Information recommendation method and device based on artificial intelligence, equipment and storage medium

Abstract

The application belongs to the field of artificial intelligence and the field of financial technology, and relates to an information recommendation method based on artificial intelligence, which comprises the following steps: in the process of business communication work, obtaining a specified communication theme of a target user and a target user; obtaining a seat image of the target user and a user image of the target user; obtaining historical communication timing feature information of the target user; obtaining initial theme feature information; calculating the specified communication theme, the seat image, the user image and the initial theme feature information based on a double-tower model to determine target theme feature information; obtaining target dialogue and pushing the target dialogue to the target user. The application also provides an information recommendation device based on artificial intelligence, a computer device and a storage medium. In addition, the target dialogue of the application can be stored in a blockchain. The application can be applied to a communication theme recommendation scene in the financial field, and the use of the double-tower model effectively improves the recommendation accuracy of the dialogue.

Description

Technical Field

[0001] This application relates to the fields of artificial intelligence development technology and fintech, and in particular to information recommendation methods, devices, computer equipment and storage media based on artificial intelligence. Background Technology

[0002] In financial and insurance service operations, agents typically need to proactively communicate with customers and complete the core communication process to improve the success rate of tasks. For example, in the process of providing car insurance telephone service, agents need to conduct multiple rounds of telephone communication with users to gradually discover their needs and encourage them to purchase, renew, or obtain a satisfactory claims experience.

[0003] Existing communication workflows often provide agents with a set of generic scripts to facilitate better communication with customers and ensure smooth workflow completion. However, this approach requires agents to pre-determine the necessary communication topic based on their personal experience and spend considerable time selecting appropriate scripts from the generic scripts. Because the accuracy of the communication topic determined based on personal experience is often low, the scripts selected by agents from the generic scripts may be inaccurate, thus impacting communication efficiency between agents and customers. Summary of the Invention

[0004] The purpose of this application is to propose an information recommendation method, apparatus, computer device, and storage medium based on artificial intelligence, in order to solve the problem that existing processing methods for providing general scripts have low accuracy in determining the current required communication topics based on personal experience, which easily leads to inaccurate script recommendations by agents from general scripts, thereby affecting the communication efficiency between agents and customers.

[0005] To address the aforementioned technical problems, this application provides an information recommendation method based on artificial intelligence, employing the following technical solution:

[0006] During the business communication process between the target agent and the target user, the current designated communication topic between the target agent and the target user is obtained;

[0007] Obtain the seat profile of the target agent and the user profile of the target user;

[0008] Obtain historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions;

[0009] The initial topic feature information to be recommended is obtained from a preset topic information database; wherein, the number of initial topic feature information includes multiple types;

[0010] Based on the preset dual-tower model, the specified communication topic, the agent profile, the user profile, and the initial topic feature information are calculated and processed to determine the target topic feature information from all the initial topic feature information;

[0011] Obtain the target dialogue corresponding to the target topic feature information;

[0012] The target topic feature information and the target script are pushed to the target agent.

[0013] Furthermore, the step of calculating and processing the specified communication topic, the agent profile, the user profile, and the initial topic feature information based on the preset dual-tower model, and determining the target topic feature information from all the initial topic feature information, specifically includes:

[0014] The specified communication topic, the agent profile, and the user profile are input into the user-side tower model in the dual-tower model. Vector extraction is performed on the specified communication topic, agent profile, and user profile through the user-side tower model to obtain the corresponding first feature vector.

[0015] Each of the initial theme feature information is input into the item side tower model in the dual tower model. The item side tower model is used to extract vectors from each of the initial theme feature information to obtain multiple corresponding second feature vectors.

[0016] The similarity between the first feature vector and each of the second feature vectors is calculated through the interoperability layer in the dual-tower model.

[0017] Based on the similarity, target topic feature information is determined from all the initial topic feature information.

[0018] Furthermore, the step of calculating the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model specifically includes:

[0019] Obtain the preset similarity calculation strategy;

[0020] The target similarity calculation strategy is determined from all the aforementioned similarity calculation strategies;

[0021] Based on the target similarity calculation strategy, the similarity between the first feature vector and each of the second feature vectors is calculated through the interoperability layer in the dual-tower model.

[0022] Furthermore, the step of determining the target topic feature information from all the initial topic feature information based on the similarity specifically includes:

[0023] Compare all the aforementioned similarities numerically, and select the specified similarity with the highest numerical value from all the aforementioned similarities;

[0024] Obtain the third feature vector corresponding to the specified similarity from all the second feature vectors;

[0025] Obtain the specified initial topic feature information corresponding to the third feature vector from all the initial topic feature information;

[0026] The specified initial topic feature information is used as the target topic feature information.

[0027] Furthermore, the step of obtaining the target dialogue corresponding to the target topic feature information specifically includes:

[0028] Call the preset script library;

[0029] Extract the first dialogue script corresponding to the target topic feature information from the dialogue script library;

[0030] Obtain the communication success rate for each of the first dialogue phrases;

[0031] Select a second script from all the first scripts, whose success rate is greater than a preset success rate threshold;

[0032] The second script is used as the target script.

[0033] Furthermore, the step of obtaining historical communication time sequence feature information corresponding to the target user specifically includes:

[0034] Call the preset communication information database;

[0035] Obtain the target user information of the target user;

[0036] Based on the target user information, obtain the associated call information corresponding to the business communication task;

[0037] Filter the specified communication information corresponding to the associated call information from the communication information database;

[0038] The specified communication information is used as the historical communication time sequence feature information.

[0039] Furthermore, the step of obtaining the initial topic feature information to be recommended from the preset topic information database specifically includes:

[0040] Call the aforementioned topic information database;

[0041] Obtain the specified job type for the business communication task;

[0042] Select the specified topic feature information corresponding to the specified job type from the topic information database;

[0043] The specified topic feature information is used as the initial topic feature information.

[0044] To address the aforementioned technical problems, this application also provides an information recommendation device based on artificial intelligence, employing the following technical solution:

[0045] The first acquisition module is used to acquire the current designated communication topic between the target agent and the target user during the business communication process between the target agent and the target user.

[0046] The second acquisition module is used to acquire the seat profile of the target agent and the user profile of the target user.

[0047] The third acquisition module is used to acquire historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions;

[0048] The fourth acquisition module is used to acquire initial topic feature information to be recommended from a preset topic information database; wherein, the number of initial topic feature information includes multiple types;

[0049] The determination module is used to perform calculations on the specified communication topic, the agent profile, the user profile, and the initial topic feature information based on a preset dual-tower model, and determine the target topic feature information from all the initial topic feature information;

[0050] The fifth acquisition module is used to acquire the target dialogue corresponding to the target topic feature information;

[0051] The push module is used to push the target topic feature information and the target script to the target agent.

[0052] To address the aforementioned technical problems, this application also provides a computer device that employs the following technical solution:

[0053] During the business communication process between the target agent and the target user, the current designated communication topic between the target agent and the target user is obtained;

[0054] Obtain the seat profile of the target agent and the user profile of the target user;

[0055] Obtain historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions;

[0056] The initial topic feature information to be recommended is obtained from a preset topic information database; wherein, the number of initial topic feature information includes multiple types;

[0057] Based on the preset dual-tower model, the specified communication topic, the agent profile, the user profile, and the initial topic feature information are calculated and processed to determine the target topic feature information from all the initial topic feature information;

[0058] Obtain the target dialogue corresponding to the target topic feature information;

[0059] The target topic feature information and the target script are pushed to the target agent.

[0060] To address the aforementioned technical problems, this application also provides a computer-readable storage medium, employing the technical solution described below:

[0061] During the business communication process between the target agent and the target user, the current designated communication topic between the target agent and the target user is obtained;

[0062] Obtain the seat profile of the target agent and the user profile of the target user;

[0063] Obtain historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions;

[0064] The initial topic feature information to be recommended is obtained from a preset topic information database; wherein, the number of initial topic feature information includes multiple types;

[0065] Based on the preset dual-tower model, the specified communication topic, the agent profile, the user profile, and the initial topic feature information are calculated and processed to determine the target topic feature information from all the initial topic feature information;

[0066] Obtain the target dialogue corresponding to the target topic feature information;

[0067] The target topic feature information and the target script are pushed to the target agent.

[0068] Compared with the prior art, the embodiments of this application have the following main advantages:

[0069] In this embodiment of the application, during the business communication process between the target agent and the target user, the following steps are taken: First, the current designated communication topic between the target agent and the target user is obtained; then, the agent profile of the target agent and the user profile of the target user are obtained; next, historical communication time sequence feature information corresponding to the target user is obtained; and initial topic feature information to be recommended is obtained from a preset topic information database; wherein, the number of initial topic feature information includes multiple; subsequently, based on a preset dual-tower model, the designated communication topic, the agent profile, the user profile, and the initial topic feature information are calculated and processed to determine the target topic feature information from all the initial topic feature information; further, the target script corresponding to the target topic feature information is obtained; finally, the target topic feature information and the target script are pushed to the target agent. In the process of business communication between a target agent and a target user, this application, after obtaining the current designated communication topic, the agent's profile, the user's profile, historical communication sequence characteristics, and initial topic characteristics to be recommended, uses a preset dual-tower model to calculate and process the designated communication topic, agent profile, user profile, and initial topic characteristics. This allows for the rapid and accurate determination of the target topic characteristics from all the initial topic characteristics, ensuring the accuracy of the generated target topic characteristics. Furthermore, by obtaining the target script corresponding to the target topic characteristics, and pushing the target topic characteristics and the target script to the target agent, the accuracy of the target script recommendation is effectively guaranteed. This facilitates subsequent business communication between the target agent and the target user using the target script, thereby improving the efficiency of business communication. Attached Figure Description

[0070] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0071] Figure 1 This is an exemplary system architecture diagram to which this application can be applied;

[0072] Figure 2 A flowchart of an embodiment of the AI-based information recommendation method according to this application;

[0073] Figure 3This is a schematic diagram of a structure of an embodiment of the artificial intelligence-based information recommendation device according to this application;

[0074] Figure 4 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0075] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0076] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0077] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0078] like Figure 1 As shown, system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105. Network 104 serves as the medium for providing communication links between terminal devices 101, 102, and 103 and server 105. Network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0079] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, and 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0080] Terminal devices 101, 102, and 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 players (Moving Picture Experts Group Audio Layer IV), laptops, and desktop computers, etc.

[0081] Server 105 can be a server that provides various services, such as a backend server that supports the pages displayed on terminal devices 101, 102, and 103.

[0082] It should be noted that the information recommendation method based on artificial intelligence provided in this application is generally executed by a server / terminal device, and correspondingly, the information recommendation device based on artificial intelligence is generally set in the server / terminal device.

[0083] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0084] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0085] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0086] Continue to refer to Figure 2The flowchart illustrates an embodiment of the AI-based information recommendation method according to this application. The order of steps in the flowchart can be changed, and some steps can be omitted, depending on different needs. The AI-based information recommendation method provided in this application can be applied to any scenario requiring communication topic recommendation and script recommendation, and thus can be applied to products in these scenarios, such as communication topic recommendation and script recommendation in the financial insurance field. The AI-based information recommendation method includes the following steps:

[0087] Step S201: During the business communication between the target agent and the target user, obtain the current designated communication topic between the target agent and the target user.

[0088] In this embodiment, the information recommendation method based on artificial intelligence runs on an electronic device (e.g., Figure 1 The server / terminal device shown can obtain the specified communication topic through wired or wireless connection. It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra-wideband) connections, and other currently known or future-developed wireless connection methods. In the communication operation scenario of insurance claims services, business communication operations may include auto insurance business communication operations, life insurance business communication operations, accident insurance business communication operations, etc. The aforementioned specified communication topic refers to the topic of communication between the agent and the customer. For example, after confirming the user's purchase intention, the process will proceed to the quotation stage, and this quotation stage becomes the current communication topic. For example, the communication topic may also include confirming the customer's claims service requests, purchase intention, introducing product highlights, comparing product prices, and confirming customer information.

[0089] Step S202: Obtain the seat profile of the target agent and the user profile of the target user.

[0090] In this embodiment, a profile database storing profile data is pre-built. By obtaining the target agent's agent information, the corresponding agent profile can be retrieved from the profile database. Similarly, by obtaining the target user's user information, the target user's user profile can be retrieved from the profile database. The agent's profile may include at least the agent's rank, region, and gender. The user's profile may include at least the user's gender, age, communication focus, and wealth level.

[0091] Step S203: Obtain historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions.

[0092] In this embodiment, the specific implementation process of obtaining the historical communication time sequence feature information corresponding to the target user will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0093] Step S204: Obtain initial topic feature information to be recommended from a preset topic information database; wherein, the number of initial topic feature information includes multiple.

[0094] In this embodiment, the specific implementation process of obtaining the initial topic feature information to be recommended from the preset topic information database will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0095] Step S205: Based on the preset dual-tower model, calculate and process the specified communication topic, the agent profile, the user profile, and the initial topic feature information, and determine the target topic feature information from all the initial topic feature information.

[0096] In this embodiment, the aforementioned dual-tower model is a model constructed based on a user-side tower model, an interoperability layer, and a user-side tower model. The model is divided into two parts: a user-side tower model and an item-side tower model. The interoperability layer then combines these two parts to generate a final prediction score, thereby achieving the information recommendation function. The specific implementation process of calculating and processing the specified communication topic, the agent profile, the user profile, and the initial topic feature information based on the preset dual-tower model to determine the target topic feature information from all the initial topic feature information will be further described in detail in subsequent embodiments of this application, and will not be elaborated upon here.

[0097] Step S206: Obtain the target dialogue corresponding to the target topic feature information.

[0098] In this embodiment, the specific implementation process of obtaining the target speech corresponding to the target topic feature information will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0099] Step S207: Push the target topic feature information and the target script to the target agent.

[0100] In this embodiment, the target agent's work terminal's communication information can be obtained, and then, based on this communication information, the target topic feature information and the target script can be pushed to the work interface of the target agent's work terminal.

[0101] In the process of business communication between a target agent and a target user, this application first obtains the current designated communication topic between the target agent and the target user; then, it obtains the agent profile of the target agent and the user profile of the target user; next, it obtains the historical communication time sequence feature information corresponding to the target user; and it obtains the initial topic feature information to be recommended from a preset topic information database; wherein, the number of initial topic feature information includes multiple; subsequently, based on a preset dual-tower model, it calculates and processes the designated communication topic, the agent profile, the user profile, and the initial topic feature information to determine the target topic feature information from all the initial topic feature information; further, it obtains the target script corresponding to the target topic feature information; finally, it pushes the target topic feature information and the target script to the target agent. In the process of business communication between a target agent and a target user, this application, after obtaining the current designated communication topic, the agent's profile, the user's profile, historical communication sequence characteristics, and initial topic characteristics to be recommended, uses a preset dual-tower model to calculate and process the designated communication topic, agent profile, user profile, and initial topic characteristics. This allows for the rapid and accurate determination of the target topic characteristics from all the initial topic characteristics, ensuring the accuracy of the generated target topic characteristics. Furthermore, by obtaining the target script corresponding to the target topic characteristics, and pushing the target topic characteristics and the target script to the target agent, the accuracy of the target script recommendation is effectively guaranteed. This facilitates subsequent business communication between the target agent and the target user using the target script, thereby improving the efficiency of business communication.

[0102] In some alternative implementations, step S205 includes the following steps:

[0103] The specified communication topic, the agent profile, and the user profile are input into the user-side tower model in the dual-tower model. Vector extraction is performed on the specified communication topic, the agent profile, and the user profile through the user-side tower model to obtain the corresponding first feature vector.

[0104] In this embodiment, the user-side tower model specifically refers to the left User tower in the dual-tower model. By using the left User tower, vector extraction is performed on the specified communication topic, the agent profile, and the user profile to output UserEmbedding, which yields the aforementioned first feature vector.

[0105] Each of the initial theme feature information is input into the item side tower model in the dual tower model. The item side tower model is used to extract vectors from each of the initial theme feature information to obtain multiple corresponding second feature vectors.

[0106] In this embodiment, the above-mentioned item side tower model specifically refers to the right Item tower in the dual-tower model. By using this side Item tower, vector extraction is performed on the initial topic feature information to output Item Embedding, thus obtaining multiple second feature vectors.

[0107] The similarity between the first feature vector and each of the second feature vectors is calculated through the interoperability layer in the dual-tower model.

[0108] In this embodiment, the similarity between the obtained first feature vector and multiple second feature vectors can be calculated by using them as inputs to the interoperability layer in the dual-tower model. The specific implementation process of calculating the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model will be described in further detail in subsequent embodiments of this application, and will not be elaborated upon here.

[0109] Based on the similarity, target topic feature information is determined from all the initial topic feature information.

[0110] In this embodiment, the specific implementation process of determining the target topic feature information from all the initial topic feature information based on the similarity will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0111] This application inputs the specified communication topic, the agent profile, and the user profile into the user-side tower model of the dual-tower model. The user-side tower model extracts vectors from the specified communication topic, the agent profile, and the user profile to obtain corresponding first feature vectors. Then, it inputs each initial topic feature information into the item-side tower model of the dual-tower model. The item-side tower model extracts vectors from each initial topic feature information to obtain multiple corresponding second feature vectors. The interoperability layer in the dual-tower model then calculates the similarity between the first feature vectors and each of the second feature vectors. Based on this similarity, the target topic feature information is determined from all the initial topic feature information. By using the dual-tower model to calculate and process the specified communication topic, the agent profile, the user profile, and the initial topic feature information, this application can quickly and accurately determine the target topic feature information from all the initial topic feature information, improving the generation efficiency of the target topic feature information and ensuring the accuracy of the generated target topic feature information.

[0112] In some optional implementations of this embodiment, calculating the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model includes the following steps:

[0113] Obtain the preset similarity calculation strategy.

[0114] In this embodiment, the above-mentioned similarity calculation strategies include dot product operation, cosine similarity calculation, MLP structure, and other strategies.

[0115] The target similarity calculation strategy is determined from all the aforementioned similarity calculation strategies.

[0116] In this embodiment, the method for determining the target similarity calculation strategy is not specifically limited and can be set according to actual usage requirements. Specifically, the processing efficiency of various similarity calculation strategies can be obtained, and then the similarity calculation strategy with the highest processing efficiency can be selected from all similarity calculation strategies to be used as the target similarity calculation strategy.

[0117] Based on the target similarity calculation strategy, the similarity between the first feature vector and each of the second feature vectors is calculated through the interoperability layer in the dual-tower model.

[0118] In this embodiment, the similarity between the first feature vector and each of the second feature vectors can be calculated using the target similarity calculation strategy through the interoperability layer in the dual-tower model, so as to improve the calculation efficiency of similarity and increase the acquisition rate of similarity.

[0119] This application obtains a preset similarity calculation strategy; then determines a target similarity calculation strategy from all the aforementioned similarity calculation strategies; subsequently, based on the target similarity calculation strategy, it calculates the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model. This application determines the target similarity calculation strategy from the obtained similarity calculation strategies, and then uses the target similarity calculation strategy to calculate the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model, thereby achieving rapid completion of the similarity calculation process between the first feature vector and each of the second feature vectors, improving the similarity acquisition rate and the intelligence of the similarity calculation.

[0120] In some optional implementations, determining the target topic feature information from all the initial topic feature information based on the similarity includes the following steps:

[0121] Numerical comparisons are performed on all the aforementioned similarities, and the specified similarity with the highest numerical value is selected from all the aforementioned similarities.

[0122] In this embodiment, the number of specified similarities may include one or more.

[0123] Obtain the third feature vector corresponding to the specified similarity from all the second feature vectors.

[0124] In this embodiment, a third feature vector corresponding to the specified similarity can be obtained from all the second feature vectors based on the correspondence between feature vectors and similarity. The number of the third feature vectors may include one or more.

[0125] Obtain the specified initial topic feature information corresponding to the third feature vector from all the initial topic feature information.

[0126] In this embodiment, after the third feature vector is determined, the topic feature information corresponding to the third feature vector can be obtained from all the initial topic feature information to obtain the specified initial topic feature information.

[0127] The specified initial topic feature information is used as the target topic feature information.

[0128] This application compares the numerical similarities of all the aforementioned similarities and selects the highest-valued specified similarity from them. Then, it obtains a third feature vector corresponding to the specified similarity from all the second feature vectors. Subsequently, it obtains specified initial topic feature information corresponding to the third feature vector from all the initial topic feature information and uses this specified initial topic feature information as the target topic feature information. This application obtains the third feature vector corresponding to the highest-valued specified similarity from all the second feature vectors, and then obtains specified initial topic feature information corresponding to the third feature vector from all the initial topic feature information to serve as the final target topic feature information. By using the topic feature information corresponding to the third feature vector with the highest-valued specified similarity as the target topic feature information, the obtained target topic feature information is the most suitable topic for the current context, ensuring the accuracy of the generated target topic feature information.

[0129] In some alternative implementations, step S206 includes the following steps:

[0130] Call the preset script library.

[0131] In this embodiment, the aforementioned script library is a pre-built database that stores communication scripts applicable to various dialogue topics, based on actual agent business communication records.

[0132] Extract the first dialogue script corresponding to the target topic feature information from the dialogue script library.

[0133] In this embodiment, a target dialogue topic matching the target topic feature information can be retrieved from the dialogue script library, and then the dialogue script corresponding to the target dialogue topic can be extracted from the dialogue script library to obtain a first dialogue script corresponding to the target topic feature information.

[0134] Obtain the communication success rate for each of the first dialogue scripts.

[0135] In this embodiment, the communication success rate of each first script can be extracted by obtaining communication statistics data of all first scripts and then extracting information from the communication statistics data.

[0136] Select a second script from all the first scripts, choosing one with a communication success rate greater than a preset success rate threshold.

[0137] In this embodiment, the value of the success rate threshold is not specifically limited and can be set according to actual usage requirements.

[0138] The second script is used as the target script.

[0139] This application calls a preset script library; then extracts a first script corresponding to the target topic feature information from the script library; subsequently, it obtains the communication success rate of each first script; and then selects a second script with a communication success rate greater than a preset success rate threshold from all the first scripts, and uses the second script as the target script. This application uses a script library to extract the first script corresponding to the target topic feature information, and then intelligently selects a second script with a communication success rate greater than a preset success rate threshold from all the first scripts to use as the target script. Because the generated target script has a high communication success rate, the accuracy of script recommendation is improved, which is beneficial for subsequent target agents to use the target script to communicate with target users and effectively improve the success rate of business transactions.

[0140] In some optional implementations of this embodiment, step S203 includes the following steps:

[0141] Call the preset communication information database.

[0142] In this embodiment, the aforementioned communication information database is a pre-built database that stores call information generated by agents during communication with users, as well as communication information matching the call information. Typically, a communication process involves multiple rounds of business communication. Communication information includes at least the communication topic and the user's intent. For example, in an insurance claims service communication scenario, the communication topic may include confirming the customer's claims service requests, purchase intentions, introducing product highlights, comparing product prices, and confirming customer information. User intent refers to the objections the user focuses on during communication with the agent; for example, user intent may include price concerns or comparisons with other companies.

[0143] Obtain the target user information of the target user.

[0144] In this embodiment, the aforementioned target user information is the target user's identity information, such as the target user's name.

[0145] Based on the target user information, obtain the associated call information corresponding to the business communication task.

[0146] In this embodiment, the aforementioned associated call information refers to previous call information of the target user related to the business communication operation. For example, a communication operation process may include multiple rounds of communication, and the associated call information may refer to the previous rounds of calls related to the communication operation process, such as multiple call information that occurred a few days or 10 days ago.

[0147] Select the specified communication information corresponding to the associated call information from the communication information database.

[0148] In this embodiment, the communication information database can be queried using associated call information to filter out specific communication information corresponding to the associated call information. The specific communication information refers to the communication topic and user intent corresponding to the associated call information.

[0149] The specified communication information is used as the historical communication time sequence feature information.

[0150] This application retrieves historical communication time-series feature information by calling a pre-defined communication information database, obtaining target user information of the target user, and then, based on the target user information, obtaining associated call information corresponding to the business communication task. Subsequently, it filters out specific communication information corresponding to the associated call information from the communication information database and uses this specific communication information as the historical communication time-series feature information. This application achieves rapid and accurate retrieval of historical communication time-series feature information by querying the communication information database using the target user information to obtain associated call information corresponding to the business communication task, and then filtering out specific communication information corresponding to the associated call information from the communication information database. This improves the efficiency of obtaining historical communication time-series feature information and ensures the accuracy of the obtained historical communication time-series feature information.

[0151] In some optional implementations of this embodiment, step S204 includes the following steps:

[0152] Call the aforementioned topic information database.

[0153] In this embodiment, the aforementioned topic information database is a database constructed based on actual dialogue topic collection needs, which stores multiple job types and topic feature information corresponding to each job type.

[0154] Obtain the specified job type for the business communication task.

[0155] In this embodiment, keyword analysis can be performed on business communication tasks to determine their specific task type. In the business scenario of insurance business communication tasks, task types may include auto insurance claims, life insurance claims, personal insurance claims, and so on.

[0156] Select the specified topic feature information corresponding to the specified job type from the topic information database.

[0157] In this embodiment, information can be queried from the topic information database using the specified job type to retrieve the specified topic feature information corresponding to the specified job type.

[0158] The specified topic feature information is used as the initial topic feature information.

[0159] This application achieves rapid and accurate acquisition of initial topic feature information by calling the topic information database, obtaining the specified job type of the business communication task, and then filtering the specified topic feature information corresponding to the specified job type from the topic information database, using the specified topic feature information as the initial topic feature information. This improves the efficiency of initial topic feature information acquisition and ensures the accuracy of the acquired initial topic feature information.

[0160] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0161] It should be emphasized that, in order to further ensure the privacy and security of the aforementioned target scripts, the target scripts can also be stored in a blockchain node.

[0162] The blockchain referred to in this application is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0163] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0164] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0165] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware with computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When executed, the program can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0166] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0167] Further reference Figure 3 As a response to the above Figure 2 To implement the method shown, this application provides an embodiment of an information recommendation device based on artificial intelligence, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0168] like Figure 3 As shown, the artificial intelligence-based information recommendation device 300 described in this embodiment includes: a first acquisition module 301, a second acquisition module 302, a third acquisition module 303, a fourth acquisition module 304, a determination module 305, a fifth acquisition module 306, and a push module 307. Wherein:

[0169] The first acquisition module 301 is used to acquire the current designated communication topic between the target agent and the target user during the business communication process between the target agent and the target user.

[0170] The second acquisition module 302 is used to acquire the seat profile of the target agent and the user profile of the target user.

[0171] The third acquisition module 303 is used to acquire historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions;

[0172] The fourth acquisition module 304 is used to acquire initial topic feature information to be recommended from a preset topic information database; wherein, the number of initial topic feature information includes multiple;

[0173] The determination module 305 is used to perform calculations on the specified communication topic, the agent profile, the user profile, and the initial topic feature information based on a preset dual-tower model, and determine the target topic feature information from all the initial topic feature information;

[0174] The fifth acquisition module 306 is used to acquire the target speech corresponding to the target topic feature information;

[0175] The push module 307 is used to push the target topic feature information and the target script to the target agent.

[0176] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based information recommendation method in the aforementioned implementation method, and will not be repeated here.

[0177] In some optional implementations of this embodiment, the determining module 305 includes:

[0178] The first extraction submodule is used to input the specified communication topic, the agent profile, and the user profile into the user side tower model in the dual-tower model, and to extract vectors from the specified communication topic, agent profile, and user profile through the user side tower model to obtain the corresponding first feature vector;

[0179] The second extraction submodule is used to input the initial theme feature information into the item side tower model in the dual tower model, and extract vectors from the initial theme feature information through the item side tower model to obtain multiple corresponding second feature vectors.

[0180] The computational submodule is used to calculate the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model;

[0181] The first determining submodule is used to determine the target topic feature information from all the initial topic feature information based on the similarity.

[0182] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based information recommendation method in the aforementioned implementation method, and will not be repeated here.

[0183] In some optional implementations of this embodiment, the computation submodule includes:

[0184] The first acquisition unit is used to acquire a preset similarity calculation strategy;

[0185] The first determining unit is used to determine the target similarity calculation strategy from all the similarity calculation strategies;

[0186] The computing unit is used to calculate the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model based on the target similarity calculation strategy.

[0187] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based information recommendation method in the aforementioned implementation method, and will not be repeated here.

[0188] In some optional implementations of this embodiment, the first determining submodule includes:

[0189] A comparison unit is used to perform numerical comparisons on all the similarities and select the specified similarity with the highest value from all the similarities.

[0190] The second acquisition unit is used to acquire a third feature vector corresponding to the specified similarity from all the second feature vectors;

[0191] The third acquisition unit is used to acquire the specified initial topic feature information corresponding to the third feature vector from all the initial topic feature information;

[0192] The second determining unit is used to use the specified initial topic feature information as the target topic feature information.

[0193] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based information recommendation method in the aforementioned implementation method, and will not be repeated here.

[0194] In some optional implementations of this embodiment, the fifth acquisition module 306 includes:

[0195] The first submodule is used to call the preset script library;

[0196] The third extraction submodule is used to extract the first dialogue script corresponding to the target topic feature information from the dialogue script library;

[0197] The first acquisition submodule is used to acquire the communication success rate of each of the first dialogue scripts;

[0198] The first filtering submodule is used to filter out second scripts from all the first scripts, which have a communication success rate greater than a preset success rate threshold.

[0199] The second determining submodule is used to use the second script as the target script.

[0200] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based information recommendation method in the aforementioned implementation method, and will not be repeated here.

[0201] In some optional implementations of this embodiment, the third acquisition module 303 includes:

[0202] The second calling submodule is used to call the preset communication information database;

[0203] The second acquisition submodule is used to acquire the target user information of the target user.

[0204] The third acquisition submodule is used to acquire associated call information corresponding to the business communication operation based on the target user information;

[0205] The second filtering submodule is used to filter out specified communication information corresponding to the associated call information from the communication information database;

[0206] The third determining submodule is used to use the specified communication information as the historical communication time sequence feature information.

[0207] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based information recommendation method in the aforementioned implementation method, and will not be repeated here.

[0208] In some optional implementations of this embodiment, the fourth acquisition module 304 includes:

[0209] The third calling submodule is used to call the topic information database;

[0210] The fourth acquisition submodule is used to acquire the specified job type of the business communication job;

[0211] The third filtering submodule is used to filter out the specified topic feature information corresponding to the specified job type from the topic information database;

[0212] The fourth determining submodule is used to use the specified topic feature information as the initial topic feature information.

[0213] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based information recommendation method in the aforementioned implementation method, and will not be repeated here.

[0214] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed] for details. Figure 4 , Figure 4This is a basic structural block diagram of the computer device in this embodiment.

[0215] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected via a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0216] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0217] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 4. Of course, the memory 41 may also include both the internal storage unit and its external storage device of the computer device 4. In this embodiment, the memory 41 is typically used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions for information recommendation methods based on artificial intelligence. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.

[0218] In some embodiments, the processor 42 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to execute computer-readable instructions stored in the memory 41 or to process data, for example, to execute computer-readable instructions for the AI-based information recommendation method.

[0219] The network interface 43 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 4 and other electronic devices.

[0220] Compared with the prior art, the embodiments of this application have the following main advantages:

[0221] In this embodiment, during business communication between the target agent and the target user, after obtaining the current designated communication topic, the agent's profile, the user's profile, historical communication sequence characteristics, and initial topic characteristics to be recommended, a preset dual-tower model is used to calculate and process the designated communication topic, agent profile, user profile, and initial topic characteristics. This allows for the rapid and accurate determination of the target topic characteristics from all the initial topic characteristics, ensuring the accuracy of the generated target topic characteristics. Furthermore, by obtaining the target script corresponding to the target topic characteristics and pushing the target script to the target agent, the accuracy of the target script recommendation is effectively guaranteed. This facilitates subsequent business communication between the target agent and the target user using the target script, thereby improving the efficiency of business communication.

[0222] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the artificial intelligence-based information recommendation method described above.

[0223] Compared with the prior art, the embodiments of this application have the following main advantages:

[0224] In this embodiment, during business communication between the target agent and the target user, after obtaining the current designated communication topic, the agent's profile, the user's profile, historical communication sequence characteristics, and initial topic characteristics to be recommended, a preset dual-tower model is used to calculate and process the designated communication topic, agent profile, user profile, and initial topic characteristics. This allows for the rapid and accurate determination of the target topic characteristics from all the initial topic characteristics, ensuring the accuracy of the generated target topic characteristics. Furthermore, by obtaining the target script corresponding to the target topic characteristics and pushing the target script to the target agent, the accuracy of the target script recommendation is effectively guaranteed. This facilitates subsequent business communication between the target agent and the target user using the target script, thereby improving the efficiency of business communication.

[0225] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0226] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. An information recommendation method based on artificial intelligence, characterized in that, Includes the following steps: During the business communication process between the target agent and the target user, the current designated communication topic between the target agent and the target user is obtained; Obtain the seat profile of the target agent and the user profile of the target user; Obtain historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions; The initial topic feature information to be recommended is obtained from a preset topic information database; wherein, the number of initial topic feature information includes multiple types; Based on the preset dual-tower model, the specified communication topic, the agent profile, the user profile, and the initial topic feature information are calculated and processed to determine the target topic feature information from all the initial topic feature information; Obtain the target dialogue corresponding to the target topic feature information; The target topic feature information and the target script are pushed to the target agent; The step of calculating and processing the specified communication topic, the agent profile, the user profile, and the initial topic feature information based on the preset dual-tower model, and determining the target topic feature information from all the initial topic feature information, specifically includes: The specified communication topic, the agent profile, and the user profile are input into the user-side tower model in the dual-tower model. Vector extraction is performed on the specified communication topic, agent profile, and user profile through the user-side tower model to obtain the corresponding first feature vector. Each of the initial theme feature information is input into the item side tower model in the dual tower model. The item side tower model is used to extract vectors from each of the initial theme feature information to obtain multiple corresponding second feature vectors. The similarity between the first feature vector and each of the second feature vectors is calculated through the interoperability layer in the dual-tower model. Based on the similarity, target topic feature information is determined from all the initial topic feature information; Specifically, the user-side tower model refers to the left User tower in the dual-tower model. By using the left User tower, vectors are extracted from the specified communication topic, the agent profile, and the user profile to output User Embedded, thus obtaining the aforementioned first feature vector. The item-side tower model refers to the right Item tower in the dual-tower model. By using the item tower, vectors are extracted from the feature information of each initial topic to output Item Embedded, thus obtaining multiple aforementioned second feature vectors.

2. The information recommendation method based on artificial intelligence according to claim 1, characterized in that, The step of calculating the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model specifically includes: Obtain the preset similarity calculation strategy; The target similarity calculation strategy is determined from all the aforementioned similarity calculation strategies; Based on the target similarity calculation strategy, the similarity between the first feature vector and each of the second feature vectors is calculated through the interoperability layer in the dual-tower model.

3. The information recommendation method based on artificial intelligence according to claim 1, characterized in that, The step of determining the target topic feature information from all the initial topic feature information based on the similarity specifically includes: Compare all the aforementioned similarities numerically, and select the specified similarity with the highest numerical value from all the aforementioned similarities; Obtain the third feature vector corresponding to the specified similarity from all the second feature vectors; Obtain the specified initial topic feature information corresponding to the third feature vector from all the initial topic feature information; The specified initial topic feature information is used as the target topic feature information.

4. The information recommendation method based on artificial intelligence according to claim 1, characterized in that, The step of obtaining the target speech corresponding to the target topic feature information specifically includes: Call the preset script library; Extract the first dialogue script corresponding to the target topic feature information from the dialogue script library; Obtain the communication success rate for each of the first dialogue scripts; Select a second script from all the first scripts, whose success rate is greater than a preset success rate threshold; The second script is used as the target script.

5. The information recommendation method based on artificial intelligence according to claim 1, characterized in that, The step of obtaining historical communication time sequence feature information corresponding to the target user specifically includes: Call the preset communication information database; Obtain the target user information of the target user; Based on the target user information, obtain the associated call information corresponding to the business communication task; Filter the specified communication information corresponding to the associated call information from the communication information database; The specified communication information is used as the historical communication time sequence feature information.

6. The information recommendation method based on artificial intelligence according to claim 1, characterized in that, The step of obtaining the initial topic feature information to be recommended from the preset topic information database specifically includes: Call the aforementioned topic information database; Obtain the specified job type for the business communication task; Select the specified topic feature information corresponding to the specified job type from the topic information database; The specified topic feature information is used as the initial topic feature information.

7. An information recommendation device based on artificial intelligence, characterized in that, include: The first acquisition module is used to acquire the current designated communication topic between the target agent and the target user during the business communication process between the target agent and the target user. The second acquisition module is used to acquire the seat profile of the target agent and the user profile of the target user. The third acquisition module is used to acquire historical communication time sequence feature information corresponding to the target user; wherein, the historical communication time sequence feature information includes at least historical topics and historical intentions; The fourth acquisition module is used to acquire initial topic feature information to be recommended from a preset topic information database; wherein, the number of initial topic feature information includes multiple types; The determination module is used to perform calculations on the specified communication topic, the agent profile, the user profile, and the initial topic feature information based on a preset dual-tower model, and determine the target topic feature information from all the initial topic feature information; The fifth acquisition module is used to acquire the target dialogue corresponding to the target topic feature information; The push module is used to push the target topic feature information and the target script to the target agent; The determining module includes: The first extraction submodule is used to input the specified communication topic, the agent profile, and the user profile into the user side tower model in the dual-tower model, and to extract vectors from the specified communication topic, agent profile, and user profile through the user side tower model to obtain the corresponding first feature vector; The second extraction submodule is used to input the initial theme feature information into the item side tower model in the dual tower model, and extract vectors from the initial theme feature information through the item side tower model to obtain multiple corresponding second feature vectors. The computational submodule is used to calculate the similarity between the first feature vector and each of the second feature vectors through the interoperability layer in the dual-tower model; The first determining submodule is used to determine the target topic feature information from all the initial topic feature information based on the similarity. Specifically, the user-side tower model refers to the left User tower in the dual-tower model. By using the left User tower, vectors are extracted from the specified communication topic, the agent profile, and the user profile to output User Embedded, thus obtaining the aforementioned first feature vector. The item-side tower model refers to the right Item tower in the dual-tower model. By using the item tower, vectors are extracted from the feature information of each initial topic to output Item Embedded, thus obtaining multiple aforementioned second feature vectors.

8. A computer device comprising a memory and a processor, the memory storing computer-readable instructions, wherein the processor, when executing the computer-readable instructions, implements the steps of the information recommendation method based on artificial intelligence as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the information recommendation method based on artificial intelligence as described in any one of claims 1 to 6.

Images