information published by offline merchants on online platforms or social media is collected and intelligently processed to automatically generate information (such as information detail pages and associated merchant detail pages) for a local information service platform, allowing merchants to go online without actively registering. Specifically, merchants can go online without participating in the entire process from information generation to publication and display, or they can simply authorize and confirm before the information is published and displayed, or they can intervene in the later stages of store information operation by claiming the store.

[0194] Based on the content generation mechanism in the cold start mode of the embodiments of this specification, the original information related to the vertical field is collected in a targeted manner, and after understanding and processing, the information required by the local information service platform is generated, which can break the traditional dependence on B-end entry.

[0195] Figure 4 This is a schematic diagram of an information details page provided in an embodiment of this specification.

[0196] like Figure 4 As shown, the information details page can display secondary information. Specifically, it can display secondary multimedia information, title information, body information, location information, time information, etc.

[0197] The information details page can display related tags for the second piece of information. Specifically, it can display service type tags, intellectual property tags, rights and interests tags, etc. These related tags help users quickly understand the second piece of information and facilitate routing users from upstream pages to the current information details page.

[0198] The information details page may also include interactive location elements associated with points of interest matching the offline merchants of the published second information (or, in other words, the first information that generated the second information). These interactive location elements can be configured to visualize the geographical location of the offline merchant "Toy-c (Hangzhou XX Mall Store)" on an electronic map when triggered.

[0199] Figure 4 The examples provided are merely examples. In actual applications, information detail pages can be displayed in different forms or contain different page content.

[0200] Based on the same idea, embodiments of this specification also provide apparatus corresponding to the above methods.

[0201] Figure 5 The embodiments provided in this specification correspond to Figure 2 A schematic diagram of the structure of an information providing device.

[0202] like Figure 5 As shown, the device may include:

[0203] The first information acquisition module 502 is used to acquire first information published on an external content platform; the first information is used to describe the offline services provided at the target offline merchant.

[0204] The second information generation module 504 is used to perform structured processing on the first information to generate the second information; the second information includes the time information, location information and service introduction information of the offline service;

[0205] The point of interest determination module 506 is used to associate and match the target offline merchant with the points of interest in the point of interest database based on at least part of the content in the first information and the second information, and determine the target point of interest corresponding to the target offline merchant.

[0206] The second information providing module 508 is used to provide the second information for display in a user interface corresponding to the target point of interest.

[0207] based on Figure 5 The embodiments of this specification also provide some specific implementation schemes of the method, which are described below.

[0208] Optionally, the first information acquisition module 502 is specifically used to: collect the first information published by the content publishing account of the target offline merchant from the external content platform; and / or, collect the first information matching the vertical domain keywords from the external content platform based on specified vertical domain keywords; the vertical domain keywords include intellectual property keywords.

[0209] Optionally, the first information includes at least one of information title, information content, and first multimedia information; wherein the type of the first multimedia information includes at least one of static image, dynamic image, and video.

[0210] Optionally, the second information further includes second multimedia information; the second multimedia information includes at least a portion of the information in the first multimedia information.

[0211] Optionally, the second information generation module 504 is specifically used to: input the first information into the first large language model to obtain the time information and the location information extracted by the first large language model from the first information; or, extract the time information and the location information from the first information based on preset structured information extraction rules.

[0212] Optionally, the information providing device is further configured to: input the first information into a first large language model to obtain a benefit point label of the offline service determined by the first large language model based on the first information; the benefit point label is used to represent the benefits that the offline service can provide.

[0213] Optionally, the benefit point label is used to display on the information details page of the second information; and / or, the benefit point label is used as the basis for filtering target information carrying the benefit point label on the information collection page; the target information includes the second information.

[0214] Optionally, the service introduction information specifically includes title information and body information; the second information generation module 504 is specifically used to: input the first information into a second large language model to obtain the title information and body information generated by the second large language model based on the first information; wherein, the second large language model is trained based on first training data of the vertical domain to which the first information belongs; the first training data includes sample information and sample tags for the sample information; the sample tags include title tags and body tags corresponding to the sample information.

[0215] Optionally, the information providing device is further configured to: input the first information into a second language model to obtain a service type label generated by the second language model based on the first information for labeling the second information; the service type label is used to represent the service form of the offline service; the sample label in the first training data further includes a service type sample label corresponding to the sample information.

[0216] Optionally, the service type tag is used to display on the information details page of the second information; and / or, the service type tag is used as the basis for filtering target information carrying the service type tag on the information collection page; the target information includes the second information.

[0217] Optionally, the information providing device is further configured to: input the first information into a third language model to obtain an intellectual property label generated by the third language model based on the first information for marking the second information; the intellectual property label is used to represent the intellectual property object involved in the offline service; the intellectual property object includes at least one of intellectual property works and virtual characters; the third language model is trained based on second training data in the vertical domain to which the first information belongs; the second training data includes at least one of text data and image-text data; the text data includes at least one of conceptual text and relational text, the conceptual text is used to explain a specified intellectual property object, and the relational text is used to explain the relationship between multiple intellectual property objects; the image-text data includes a specified image and descriptive information about the intellectual property object in the specified image.

[0218] Optionally, the intellectual property label is used to display on the information details page of the second information; and / or, the intellectual property label is used as a basis for filtering target information carrying the intellectual property label on the information collection page; the target information includes the second information.

[0219] Optionally, the information providing device is further configured to: obtain the target user's preference for different intellectual property objects; determine the matching degree between the second information and the target user based on the preference and the intellectual property tag; and determine the push priority of the second information in the information collection page for the target user based on the matching degree.

[0220] Optionally, the information providing device is further configured to: determine whether the first information matches a preset filtering rule before performing structured processing on the first information to generate the second information; the preset filtering rule includes at least one of the following: the content length is lower than a preset length threshold, the content is unrelated to the target vertical domain, and the service location is not within the target service area; the second information generation module 504 is specifically configured to: perform structured processing on the first information that does not match the preset filtering rule to generate the second information.

[0221] Optionally, the point of interest database includes point of interest address information for each point of interest; the point of interest determination module 506 is specifically used to: match the location information with the point of interest address information in the point of interest database; if the match is successful, the matched point of interest is determined as the target point of interest corresponding to the target offline merchant.

[0222] Optionally, the point of interest database includes point of interest account information for each point of interest; the first information includes a content publishing account identifier; the point of interest determination module 506 is specifically used to: match the content publishing account identifier with the point of interest account information in the point of interest database; if the match is successful, the matched point of interest is determined as the target point of interest corresponding to the target offline merchant.

[0223] Optionally, the point of interest determination module 506 is specifically used to: retrieve a set of candidate points of interest from the point of interest database based on the query vector corresponding to the first information; filter the set of candidate points of interest, and determine the target point of interest from the set of candidate points of interest.

[0224] Optionally, the information providing device is further configured to: acquire user historical interaction information of the target point of interest; the user historical interaction information includes at least one of historical access information and historical evaluation information for the target point of interest; and determine the overall recommendation tag of the target offline merchant based on the user historical interaction information.

[0225] Optionally, the information providing device is further configured to: acquire individual historical interaction information of a target user; the individual historical interaction information includes at least one of the target user's individual access information and individual evaluation information for each offline merchant; determine individual recommendation tags for the target offline merchant based on the user's historical interaction information and the individual historical interaction information; the individual recommendation tags are used to be displayed in association with the target offline merchant in the user interface shown to the target user.

[0226] Optionally, the first information specifically includes information from a vertical field, which may include the field of anime and manga.

[0227] It is understood that the modules mentioned above refer to computer programs or program segments used to perform one or more specific functions. Furthermore, the distinction between these modules does not imply that the actual program code must also be separate.

[0228] For ease of description, the above devices are described by dividing them into various modules or units based on their functions. Of course, when implementing one or more of these specifications, the functions of each module or unit can be implemented in the same or different software and / or hardware, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0229] The above is an illustrative scheme of an information providing device according to this embodiment. It should be noted that the technical solution of this information providing device and the technical solution of the information providing method described above belong to the same concept. For details not described in detail in the technical solution of the information providing device, please refer to the description of the technical solution of the information providing method described above.

[0230] Based on the same idea, this specification also provides devices corresponding to the above methods in its embodiments.

[0231] Figure 6 This is a structural block diagram of a computing device provided as an embodiment of this specification.

[0232] The computing device 600 includes:

[0233] Memory 610 and processor 620;

[0234] The memory 610 is used to store computer programs / instructions, and the processor 620 is used to execute the computer programs / instructions, which, when executed by the processor 620, implement the steps of the information providing method.

[0235] Specifically, the components of the computing device 600 include, but are not limited to, a memory 610 and a processor 620. The processor 620 is connected to the memory 610 via a bus 630, and the database 650 is used to store data.

[0236] The computing device 600 also includes an access device 640, which enables the computing device 600 to communicate via one or more networks 640. Examples of these networks include Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or combinations of communication networks such as the Internet. The access device 640 may include one or more of any type of wired or wireless network interface (such as a network interface card (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, a Wi-MAX (Worldwide Interoperability for Microwave Access) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, a Near Field Communication (NFC) interface, and so on.

[0237] In one embodiment of this specification, the above-described components of the computing device 600 and Figure 6 Other components, not shown, can also be connected to each other, for example, via a bus. It should be understood that... Figure 6 The block diagram of the computing device shown is for illustrative purposes only and is not intended to limit the scope of this application. Those skilled in the art can add or replace other components as needed.

[0238] The computing device 600 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (such as tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (such as smartphones), wearable computing devices (such as smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or personal computers (PCs). The computing device 600 can also be a mobile or stationary server.

[0239] The processor 620 implements the information provision method when executing the computer instructions.

[0240] The above is an illustrative scheme of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the information providing method described above belong to the same concept. For details not described in detail in the technical solution of the computing device, please refer to the description of the technical solution of the information providing method described above.

[0241] An embodiment of this specification also provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the steps of the information provision method as described above.

[0242] The above is an illustrative embodiment of a computer-readable storage medium. It should be noted that the technical solution of this storage medium and the technical solution of the information providing method described above belong to the same concept. Details not described in detail in the technical solution of the storage medium can be found in the description of the technical solution of the information providing method described above.

[0243] An embodiment of this specification also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the above-described information provision method.

[0244] The above is an illustrative scheme of a computer program product according to this embodiment. It should be noted that the technical solution of this computer program product and the technical solution of the information providing method described above belong to the same concept. For details not described in detail in the technical solution of the computer program product, please refer to the description of the technical solution of the information providing method described above.

[0245] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the apparatus and device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments. The apparatus, device and method provided in the embodiments of this specification are corresponding to each other, and therefore the apparatus and device also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the corresponding apparatus and device will not be repeated here.

[0246] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0247] In the 1990s, improvements to a technology could be clearly distinguished as either hardware improvements (e.g., improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the methodology). However, with technological advancements, many methodological improvements today can be considered direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved methodology into the hardware circuit. Therefore, it cannot be said that a methodological improvement cannot be implemented using hardware physical modules. For example, a Programmable Logic Device (PLD) (such as a Field Programmable Gate Array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. Designers can program a digital system themselves to "integrate" it onto a PLD, without needing chip manufacturers to design and manufacture dedicated integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing integrated circuit chips, this programming is mostly implemented using "logic compiler" software. Similar to the software compiler used in program development, the original code before compilation must be written in a specific programming language, called a Hardware Description Language (HDL). There are many HDLs, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, and RHDL (Ruby Hardware Description Language). Currently, the most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should understand that by simply performing some logic programming on the method flow using one of these hardware description languages ​​and programming it into an integrated circuit, the hardware circuit implementing the logical method flow can be easily obtained.

[0248] The controller can be implemented in any suitable manner. For example, it can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code form, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included therein for implementing various functions can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0249] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0250] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0251] Those skilled in the art will understand that one or more embodiments of this specification can be provided as a method, system, or computer program product. Therefore, the invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0252] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0253] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0254] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0255] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0256] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0257] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital character versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0258] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0259] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An information provision method, comprising: Get the first information published on external content platforms; The first piece of information is used to describe the offline services provided at the target offline merchant; The first information is processed in a structured manner to generate the second information; The second information includes the time, location, and service description of the offline service; Based on at least some of the content in the first information and the second information, the target offline merchants are associated and matched with the points of interest in the point of interest database to determine the target points of interest corresponding to the target offline merchants; The second information is provided for display in the user interface corresponding to the target point of interest.

2. The method as described in claim 1, wherein obtaining the first information published on an external content platform specifically includes: Based on the content publishing account of the target offline merchant, the first information published by the content publishing account is collected from the external content platform; And / or, Based on specified vertical domain keywords, the first information matching the vertical domain keywords is collected from the external content platform; the vertical domain keywords include intellectual property keywords.

3. The method as described in claim 1, wherein, The first information includes at least one of information title, information content, and first multimedia information; wherein the type of the first multimedia information includes at least one of static image, dynamic image, and video.

4. The method of claim 3, wherein, The second information also includes second multimedia information; the second multimedia information includes at least a portion of the information in the first multimedia information.

5. The method as described in claim 1, wherein the step of structuring the first information to generate the second information specifically includes: The first information is input into the first large language model to obtain the time information and the location information extracted by the first large language model from the first information; or, Based on preset structured information extraction rules, the time information and the location information are extracted from the first information information.

6. The method of claim 1, further comprising: Input the first information into the first language model to obtain the benefit point label of the offline service determined by the first language model based on the first information; The benefit point label is used to indicate the benefits that the offline service can provide.

7. The method of claim 6, wherein, The rights and benefits tags are used to display in the information details page of the second information; and / or, The benefit point tag is used as a basis for filtering target information carrying the benefit point tag on the information collection page; the target information includes the second information.

8. The method as described in claim 1, wherein the service introduction information specifically includes title information and body information; the step of performing structured processing on the first information to generate the second information specifically includes: The first information is input into a second language model to obtain the title information and the body text information generated by the second language model based on the first information; wherein, the second language model is trained based on first training data of the vertical domain to which the first information belongs; the first training data includes sample information and sample tags for the sample information; the sample tags include title tags and body text tags corresponding to the sample information.

9. The method of claim 8, further comprising: The first information is input into the second language model to obtain a service type label generated by the second language model based on the first information; the service type label is used to represent the service form of the offline service. The sample labels in the first training data also include service type sample labels corresponding to the sample information.

10. The method of claim 9, wherein, The service type tag is used to display in the information details page of the second information; and / or, The service type tag is used as a basis for filtering target information carrying the service type tag on the information collection page; the target information includes the second information.

11. The method of claim 1, further comprising: The first information is input into the third language model to obtain an intellectual property tag generated by the third language model based on the first information for marking the second information; The intellectual property label is used to indicate the intellectual property objects involved in the offline service; The intellectual property objects include at least one of intellectual property works and virtual characters; The third language model is trained based on the second training data in the vertical domain to which the first information belongs; The second training data includes at least one of text-based data and image-text data; the text-based data includes at least one of conceptual text and relational text, wherein the conceptual text is used to explain a specified intellectual property object, and the relational text is used to explain the relationship between multiple intellectual property objects; the image-text data includes a specified image and descriptive information about the intellectual property objects in the specified image.

12. The method of claim 11, wherein, The intellectual property label is used to display on the information details page of the second information; and / or, The intellectual property label is used as a basis for filtering target information carrying the intellectual property label on the information collection page; the target information includes the second information.

13. The method of claim 11, further comprising: To obtain the target users' preferences for different intellectual property objects; Based on the preference level and the intellectual property tag, determine the matching degree between the second information and the target user; Based on the matching degree, the priority of pushing the second information in the information collection page for the target user is determined.

14. The method of claim 1, further comprising, before performing structuring processing on the first information to generate the second information: Determine whether the first information matches a preset filtering rule; the preset filtering rule includes at least one of the following: the content length is less than a preset length threshold, the content is unrelated to the target vertical domain, or the service location is not within the target service area; The step of structuring the first information to generate the second information specifically includes: The first information that does not match the preset filtering rules is processed in a structured manner to generate the second information.

15. The method as described in claim 1, wherein the point of interest database includes point of interest address information for each point of interest; the step of associating and matching the target offline merchant with the points of interest in the point of interest database based on at least a portion of the first information and the second information to determine the target point of interest corresponding to the target offline merchant specifically includes: The location information is matched with the address information of the points of interest in the point of interest database; If a match is successful, the matched point of interest will be determined as the target point of interest corresponding to the target offline merchant.

16. The method as described in claim 1, wherein the point of interest database includes point of interest account information for each point of interest; the first information includes a content publishing account identifier; and the step of associating and matching the target offline merchant with the points of interest in the point of interest database based on at least a portion of the first information and the second information to determine the target point of interest corresponding to the target offline merchant specifically includes: The content publishing account identifier is matched with the interest point account information in the interest point database; If a match is successful, the matched point of interest will be determined as the target point of interest corresponding to the target offline merchant.

17. The method of claim 1, wherein the step of associating and matching the target offline merchant with points of interest in the point of interest database based on at least a portion of the first information and the second information, to determine the target point of interest corresponding to the target offline merchant, specifically includes: Based on the query vector corresponding to the first information, a set of candidate points of interest is recalled from the point of interest database; The candidate interest point set is filtered to determine the target interest point.

18. The method of claim 1, further comprising: Obtain the user's historical interaction information for the target point of interest; The user's historical interaction information includes at least one of historical access information and historical evaluation information for the target point of interest. Based on the user's historical interaction information, the overall recommendation tags for the target offline merchants are determined.

19. The method of claim 18, further comprising: Obtain individual historical interaction information of the target user; the individual historical interaction information includes at least one of the target user's individual access information and individual evaluation information for each offline merchant; Based on the user's historical interaction information and the individual's historical interaction information, determine the individual recommendation tags for the target offline merchants; The individual recommendation tags are used to associate and display the target offline merchants in the user interface shown to the target user.

20. The method according to any one of claims 1 to 19, wherein the first information specifically includes information in a vertical field, and the vertical field includes the two-dimensional field.

21. An information providing device, comprising: The first information acquisition module is used to acquire first information published on external content platforms; The first piece of information is used to describe the offline services provided at the target offline merchant; The second information generation module is used to perform structured processing on the first information to generate the second information; The second information includes the time, location, and service description of the offline service; The point of interest determination module is used to associate and match the target offline merchant with the points of interest in the point of interest database based on at least part of the first information and the second information, and determine the target point of interest corresponding to the target offline merchant; The second information providing module is used to provide the second information for display in the user interface corresponding to the target point of interest.

22. A computing device, comprising: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 to 20.

Images