Information retrieval method and device, equipment and storage medium

Abstract

The invention relates to an information retrieval method and device, equipment and a storage medium, relates to the field of terminals, and solves the problem of low information acquisition efficiency during user retrieval. The specific scheme comprises the steps that a terminal receives to-be-retrieved information input by a user and displays a first abstract corresponding to the to-be-retrieved information; under the condition that the user retrieves the historical retrieval information matched with the to-be-retrieved information in the past, the first abstract is used for indicating the difference between the retrieval result corresponding to the to-be-retrieved information and the retrieval result corresponding to the similar historical retrieval information. Under the condition that the user does not retrieve historical retrieval information similar to the to-be-retrieved information in the past, the first abstract is used for indicating differences of retrieval results corresponding to the to-be-retrieved information in two different time periods. The first abstract indicates the difference of the two retrieval results, and the difference of the multiple retrieval results can be displayed through the abstract under the condition that the user retrieves for multiple times, so that the efficiency of acquiring new information by the user is improved.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of terminals, and in particular, to an information retrieval method, apparatus, device, and storage medium. Background Art

[0002] In the prior art, when a user uses a search engine for retrieval, the search engine will return corresponding retrieval results according to the retrieval information (query) input by the user. However, there is often a head effect in the retrieval results, that is, the content on the first few pages is mostly hot content or newly released content, and some of this hot content or newly released content may be content that the user has already seen, resulting in low retrieval efficiency for the user using the search engine.

[0003] Moreover, in the prior art, when a user uses a search engine for retrieval, if the same or similar retrieval information is retrieved multiple times, there are often a large number of identical or similar contents in the retrieval results returned by the search engine in different retrievals. Even if there may be newly released content in the retrieval results returned by the search engine multiple times, the newly released content may only be a copy or summary of the existing content. This makes it difficult for the user to obtain new information when retrieving the same or similar retrieval information multiple times.

[0004] It can be seen that in the case where the user has a retrieval requirement, there is a lack of a method in the prior art that enables the user to obtain new information with high efficiency. Summary of the Invention

[0005] Embodiments of the present application provide an information retrieval method, apparatus, device, and storage medium, which solve the problem of low efficiency in obtaining information during user retrieval.

[0006] According to a first aspect of the embodiments of the present application, an information retrieval method is provided. The terminal receives the information to be retrieved input by the user and displays a first summary corresponding to the information to be retrieved. Wherein, in the case where the user has retrieved historical retrieval information that matches the information to be retrieved in the past, the first summary is used to indicate the differences between the retrieval results corresponding to the information to be retrieved and the retrieval results corresponding to the similar historical retrieval information. In the case where the user has not retrieved historical retrieval information similar to the information to be retrieved in the past, the first summary is used to indicate the differences between the retrieval results corresponding to the information to be retrieved in two different time periods. In the present application, by generating a summary of the information to be retrieved, the user does not need to read a large amount of repetitive content, improving the efficiency of the user in obtaining information. And the first summary is used to indicate the differences between the two retrieval results. In the case where the user retrieves similar problems multiple times, the differences between the multiple retrieval results can be shown to the user through the summary, and the user can obtain new information faster, improving the efficiency of the user in obtaining new information.

[0007] In some embodiments, the process of displaying the first abstract includes: displaying a second abstract and a first control, where the second abstract is the abstract of the first search result; in the presence of second search information, the first control is used to prompt the user to obtain the abstract of the search result of the updated first search information, and the first abstract is displayed after receiving an operation; or, in the absence of second search information, the first control is used to prompt the user to subscribe to the search result of the first search information, and the first abstract is displayed after a preset time of receiving an operation. Receive the user's operation on the first control, and in response to this operation, display the first abstract. Through the above method, the user can obtain the first abstract by operating, giving the user more optional options and improving the user experience.

[0008] In some embodiments, the process of displaying the second abstract and the first control includes: sending the first search information to the server; receiving the second abstract and the first indication information from the server, where the first indication information is used to indicate whether there is second search information. Display the second abstract and display the first control according to the first indication information. Through the first indication information, the terminal can determine the content that the first control should display, so as to bring a better user experience to the user through the first control.

[0009] In some embodiments, the above process of displaying the first abstract includes: in response to an operation, sending second indication information to the server; receiving the first abstract from the server; where, in the presence of second search information, the second indication information is used to indicate to update the abstract of the search result of the first search information; or, in the absence of second search information, the second indication information is used to indicate subscribing to the search result of the first search information. Through the second indication information, the server can generate the corresponding first abstract, so that the terminal can obtain the corresponding first abstract.

[0010] In some embodiments, the second abstract can also be generated by the terminal. Specifically, generate the second abstract according to the first search information; the second abstract is the abstract of the first search result; according to the first search information, obtain the interest point matching result, where the interest point matching result is used to indicate whether there is second search information with the same interest point as the first search information, and display the first control according to the interest point matching result. In this way, the interest point matching can be performed locally to protect user privacy.

[0011] In some embodiments, the first summary may also be generated by the terminal. Specifically, obtain the first search result corresponding to the first search information; receive the user's operation on the first control; in the case where the first control is used to prompt an update of the summary of the search result of the first search information, obtain the second search result corresponding to the second search information, and generate the first summary according to the first search result and the second search result; or, in the case where the first control is used to prompt a subscription to the search result of the first search information, obtain the second search result corresponding to the first search information after a preset time, and generate the first summary according to the first search result and the second search result. In this way, the terminal can meet the requirements in different scenarios and improve the user experience.

[0012] In some embodiments, the process of the terminal generating the first summary may specifically include: using the first search result, the second search result, and the first data as inputs, and generating the first summary through an artificial intelligence model; where the first data includes at least one of the following: the first search information, the user's feedback information, and the point of interest. In the case where there is second search information, the feedback information is used to indicate the interaction information of the user with the page corresponding to the second search result; or, in the case where there is no second search information, the feedback information is the interaction information of the user on the page corresponding to the first search result. Generating the first summary through multiple pieces of information in this way can make the generated first summary more accurate and improve the user experience.

[0013] According to the second aspect of the embodiments of the present application, there is provided an information retrieval method applied to a server. The method includes: receiving the first search information from the terminal; sending the first summary corresponding to the first search information to the terminal; where the first summary is used to indicate the difference between the first search result and the second search result, and the first search result is the search result of the first search information retrieved at the first time; where, in the case where there is second search information, the second search result is the search result of the second search information; or, in the case where there is no second search information, the second search result is the search result of the first search information retrieved at the second time, the second time is later than the first time, and the second search information is the search information retrieved by the user before receiving the first search information and having the same point of interest as the first search information.

[0014] In some embodiments, before sending the first summary corresponding to the first search information to the terminal, the method further includes: generating a second summary according to the first search information; the second summary is the summary of the first search result; obtaining an interest point matching result according to the first search information, and the interest point matching result is used to indicate whether there is second search information having the same point of interest as the first search information; sending the second summary to the terminal, and sending a first indication information to the terminal according to the interest point matching result, where the first indication information is used to indicate whether there is second search information.

[0015] In some embodiments, before sending the first summary corresponding to the first retrieval information to the terminal, the method further includes: obtaining a first retrieval result corresponding to the first retrieval information; receiving second indication information from the terminal; in a case where the second indication information is used to indicate an update of the summary of the retrieval result of the first retrieval information, obtaining a second retrieval result corresponding to second retrieval information, and generating a first summary according to the first retrieval result and the second retrieval result; or, in a case where the second indication information is used to indicate a subscription to the retrieval result of the first retrieval information, obtaining a second retrieval result corresponding to the first retrieval information after a preset time, and generating a first summary according to the first retrieval result and the second retrieval result.

[0016] In some embodiments, generating a first summary according to the first retrieval result and the second retrieval result includes: using the first retrieval result, the second retrieval result, and first data as inputs, and generating a first summary through an artificial intelligence model; where the first data includes at least one of the following: the first retrieval information, the user's feedback information, and points of interest, and in a case where there is second retrieval information, the feedback information is used to indicate the interaction information of the user with the page corresponding to the second retrieval result; or, in a case where there is no second retrieval information, the feedback information is the interaction information of the user with the page corresponding to the first retrieval result.

[0017] According to a third aspect of the embodiments of the present application, there is provided an electronic device, including a memory and a processor, the memory is used to store a computer program, and the processor is used to execute the computer program to implement the information retrieval method in the first aspect or the second aspect above.

[0018] According to a fourth aspect of the embodiments of the present application, there is provided a computer-readable storage medium, on which a computer program is stored, and the computer program is used to implement the information retrieval method in the first aspect or the second aspect above.

[0019] According to a fifth aspect of the embodiments of the present application, there is provided a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying the computer-readable code, and when the computer-readable code runs in an electronic device, the processor in the electronic device executes the information retrieval method described above.

[0020] According to a sixth aspect of the embodiments of the present application, there is provided an information retrieval device, and the device has a function of implementing the behavior of the terminal or the server in the method in the first aspect or the second aspect above. The function can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions, for example, a receiving module, a processing module, a sending module, and a display module.

[0021] It should be understood that for the technical solutions provided in the above second aspect, third aspect, fourth aspect, fifth aspect, and sixth aspect, their technical features can all correspond to the methods provided in the first aspect and its optional embodiments, so the beneficial effects that can be achieved are similar and will not be elaborated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Flowchart of an information retrieval method provided for the related art;

[0023] Figure 2 Flowchart of an abstract generation method provided for the related art;

[0024] Figure 3A Simplified schematic diagram of a system architecture provided for an embodiment of the present application;

[0025] Figure 3B Structural diagram of a terminal provided for an embodiment of the present application;

[0026] Figure 3C Structural diagram of a server provided for an embodiment of the present application;

[0027] Figure 4 Simplified schematic diagram of another system architecture provided for an embodiment of the present application;

[0028] Figure 5 Flowchart of an information retrieval method provided for an embodiment of the present application;

[0029] Figure 6 Schematic diagram of a retrieval interface provided for an embodiment of the present application;

[0030] Figure 7 Flowchart of another information retrieval method provided for an embodiment of the present application;

[0031] Figure 8 Schematic diagram of another information retrieval interface provided for an embodiment of the present application;

[0032] Figure 9 Schematic diagram of yet another information retrieval interface provided for an embodiment of the present application;

[0033] Figure 10 Schematic diagram of an information retrieval method provided for an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] In the related art, users can often retrieve the information they want to know through a search engine. The process of user retrieval in the related art is as follows: The terminal receives the retrieval information input by the user in the web page of the search engine or the input box of the client, and receives the user's retrieval operation, such as a click or touch operation on the retrieval button. The terminal sends a retrieval request to the server in response to the above retrieval operation, and the retrieval request carries the retrieval information input by the user. After receiving the user's retrieval information, the server retrieves the retrieval information through the search engine to obtain retrieval results. The retrieval results generally include multiple links. The server sends the retrieval results to the terminal, and the terminal displays the retrieval results after receiving them.

[0035] Since the retrieval results include multiple links, users often need to view the content in the links to obtain the information they want to know, and there may be a large amount of duplicate content among the multiple links, which makes the efficiency of users obtaining the information they want through the retrieval results relatively low. Moreover, since the retrieval results retrieved by the search engine often have a head effect, that is, the content on the first few pages is mostly hot results or newly released content, but the newly released content does not necessarily mean that there is content update. It may just be a copy or summary of the existing content, which makes these results may be the content that users have already seen or are not interested in, and may not necessarily meet the user's needs. All these make it possible for users to need to read a large amount of content to obtain the information they want to know, and the efficiency of users obtaining information through retrieval is relatively low.

[0036] In addition, during the retrieval process, users always expect to obtain new information rather than duplicate and already familiar information. However, in some cases, if users cannot obtain effective information from the current retrieval results, they may retrieve similar retrieval information again in another expression. In addition, in some cases, users may also retrieve the same or similar retrieval information at different time periods to obtain the latest information. And if users frequently retrieve the same or similar retrieval information, a large amount of duplicate content will also be included in the retrieval results of different retrievals, which also makes the efficiency of users obtaining new information relatively low.

[0037] It can be seen that in the information retrieval method of the related art, the efficiency of users obtaining information is relatively low, especially during multiple retrievals, users cannot quickly obtain new information.

[0038] In order to improve the retrieval efficiency, improve the efficiency of users obtaining information and the user experience, in the related art, the retrieval results can be summarized based on an artificial intelligence model for users to quickly browse and screen. For example, reference can be made to Related Art 1 and Related Art 2 introduced below.

[0039] In Related Art 1, such as Figure 1As shown, the server retrieves and analyzes the user's retrieval logs, establishes or updates the user interest model based on the concept clustering method, that is, determines the keywords of the content that the user is interested in from the content retrieved by the user in the past. And perform user interest and sentence similarity analysis between the retrieval results based on the user interest model to obtain the current retrieval results. Calculate the inverse document frequency (IDF) of the retrieval keywords. Calculate the information content of the sentences. Use the sentence similarity between the user interest model and the retrieval results, the IDF of the retrieval keywords, and the information content of the sentences, and adopt a personalized sentence scoring processing method to generate a personalized summary that meets the user. This enables in the related technology to generate a summary that better conforms to the interest preferences of different users according to the interest models of different users. For example, if the user has viewed more content related to planting in the past, then when the user retrieves a certain piece of information, the content related to planting will be extracted from the retrieval results, and a summary will be generated through the content related to planting.

[0040] Although the above method can generate a summary according to the user's interest and the current retrieval results through an artificial intelligence model, improving the user's information acquisition efficiency, the problem of duplicate content in the retrieval results of multiple retrievals is not solved in the above method. Moreover, in the above method, the user interest model is generated based on the content retrieved by the user in the past, which makes the generated summary more likely to be related to the content browsed by the user in the past, and may further exacerbate the problem of duplicate content in the retrieval results of multiple retrievals. This causes a lot of duplicate information in the summary seen by the user when retrieving the same or similar retrieval information multiple times, so that the efficiency of obtaining new information in the process of multiple retrievals by the user cannot be improved, affecting the user's retrieval experience.

[0041] In addition to the above related technology 1, another related technology 2 for generating summaries in the news push scenario will be introduced here. See Figure 2 , the server can determine the user's target news category according to the user's historical access data; match the target news category with a preset event graph to obtain multiple news related to the target news category; extract news summaries from each piece of news to obtain multiple news summaries; screen out some news summaries from the multiple news summaries and generate a briefing based on the some news summaries.

[0042] Although the related technology 2 is also a method for generating summaries, this method can only generate a briefing for the target news category corresponding to the preset event graph, cannot cover all the content that the user is interested in, lacks flexibility and generality, and cannot be applied to the search engine scenario of this application. In addition, similar to the method of related technology 1, this method does not discriminate against duplicate content, resulting in possible duplicate information in the briefings generated for the same user multiple times, making the efficiency of the user obtaining new information not high and affecting the user's usage experience.

[0043] It can be seen that there is a lack of a method for improving the efficiency of user retrieval and obtaining new information in the related art.

[0044] The present application proposes an information retrieval method, and the main content of this method includes: the terminal receives the information to be retrieved input by the user and displays the first summary corresponding to the information to be retrieved. Among them, in the case where the user has retrieved historical retrieval information that matches the information to be retrieved in the past, the first summary is used to indicate the differences between the retrieval results corresponding to the information to be retrieved and the retrieval results corresponding to the similar historical retrieval information. In the case where the user has not retrieved historical retrieval information similar to the information to be retrieved in the past, the first summary is used to indicate the differences between the retrieval results corresponding to the information to be retrieved in two different time periods.

[0045] In the present application, by generating a summary of the information to be retrieved, the user does not need to read a large amount of repetitive content, improving the efficiency of the user to obtain information. And the first summary is used to indicate the differences between the two retrieval results. In the case where the user retrieves similar problems multiple times, the differences between the multiple retrieval results can be shown to the user through the summary, and the user can obtain new information faster, improving the efficiency of the user to obtain new information.

[0046] It should be noted that in the technical solution of the present application, the collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information and other processes all comply with the provisions of relevant laws and regulations and do not violate public order and good customs. For example, in the technical solution of the present application, the processing of the user's personal information is carried out under the authorization of the user. This is uniformly stated here and will not be repeated below.

[0047] For the application scenario of the present application, the method of the present application can be executed by the terminal or can be implemented through the interaction between the terminal and the server. Next, taking the interaction between the terminal and the server to implement the method of the present application as an example, the system architecture of the embodiments of the present application will be described in detail.

[0048] Figure 3A Shown is a simplified schematic diagram of the system architecture to which the embodiments of the present application can be applied. As Figure 3A shown, the system architecture may include: at least one terminal 31 and at least one server 32.

[0049] Among them, the structure of the terminal 31 can be referred to Figure 3B As Figure 3B shown, the terminal 31 may include a processor 311, an internal memory 312, a communication module 313, a sensor module 314, and a display screen 315, etc. Among them, the communication module may include a first antenna, a second antenna, a radio frequency sub-module, a communication sub-module, a modem, and a baseband processor, etc., and the sensor module 314 may at least include a touch sensor.

[0050] The structure illustrated in the embodiments of the present invention does not limit the terminal 31. It may include more or fewer components than shown, or combine certain components, or split certain components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0051] The processor 311 may include one or more processing units. For example, the processor 311 may include an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units may be independent devices or integrated in the same processor.

[0052] The controller may be the decision-maker that commands the various components of the terminal 31 to work in coordination according to instructions. It is the nerve center and command center of the terminal 31. The controller generates operation control signals according to the instruction operation code and timing signals to complete the control of fetching and executing instructions.

[0053] A memory may also be provided in the processor 311 for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. It can save the instructions or data that the processor has just used or recycled. If the processor needs to use the instruction or data again, it can directly call it from the memory. This avoids repeated accesses, reduces the waiting time of the processor, and thus improves the efficiency of the system.

[0054] In some embodiments, the processor 311 may include an interface. The interface may include an inter-integrated circuit (I2C) interface, a mobile industry processor interface (MIPI), and / or a general-purpose input / output (GPIO) interface, etc.

[0055] The I2C interface is a bidirectional synchronous serial bus, including a serial data line (SDA) and a serial clock line (SCL). In some embodiments, the processor may include multiple groups of I2C buses. For example, the processor can be coupled to the touch sensor through the I2C interface, enabling the processor and the touch sensor to communicate through the I2C bus interface to implement the touch function of the terminal 31.

[0056] The MIPI interface can be used to connect the processor to peripheral devices such as the display screen 315 and the camera. The MIPI interface may include a display serial interface (DSI) of the display screen 315, etc. In some embodiments, the processor and the display screen 315 communicate through the DSI interface to implement the display function of the terminal 31.

[0057] The GPIO interface can be configured by software. The GPIO interface can be configured as a control signal or a data signal. In some embodiments, the GPIO interface can be used to connect the processor to the display screen 315, the communication module, the sensor, etc. The GPIO interface can also be configured as an I2C interface, a MIPI interface, etc.

[0058] The interface connection relationship between the modules illustrated in the embodiments of the present invention is only for illustrative purposes and does not constitute a structural limitation on the terminal 31. The terminal 31 can adopt different interface connection methods in the embodiments of the present invention, or a combination of multiple interface connection methods.

[0059] In some embodiments, the processor 311 can be used to execute the information retrieval method in the embodiments of the present application. The processor 311 can also communicate with other sensors or the display screen 315 through various interfaces. For example, it can communicate with the touch sensor through the I2C interface to obtain the information to be retrieved input by the user. For example, it can communicate with the display screen 315 through the GPIO interface, so that the first abstract can be displayed to the user through the display screen 315.

[0060] The internal memory 312 can be used to store computer-executable program code, and the executable program code includes instructions. The processor 311 executes various functional applications and data processing of the terminal 31 by running the instructions stored in the internal memory 312. The memory 312 can include a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a search engine application), etc. The data storage area can store data created during the use of the terminal 31. In addition, the memory 312 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one disk storage device, a flash memory device, other volatile solid-state storage devices, a universal flash storage (UFS), etc.

[0061] The wireless communication function of the terminal 31 can be implemented through the communication module 313, that is, the first antenna, the second antenna, the radio frequency sub-module, the communication sub-module, the modem, and the baseband processor, etc.

[0062] The first antenna and the second antenna are used to transmit and receive electromagnetic wave signals. Each antenna in the terminal 31 can be used to cover a single or multiple communication frequency bands. Different antennas can also be multiplexed to improve the utilization rate of the antennas. For example: the cellular network antenna can be multiplexed as a wireless local area network diversity antenna. In some embodiments, the antenna can be used in combination with a tuning switch.

[0063] The radio frequency sub-module can provide a communication processing module for wireless communication solutions including 2G / 3G / 4G / 5G, etc. applied to the terminal 31. It can include at least one filter, switch, power amplifier, low noise amplifier (LNA), etc. The radio frequency module receives electromagnetic waves through the first antenna, filters, amplifies, and processes the received electromagnetic waves, and transmits them to the modem for demodulation. The radio frequency module can also amplify the signal modulated by the modem and convert it into electromagnetic waves through the first antenna and radiate it out. In some embodiments, at least some functional modules of the radio frequency sub-module can be set in the processor 311. In some embodiments, at least some functional modules of the radio frequency sub-module and at least some modules of the processor 311 can be set in the same device.

[0064] The modem can include a modulator and a demodulator. The modulator is used to modulate the low-frequency baseband signal to be transmitted into a medium-high frequency signal. The demodulator is used to demodulate the received electromagnetic wave signal into a low-frequency baseband signal. Subsequently, the demodulator transmits the demodulated low-frequency baseband signal to the baseband processor for processing. In some embodiments, the modem can be an independent device. In some embodiments, the modem can be independent of the processor and set in the same device as the radio frequency module or other functional modules.

[0065] The communication sub-module can provide a communication processing module for applications on the terminal 31, including solutions for wireless communications such as wireless local area networks (WLAN), Bluetooth (BT), global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), infrared technology (IR), etc. The communication sub-module can be one or more devices integrating at least one communication processing module. The communication module receives electromagnetic waves via the second antenna, performs frequency modulation and filtering processing on the electromagnetic wave signals, and sends the processed signals to the processor. The communication sub-module can also receive the signals to be sent from the processor, perform frequency modulation and amplification on them, and convert them into electromagnetic waves through the second antenna and radiate them out.

[0066] In some embodiments, the first antenna of the terminal 31 is coupled to the radio frequency module, and the second antenna is coupled to the communication module. This enables the terminal 31 to communicate with the network and other devices through wireless communication technologies. The wireless communication technologies can include global system for mobile communications (GSM), general packet radio service (GPRS), code division multiple access (CDMA), wideband code division multiple access (WCDMA), time-division code division multiple access (TD-SCDMA), long term evolution (LTE), BT, WLAN, and / or IR technology, etc.

[0067] Through the communication module 313, the terminal can communicate with the server. For example, in some embodiments, the terminal can send the information to be retrieved to the server, and the terminal can receive the first summary, the first retrieval result, the second retrieval result, etc. sent by the server.

[0068] The touch sensor, also known as the "touch panel", can be disposed on the display screen 315. It is used to detect touch operations acting on or near it. The detected touch operations can be transmitted to the application processor to determine the type of touch event and provide corresponding visual output through the display screen 315.

[0069] Through the touch sensor, the terminal can obtain the information to be retrieved input by the user.

[0070] The terminal 31 realizes the display function through the GPU, the display screen 315, and the application processor, etc. The GPU is a microprocessor for image processing, connected to the display screen 315 and the application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. The processor 311 may include one or more GPUs, which execute program instructions to generate or change the display information.

[0071] The display screen 315 can be used to display the above-mentioned first abstract, first retrieval result, etc. The display screen 315 includes a display panel. The display panel can adopt LCD (liquid crystal display), OLED (organic light-emitting diode), active matrix organic light-emitting diode or active-matrix organic light-emitting diode (AMOLED), Miniled, MicroLed, Micro-oLed, quantum dot light-emitting diodes (QLED), etc. In some embodiments, the terminal 31 may include 1 or N display screens, where N is a positive integer greater than 1.

[0072] Through the display screen 315, the terminal can display the first abstract to the user, so that the user can obtain new information with higher efficiency.

[0073] The structure of the server can be referred to Figure 3C as Figure 3C shown, the server 32 may include: a processing module 321, a storage module 322, and a communication module 323.

[0074] The processing module 321 is the control center of the server. For example, the processing module 321 can be any one or a combination of a CPU, a GPU, a field-programmable gate array (FPGA), and an application-specific integrated circuit (ASIC).

[0075] The storage module 322 is used to store data. For example, the storage module 322 can be used to store the usage data of users and the data corresponding to multimedia content. The storage module 322 can also store computer instructions. The processing module 321 reads and executes the computer instructions stored in the storage module 322 to implement the functions of the server, such as obtaining retrieval results and generating a first summary, etc.

[0076] The communication module 323 is used for the server to communicate with the terminal or to communicate with other servers. In some embodiments, the system may include a first server and a second server. The first server is deployed with a search engine and is responsible for information retrieval, and the second server is responsible for generating a summary based on the retrieval result.

[0077] Figure 3B and Figure 3C The device structure of the application system of the present application is divided from a hardware perspective. Next, in combination with Figure 4 , the server of the application system of the present application is structurally divided from a functional perspective. And a brief description of the solution of the embodiment of the present application is given. It should be noted that Figure 4 is only an example and does not represent a limitation to the present application. For example, in addition to Figure 4 the solution shown in, it is easy to understand that Figure 4 multiple functions in can also be implemented by multiple servers. For example, the system includes a first server and a second server. The first server is deployed with a search engine module 41 for implementing the function of information retrieval, and the second server is deployed with a summary generation module 42 for implementing the function of generating a summary.

[0078] As Figure 4 shown, the server 32 includes a search engine module 41 and a summary generation module 42. The search engine module 41 is used to receive a request including the information to be retrieved from the terminal 31, retrieve the information to be retrieved, and return the retrieval result to the terminal. The search engine module 41 can also send the retrieval result to the summary generation module 42. The information to be retrieved is also the first retrieval information in the present application.

[0079] The summary generation module 42 may include a data sub-module 421 and a model sub-module 422.

[0080] The data sub-module includes a storage unit 4211 and a processing unit 4212. The storage unit 4211 is used to store the retrieval result sent by the search engine and the corresponding information to be retrieved, and the processing unit 4212 is used to process the data into a data format that can be processed by the model sub-module 422.

[0081] The processing unit 4212 can receive the retrieval results from the search engine module 41, and can also receive a request including the information to be retrieved sent by the terminal 31. According to this request, data processing is performed on the information to be retrieved and the retrieval results, and the data is processed into a data format that can be processed by the model sub-module 422. The processing unit 4212 can also communicate with the storage unit 4211, can store the processed data in the storage unit 4211, and the processing unit 4212 can also obtain the stored processed data from the storage unit 4211 and forward it to the model sub-module 422.

[0082] The storage unit 4211 can store data to facilitate the model sub-module 422 to determine whether there is historical retrieval information that matches the current information to be retrieved based on the data stored in the storage unit 4211. In some embodiments, the storage unit 4211 can store the historical retrieval information corresponding to each user and the corresponding retrieval results for each user, so as to generate a first summary in the subsequent process.

[0083] In some embodiments, the data processing performed by the processing unit 4212 can be to convert the data into an embedding format. That is, vectorization processing is performed on data in formats such as text.

[0084] The model sub-module 422 can obtain the data of the data sub-module 421 and determine whether there is historical retrieval information that matches the information to be retrieved, that is, whether there is a second retrieval information that has the same interest point as the information to be retrieved. The model sub-module 422 can also generate a first summary according to the data sent by the data sub-module 421, such as the first retrieval result and the second retrieval result sent.

[0085] In some embodiments, the model sub-module 422 can include a matching unit 4221 and a summary generation unit 4222.

[0086] The matching unit 4221 is used to determine whether there is a second retrieval information that matches the information to be retrieved. For example, the matching unit 4221 can determine whether the user has retrieved historical retrieval information with a similar or the same theme as the information to be retrieved through an artificial intelligence model according to the information to be retrieved sent by the processing unit 4212, the retrieval results corresponding to the information to be retrieved, the historical retrieval information, and the retrieval results corresponding to the historical retrieval information, that is, determine whether there is historical retrieval information that matches the information to be retrieved. For another example, the information to be retrieved and the historical retrieval information can be compared for similarity according to the information to be retrieved and the historical retrieval information sent by the processing unit 4212 to determine whether the user has retrieved similar historical retrieval information. The matching unit 4221 can also obtain the interest point corresponding to the information to be retrieved, or obtain the common interest point of the information to be retrieved and the historical retrieval information through the matching unit 4221.

[0087] The abstract generation unit 4222 is used to generate a first abstract. In some embodiments, the abstract generation unit 4222 may also generate a second abstract. The meanings and generation methods of the first abstract and the second abstract will be described in detail below and will not be elaborated here for the time being.

[0088] The abstract generation unit 4222 may generate a first abstract according to the first search result and the second search result, and the first abstract is used to indicate the difference between the first search result and the second search result. The first search result and the second search result are search results corresponding to the matching search information retrieved by the user in different time periods.

[0089] After the abstract generation unit 4222 generates the first abstract or the second abstract, it may send the generated first abstract or second abstract to the processing unit 4212, and the processing unit 4212 may forward the first abstract or second abstract to the terminal 31.

[0090] In this way, the terminal can display to the user the search results corresponding to the two matching pieces of information to be retrieved, enabling the user to quickly obtain new information and enhancing the user experience.

[0091] It should be noted that the system applied in this application is briefly described above through an example of interaction between the terminal and the server. This application can also be applied to the terminal, that is, the terminal completes the functions performed by the above-mentioned search engine module 41 and abstract generation module 42. The above embodiments do not represent a limitation to this application.

[0092] Next, an exemplary embodiment will be used to illustrate the information retrieval method shown in this application. As Figure 5 shown, Figure 5 is a flowchart of an information retrieval method shown in this application according to an exemplary embodiment, including the following steps:

[0093] Step 501, receive the information to be retrieved input by the user.

[0094] Figure 5The method shown is applied to a terminal. Herein, the user is the one who uses the terminal, and the information to be retrieved is also the query. The information to be retrieved is also referred to as the first retrieval information in this application. The information to be retrieved is the information input by the user and retrieved using a search engine, that is, the information input in the search box on the page corresponding to the search engine. The page can be the search box on the web page corresponding to the search engine of the terminal, the search box in the application corresponding to the search engine, or the search box on the negative first screen of the terminal. For the specific form of the information to be retrieved, the information to be retrieved can be a short sentence or phrase composed of text, and this short sentence or phrase is used to indicate the content that the user wants to retrieve. For example, the information to be retrieved can be "Chinese general large model", "Chinese general large model benchmark", etc. Taking the search on the negative first screen of the terminal as an example, as Figure 6 shown, the information to be retrieved is the information input in the search box 61.

[0095] Receiving the information to be retrieved input by the user can be obtaining the information to be retrieved input by the user when receiving the user's search operation. Specifically, the search operation can be the operation of the user on the search control, such as a click operation. Obtaining the information to be retrieved input by the user can be obtaining the information to be retrieved input by the user in the search box. Combining Figure 6 , as Figure 6 shown, the user can first input the information to be retrieved in the search box 61, and then click the search control 62 on the interface. The user clicking the search control 62 is the click operation of the user on the search control. After the terminal detects the click operation of the user on the search control 62, it can obtain the information to be retrieved input by the user in the search box 61.

[0096] Step 502, display a first summary according to the information to be retrieved; the first summary is used to indicate the difference between the first retrieval result and the second retrieval result, and the first retrieval result is the retrieval result of the information to be retrieved retrieved at the first time.

[0097] Among them, in the case where there is historical retrieval information that matches the information to be retrieved, the second retrieval result is the retrieval result of the above-mentioned historical retrieval information that matches the information to be retrieved; or, in the case where there is no historical retrieval information that matches the information to be retrieved, the second retrieval result is the retrieval result of the information to be retrieved retrieved at the second time, and the second time is later than the first time. The above-mentioned matching historical retrieval information is the retrieval information with the same interest point as the information to be retrieved retrieved by the user before receiving the first retrieval information.

[0098] The historical retrieval information that matches the information to be retrieved is also referred to as the second retrieval information in this application. Historical retrieval information is the retrieval information retrieved before the user inputs the information to be retrieved. For example, the information to be retrieved in step 501 may be input by the user at 12:00 on November 1st, and the historical retrieval information may be the information to be retrieved input by the user in the time period before 12:00 on November 1st.

[0099] The point of interest can be used to indicate the theme of the retrieval information. Then, the historical retrieval information that matches the information to be retrieved is the retrieval information with the same or similar theme as the information to be retrieved that the user retrieved in the past. For example, the information to be retrieved in step 501 may be "Chinese general large model" input by the user at 12:00 on November 1st, and the historical retrieval information that matches the information to be retrieved may be "Chinese general large model" input by the user at 12:00 on October 29th, or "Benchmark of Chinese general large model" input by the user at 12:00 on October 30th, etc. The specific method for determining whether there is historical retrieval information that matches the information to be retrieved can be found later in this text and will not be elaborated here for the time being.

[0100] The first abstract in this application can be generated by the terminal or by the server. When the first abstract is generated by the server, after step 501, the terminal can also send the information to be retrieved to the server so that the server can perform a retrieval based on the information to be retrieved and generate the first abstract based on the retrieval results. After the server generates the first abstract, the terminal can receive the first abstract from the server. Among them, sending the information to be retrieved to the server can be sending a retrieval request, and the retrieval request carries the information to be retrieved.

[0101] When the first abstract is generated by the terminal and the retrieval results are provided by the server, after step 501, the terminal can send the information to be retrieved to the server and receive the retrieval results from the server. The received retrieval results from the server can include the first retrieval results and the second retrieval results, that is, the server can store the user's past historical retrieval information and the corresponding retrieval results, and determine whether there is historical retrieval information that matches the information to be retrieved based on the stored historical retrieval information and the corresponding retrieval results, and send the corresponding first retrieval results and second retrieval results to the terminal. Then, the terminal can generate the first abstract based on the received first retrieval results and second retrieval results.

[0102] When the first abstract is generated by the server and the retrieval results are obtained by the terminal through retrieval, the terminal can store the user's retrieval history, that is, store the user's historical retrieval information and the corresponding retrieval results. After step 501, the terminal can determine whether there is historical retrieval information that matches the information to be retrieved based on the stored data, and send the information to be retrieved, the first retrieval result, the second retrieval result, etc. to the server, so that the server can generate the first abstract according to the received information. Saving the retrieval results in the terminal can ensure user privacy.

[0103] When the method is only implemented by the terminal, after step 501, the terminal can generate the first abstract.

[0104] Among them, the retrieval result refers to the content corresponding to multiple retrieved links. For example, if retrieving a certain information to be retrieved yields 5 links, then the retrieval result refers to the text content in the 5 links. In some cases, the content included in the retrieval result is relatively large, and the abstract can be generated only based on part of the content in the retrieval result to reduce the processing pressure. The corresponding first retrieval result and second retrieval result are also the retrieval results of the same or similar retrieval information retrieved by the user at different times.

[0105] Specifically, before the user inputs the information to be retrieved, if the user has retrieved historical retrieval information with a similar or the same theme as the information to be retrieved, that is, there is historical retrieval information that matches the information to be retrieved, the first retrieval result is the retrieval result corresponding to the information to be retrieved at the first time, and the second retrieval result is the retrieval result corresponding to the historical retrieval information that matches the information to be retrieved and has been retrieved by the user in the past. If the user has retrieved historical retrieval information that matches the information to be retrieved multiple times in the past, then the second retrieval result here can be the retrieval result corresponding to the historical retrieval information that matches and is closest to the first time.

[0106] In the above case, the first abstract is also used to show the user the update of the first retrieval result relative to the second retrieval result. Displaying the first abstract means that when the user retrieves the information to be retrieved, the update of the current retrieval result compared to the past retrieval results is shown in the results returned to the user, so that the user can quickly obtain different information in different retrievals, improving the efficiency of the user to obtain information, especially new information, and enhancing the user experience.

[0107] Before the user inputs the information to be retrieved, when the user has not retrieved any retrieval information that matches the information to be retrieved, the first retrieval result is the retrieval result corresponding to the information to be retrieved in step 501. The first retrieval result can be the retrieval result for the information to be retrieved generated immediately after step 501 is executed. The second retrieval result is the retrieval result for the information to be retrieved after a period of time after step 501 is executed, such as the retrieval result obtained for the information to be retrieved after one week or one month. That is, the second retrieval result is the retrieval result for the information to be retrieved obtained at a second time that is separated from the first time by a period of time.

[0108] In the above case, the first summary is used to show the user the update of the second retrieval result relative to the first retrieval result. Displaying the first summary, that is, after a period of time after the user retrieves the information to be retrieved, through push or other forms, show the user the update of the retrieval result corresponding to the information to be retrieved that has been retrieved in the past, so that the user can obtain new information faster. Through the way of active push, it can also enable the user to obtain new information without retrieval, improving the user experience.

[0109] In addition, for the specific method of generating the first summary, it can be generated through an artificial intelligence model, that is, through artificial intelligence generated content (AIGC), such as through large language models and other methods.

[0110] Specifically, the first retrieval result and the second retrieval result can be used as inputs, and through the artificial intelligence model, a first summary for indicating the difference between the first retrieval result and the second retrieval result can be obtained. In addition, the first retrieval result, the second retrieval result, and the first data can also be used as inputs, and through the artificial intelligence model, a first summary can be generated. The first data includes at least one of the following: the information to be retrieved, the user's feedback information, and the point of interest. When there is historical retrieval information that matches the information to be retrieved, the feedback information is used to indicate the interaction information of the user with the page corresponding to the second retrieval result; or, when there is no historical retrieval information that matches the information to be retrieved, the feedback information is the interaction information of the user with the page corresponding to the first retrieval result. If there is no historical retrieval information that matches the information to be retrieved, the point of interest input to the above artificial intelligence model is the point of interest corresponding to the information to be retrieved.

[0111] The feedback information of the user, that is, when the user retrieves the information to be retrieved corresponding to a certain point of interest, the interaction information of the user on the page corresponding to the retrieval result, such as the user clicks on or favorites a link, and the user hovers at a certain position, etc. In the case where the user has retrieved historical retrieval information that matches the information to be retrieved in the past, the above-mentioned interaction information is also the interaction information in the retrieval result of the historical retrieval information that matches the information to be retrieved in the past. In the case where the user has not retrieved historical retrieval information that matches the information to be retrieved in the past, the above-mentioned interaction information is also the interaction information on the page corresponding to the first retrieval result output after performing step 501.

[0112] These interaction information can be regarded as the content that the user is interested in. By inputting the feedback information of the user into the artificial intelligence model, the generated abstract can be more targeted, more in line with the user's interests, and improve the diversity of the abstracts corresponding to different users.

[0113] The point of interest is also the theme corresponding to the information to be retrieved. For example, if the information to be retrieved is "Chinese general large model", then the corresponding point of interest can be "Chinese large language model". In the case where the user has retrieved historical retrieval information that matches the information to be retrieved in the past, the point of interest is also the common theme of the matching historical retrieval information and the information to be retrieved. For example, if the user has retrieved "benchmark of Chinese general large model" in the past and the information to be retrieved is "Chinese general large model", then the common point of interest of the two is "Chinese large language model".

[0114] By inputting at least one of the point of interest and the information to be retrieved into the artificial intelligence model, the model can determine the content of the information to be retrieved, make the generated abstract closer to the information to be retrieved, make the first abstract more in line with the user's needs, and improve the user's experience.

[0115] In the case where the first data includes the information to be retrieved, the feedback information of the user, and the point of interest, more information can be provided for the artificial intelligence model, making the generated abstract more in line with the user's needs, thus bringing a better experience to the user.

[0116] In addition, if the user has retrieved historical retrieval information that matches the information to be retrieved in the past, and if the abstract of the historical retrieval information is stored, the abstract of the historical retrieval information can also be input into the artificial intelligence model, making the generated abstract more in line with the user's needs, thus bringing a better experience to the user.

[0117] The above display of the first abstract can be to display all the content of the first abstract. For example, in the case where there is historical retrieval information that matches the information to be retrieved, after receiving the information to be retrieved, the first abstract can be directly shown to the user.

[0118] Displaying the first abstract may also involve displaying partial content of the first abstract or the title of the first abstract. For example, when it is necessary to display the full text of the first abstract but the full text cannot be completely shown on one page, only partial content of the first abstract can be displayed. Another example is that in the case where the user has not retrieved historical retrieval information matching the information to be retrieved before step 501, after step 501 is executed for a period of time, a first abstract can be generated to indicate the difference between the second retrieval result of the information to be retrieved at the current second time and the second retrieval result of the information to be retrieved at the first time. In the above cases, the first abstract can be displayed by means of pushing. Then, displaying the first abstract can be to display the title of the first abstract or partial content of the first abstract in the push box. After the user clicks on the push box, the full content of the first abstract can be displayed.

[0119] Next, taking the Figure 4 system shown as an example, in combination with specific embodiments, an information retrieval method shown in the embodiments of the present application will be described. As Figure 7 shown, Figure 7 is a flowchart of an information retrieval method shown according to a specific embodiment of the present application, including the following steps:

[0120] Step 701, the terminal receives the information to be retrieved input by the user.

[0121] For the specific implementation manner of step 701, reference can be made to the description of step 501 above, and details will not be elaborated here.

[0122] Step 702, the terminal sends a retrieval request to the server according to the obtained information to be retrieved.

[0123] The retrieval request is also a request including the information to be retrieved. Correspondingly, the server needs to receive the information to be retrieved from the terminal, that is, the server can receive the retrieval request including the information to be retrieved from the terminal.

[0124] As Figure 4As shown, the server includes a search engine module 41 and a summary generation module 42. In some embodiments, the retrieval request may include a retrieval request sent to the search engine module 41 and a retrieval request sent to the summary generation module 42. The retrieval request sent to the search engine module 41 is used to trigger the search engine module 41 to retrieve the information to be retrieved and obtain a first retrieval result. The retrieval request sent to the summary generation module 42 is used to trigger the summary generation module 42 to match the information to be retrieved to determine whether there is historical retrieval information that matches the information to be retrieved, that is, to determine whether there is historical retrieval information with the same interest point as the information to be retrieved. In some other embodiments, the retrieval request may also be a retrieval request sent to the search engine module 41. After the search engine module 41 finishes retrieving, it sends the first retrieval result and the information to be retrieved to the summary generation module 42, so that the summary generation module 42 can determine whether the user has retrieved historical retrieval information that matches the information to be retrieved according to the first retrieval result, the information to be retrieved, the stored historical retrieval information, and the corresponding retrieval result.

[0125] Step 703, the server obtains a first retrieval result corresponding to the information to be retrieved.

[0126] Step 704, the server obtains an interest point matching result according to the first retrieval information.

[0127] Among them, the interest point matching result is used to indicate whether there is historical retrieval information with the same interest point as the information to be retrieved.

[0128] The first retrieval result is also the retrieval result corresponding to the information to be retrieved, that is, the retrieval result obtained by the above search engine module 41 according to the information to be retrieved. In step 703, the server can obtain a first retrieval result corresponding to the information to be retrieved.

[0129] The interest point matching result is also used to indicate whether the user has retrieved historical retrieval information that matches the information to be retrieved in the past. In addition, for the convenience of subsequent step processing, the interest point matching result may include historical retrieval information that matches the information to be retrieved or an identifier of the historical retrieval information, so as to facilitate the determination of the second retrieval result in subsequent steps. In the case of historical retrieval with the same interest point as the information to be retrieved, the interest point matching result may also include the common interest point of the information to be retrieved and the above historical retrieval information, and this interest point is used to facilitate the subsequent generation of the first summary. In addition, in the case of no historical retrieval with the same interest point as the information to be retrieved, the interest point matching result may also include an interest point, where the interest point refers to the interest point corresponding to the information to be retrieved, that is, the theme of the information to be retrieved, and this interest point can also facilitate the subsequent generation of the first summary.

[0130] Combined with Figure 4 , step 703 may beFigure 4 executed by the search engine module 41 in Figure 4 and the abstract generation module 42 in . In the above case, after step 703 is executed, the search engine module 41 may also send the first retrieval result to the abstract generation module 42.

[0131] Step 703 is executed in response to a retrieval request. Step 704 may be executed in response to a retrieval request sent by the terminal, or may be executed in response to the first retrieval result and the information to be retrieved sent by the search engine module 41.

[0132] For the specific implementation method of step 704, it may be to compare the similarity between the information to be retrieved and the historical retrieval information, determine the historical retrieval information with a similarity greater than the threshold with the information to be retrieved, and determine the historical retrieval information as the historical retrieval information having the same interest point as the information to be retrieved above. Or it may be to determine the cluster where the information to be retrieved is located through clustering, so as to determine the historical retrieval information meeting the above conditions.

[0133] In the case where the interest point matching result includes an interest point, the interest point may be obtained through an artificial intelligence model. For example, in the case where it is determined that the user has retrieved historical retrieval information with a similarity greater than the threshold with the information to be retrieved in the past, the historical retrieval information with a similarity greater than the threshold with the information to be retrieved and the information to be retrieved are input into the artificial intelligence model, and the above interest point is obtained through the artificial intelligence model.

[0134] The interest point matching result may also be obtained through an artificial intelligence model. For example, the information to be retrieved and the historical retrieval information may be used as inputs, and the interest point matching result is obtained through the artificial intelligence model. It may also be to use the information to be retrieved, the retrieval result corresponding to the information to be retrieved, the historical retrieval information, and the retrieval result corresponding to the historical retrieval information as inputs, and the interest point matching result is obtained through the artificial intelligence model. In the case where the interest point matching result includes an interest point, the interest point matching result including the interest point may be directly obtained through the artificial intelligence model.

[0135] The artificial intelligence model used to obtain the interest point matching result may be the same as the artificial intelligence model for generating the first abstract above. For example, it may also be obtained through a large language model.

[0136] Corresponding to Figure 4 the above steps for obtaining the interest point matching result may be Figure 4It is executed by the matching unit 4221 in []. Specifically, the processing unit 4212 can receive the first retrieval result and the information to be retrieved from the search engine module 41, or the processing unit 4212 can receive the information to be retrieved from the terminal and process the received data. The processing unit 4212 can also obtain the historical retrieval information stored in the storage unit 4211 and the retrieval results corresponding to the historical retrieval information. The processing unit 4212 can send the processed data, the obtained historical retrieval information, and the retrieval results corresponding to the historical retrieval information to the matching unit 4221. The matching unit 4221 can determine the point-of-interest matching result based on the data sent by the processing unit 4212.

[0137] The above specific implementation process is only an example and does not represent a limitation to this application.

[0138] Step 705, the server generates an initial summary according to the information to be retrieved.

[0139] The initial summary is also the second summary in this application, that is, the summary of the first retrieval result, which is used to summarize the first retrieval result. In addition, the basis for generating the initial summary can also be included in the initial summary to facilitate users to obtain information more conveniently.

[0140] Here, a specific example will be given to illustrate the initial summary. When the information to be retrieved is "Chinese general large model", the initial summary may include the following content:

[0141] A Chinese general large model refers to a large-scale pre-trained language model that can perform well in multiple Chinese natural language processing tasks, such as text classification, reading comprehension, dialogue, generation, etc.

[0142] Currently, multiple research institutions and enterprises in China have developed various Chinese general large models and tested and compared them on different evaluation benchmarks. Among them, SuperCLUB is a comprehensive evaluation benchmark specifically for Chinese general large models. It evaluates the models from single dimensions of basic capabilities, professional capabilities, and Chinese characteristic capabilities, and provides both human benchmarks and automated evaluation methods. According to the latest evaluation results of SuperCLUB, there are the following leading Chinese general large models currently: Model 1, Model 2, and Model 3. These models all have certain question-and-answer, dialogue, and generation capabilities, but there are also some limitations and deficiencies that need to be further improved.

[0143] Answer source: Web page 1, Web page 2.

[0144] The above answer source can be a link that can jump to the corresponding page after being clicked. It can be seen that through the initial summary, it can help users understand the information to be retrieved faster and improve the user experience.

[0145] For the method of generating an initial abstract, the server may use the first search result, the information to be retrieved, the point of interest, etc. as inputs, and generate the initial abstract through an artificial intelligence model. The initial abstract may also be generated by the methods in related technologies, and the present application does not limit the method of generating the initial abstract.

[0146] Combined with Figure 4 , step 705 above may be generated by the abstract generation unit 4222. Specifically, the abstract generation unit 4222 may obtain the matching result of the matching unit 4221, and obtain data such as the information to be retrieved and the first search result of the processing unit 4212, and generate the initial abstract according to the artificial intelligence model. It may also be that the matching unit 4221 sends the matching result and the input information of the matching unit 4221 to the abstract generation unit 4222, so that the abstract generation unit can generate the initial abstract according to the above information.

[0147] In some embodiments, the abstract generation unit 4222 needs to generate the above first abstract and second abstract, and it may be determined whether the abstract to be generated is the first abstract or the second abstract according to the input data of the abstract generation unit 4222. For example, the formats of the input data are different in two scenarios, or a flag bit is added to the input data to indicate whether the first abstract or the second abstract needs to be generated currently, and the abstract generation unit 4222 may generate the corresponding abstract according to the data in different formats.

[0148] Step 706, the server sends the first search result to the terminal.

[0149] Step 707, the server sends the initial abstract to the terminal, and sends the first indication information to the terminal according to the point of interest matching result.

[0150] The first indication information is used to indicate whether there is historical search information having the same point of interest as the information to be retrieved.

[0151] Correspondingly, the terminal may receive the first search result, the initial abstract and the first indication information. Correspondingly Figure 4 , step 706 may be executed by the search engine module 41, and step 707 may be executed by the abstract generation module 42.

[0152] In some embodiments, step 706 is optional, that is, only the initial abstract and the first indication information may be returned to the terminal. In other embodiments, the terminal may also obtain and display the first search result, so that the user can obtain more information from the first search result.

[0153] It should be noted that the execution order of the above steps 703 - 707 does not limit this application. For example, the server can first generate and send an initial abstract to the terminal, and then send the first search result to the terminal. It can first obtain the point of interest matching result and then obtain the first search result, etc.

[0154] Step 708, the terminal displays the initial abstract and the first control.

[0155] The initial abstract is the abstract of the first search result. In the case where there is historical search information that matches the information to be retrieved, the first control is used to prompt the user to obtain the abstract of the search result of the updated first search information, and the updated abstract is displayed after receiving an operation. Or, in the case where there is no second search information, the first control is used to prompt the user to subscribe to the search result of the first search information, and the updated abstract is displayed after a preset time after receiving an operation.

[0156] The updated abstract is also the first abstract in this application. After receiving the information to be retrieved from the user, the terminal can display the abstract corresponding to the information to be retrieved, that is, the initial abstract, and can prompt the user that the terminal can also display the updated abstract, so that the user can quickly obtain new information through the updated abstract. In addition, displaying the initial abstract first can also make it easier for the user to accept the new function and bring a better experience to the user.

[0157] If the user has retrieved historical search information with the same point of interest as the information to be retrieved in the past, the first control can be used to prompt the user that the updated abstract can be displayed. In the above case, the content displayed by the first control can be "It is detected that relevant content has been retrieved before. Generate an updated abstract for me", etc. For example, reference can be made to Figure 8 , Figure 8 The control 801 in (a) in

[0158] If the user has not retrieved historical search information with the same point of interest as the information to be retrieved in the past, the first control can be used to prompt the user that the update of the search result corresponding to the information to be retrieved can be subscribed. In the above case, the content displayed by the first control can be "Track the changes after one month", etc. For example, reference can be made to Figure 9 , Figure 9 The control 901 in (a) in

[0159] In addition to being able to display the initial abstract and the first control, the terminal can also display the first search result. For example, the search result can be displayed after the above initial abstract and the first control, so that the user can further view the search result.

[0160] In addition, the first control may be specifically displayed according to the first indication information above. When the first indication information indicates that there is historical retrieval information that matches the information to be retrieved, the first control can be used to prompt for an updated summary. When the first indication information indicates that there is no historical retrieval information that matches the information to be retrieved, the first control can be used to prompt the user to subscribe to changes in the retrieval results of the first retrieval information.

[0161] Step 709, the terminal receives the user's operation on the first control.

[0162] The above operation may be a triggering operation on the first control, such as clicking, swiping, etc. on the first control. Combining Figure 8 and Figure 9 , the above operation may be a click operation on control 801 or control 901.

[0163] Step 710, in response to the above operation, the terminal sends second indication information to the server.

[0164] Among them, when there is historical retrieval information that matches the information to be retrieved, the second indication information is used to indicate an update to the summary of the retrieval results of the information to be retrieved; or, when there is no historical retrieval information that matches the information to be retrieved, the second indication information is used to indicate subscribing to the retrieval results of the information to be retrieved.

[0165] Correspondingly, the server needs to receive the second indication information from the terminal. Corresponding to Figure 4 , the one that receives the second indication information can be the summary generation module 42.

[0166] Specifically, if the user interacts with the first control, it can be determined that the user wants to obtain an updated summary. When the first control is Figure 8 control 801 in Figure 9 , that is, when the first control is used to prompt the user to obtain the summary of the updated retrieval results of the first retrieval information, in response to the above operation, the terminal can send second indication information to the server for indicating an update to the summary of the retrieval results of the information to be retrieved. When the first control is

[0167] control 901 in

[0168] Step 711, generate an updated summary.

[0169] Step 7111, when the second indication information is used to indicate updating the summary of the retrieval result of the information to be retrieved, the server obtains the second retrieval result corresponding to the historical retrieval information that matches the information to be retrieved, and generates an updated summary according to the first retrieval result and the second retrieval result.

[0170] Step 7112, when the second indication information is used to indicate subscribing to the retrieval result of the information to be retrieved, the second retrieval result corresponding to the information to be retrieved is obtained after a preset time, and an updated summary is generated according to the first retrieval result and the second retrieval result.

[0171] That is, if the user has retrieved historical retrieval information that matches the information to be retrieved in the past, then the server can immediately generate an updated summary according to the second indication information of the terminal. If the user has not retrieved historical retrieval information that matches the information to be retrieved in the past, then the server can generate an updated summary after a preset time according to the second indication information of the terminal, so that the user can track the above changes.

[0172] For the specific method of generating the updated summary, reference can be made to the description of step 502 above, which will not be elaborated here. Correspondingly Figure 4 steps 7111 and 7112 can be Figure 4 executed by the summary generation unit 4222 in. Here, a specific example will be given to illustrate the generation process.

[0173] Figure 4 The processing unit 4212 in can integrate and embed the data input in the format into the summary generation unit 4222. The specific input data may include: the information to be retrieved by the user, such as "Chinese general large model"; the retrieval result corresponding to the information to be retrieved; the corresponding point of interest, such as "Chinese large language model", and the user's historical information obtained based on the point of interest, such as the retrieval result and summary corresponding to the user's historical retrieval information, the user's feedback information, etc.

[0174] For the specific form of the updated summary, it may include text prompting the user to update the content, and may also include the above first time and second time. Here, a specific embodiment will be used to illustrate the updated summary.

[0175] When the information to be retrieved by the user is "Chinese general large model", the updated summary may include:

[0176] "Good morning. Today is xxxx year, x month, x day. The Chinese general large model has the following updated content since your last search (yyyy year, y month, y day):

[0177] First, model 4 is released.

[0178] Model 4 is the second-generation version of the open-source Chinese-English bilingual dialogue model released by Company X last month. The performance of the new version of Model 4 has been greatly improved, and the inference speed has increased by 42%.

[0179] Second, the evaluation list of Model 5 has been updated.

[0180] In the past month, Model 5 announced its evaluation scores. The evaluation scores of Model 5 on the Chinese evaluation list far exceeded models such as Model 6, and it led Model 7 by a large margin on the English authoritative evaluation list.

[0181] Step 712, the server sends an update summary to the terminal.

[0182] Correspondingly, the terminal needs to receive the update summary from the server. Specifically, it can be the update summary sent by the summary generation module 42 in Figure 4 to the terminal.

[0183] Step 713, the terminal displays the update summary.

[0184] After receiving the update summary, the terminal can display the update summary. It should be noted that if the user has retrieved historical retrieval information that matches the information to be retrieved in the past, steps 7112, 712, and step 713 can be executed immediately. If the user has not retrieved historical retrieval information that matches the information to be retrieved in the past, steps 7111, 712, and step 713 can be executed after a period of time after step 710 is executed.

[0185] The method for the terminal to display the update summary can refer to the description above and will not be elaborated here.

[0186] Such as Figure 8 and Figure 9 shown, in the case where the user has retrieved historical retrieval information that matches the information to be retrieved in the past, as Figure 8 shown, according to the user's operation on the first control 801 in (a) of Figure 8 , even if an update summary is generated and displayed. As Figure 8 shown in (b), after the user clicks the first control 801, the content shown in (b) of Figure 8 can be displayed, enabling the user to obtain new information more conveniently, thus bringing a better user experience. In the case where the user has not retrieved historical retrieval information that matches the information to be retrieved in the past, as Figure 9 shown, according to the user's operation on the first control 901 in (a) of Figure 9 , an update summary is generated at a second time after a preset time (i.e., one month) and is presented to the user through push or other means. After the user clicks the push, the terminal can expand the push and display the content shown in (b) of Figure 9 .

[0187] Figure 10 is a summary of the above specific embodiments. As Figure 10 shown, in the above specific embodiments, the information to be retrieved can be input through the search box in Figure 10 . For the information to be retrieved, matching can be performed to obtain the point-of-interest matching results. For the information to be retrieved, the first retrieval results can also be obtained through a search engine. By inputting the point-of-interest matching results and the information to be retrieved into the summary generation model, an initial summary can be obtained. The terminal can display the initial summary and, according to the first indication information for indicating the point-of-interest matching results, display the first control as Figure 10 shown. After the terminal receives the operation of the user on the first control, second indication information can be generated, and then an updated summary can be generated according to the second indication information and the stored data, and as Figure 10 shown, the terminal displays the updated summary. Figure 10 The solid-line boxes in

[0188] represent data, and the dashed-line boxes represent the actions performed.

[0189] This application also provides an information retrieval device, which has the function of implementing the behavior of the terminal or the server in the above information retrieval method. The function can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions. For example, a receiving module, a processing module, a sending module, a display module, etc. When the device is used to implement the terminal function, the receiving module can execute the above steps 501, 701, and 709, the display module can be used to execute the above steps 502, 708, 713, and the sending module can be used to implement steps 702 and 710. When the device is used to implement the server function, the processing module can be used to implement steps 703, 704, 705, and 711, and the sending module can be used to execute steps 706 and 712.

[0189] This application also provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor is used to execute the computer program to implement the above information retrieval method.

[0190] This application also provides a computer-readable storage medium, on which a computer program is stored. The computer program is used to implement the above information retrieval method.

[0191] This application also provides a computer program product, including computer-readable code or a non-volatile computer-readable storage medium carrying the computer-readable code. When the computer-readable code runs in an electronic device, the processor in the electronic device executes the above information retrieval method.

[0192] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claimed rights.

Claims

1. An information retrieval method, characterized in that, applied to a terminal, the method includes: receiving first retrieval information input by a user; displaying a first abstract according to the first retrieval information; the first abstract is used to indicate the difference between a first retrieval result and a second retrieval result, and the first retrieval result is the retrieval result of the first retrieval information retrieved at a first time; wherein, in the presence of second retrieval information, the second retrieval result is the retrieval result of the second retrieval information; or, in the absence of the second retrieval information, the second retrieval result is the retrieval result of the first retrieval information retrieved at a second time, the second time being later than the first time, and the second retrieval information is retrieval information having the same interest point as the first retrieval information retrieved by the user before receiving the first retrieval information.

2. The method according to claim 1, characterized in that, the displaying the first abstract according to the first retrieval information includes: displaying a second abstract and a first control, the second abstract being the abstract of the first retrieval result; receiving an operation of the user on the first control; responding to the operation, displaying the first abstract; wherein, in the presence of the second retrieval information, the first control is used to prompt the user to obtain an updated abstract of the retrieval result of the first retrieval information, and the first abstract is displayed after receiving the operation; or, in the absence of the second retrieval information, the first control is used to prompt the user to subscribe to the retrieval result of the first retrieval information, and the first abstract is displayed after a preset time of receiving the operation.

3. The method according to claim 2, characterized in that, the displaying the second abstract and the first control according to the first retrieval information includes: sending the first retrieval information to a server; receiving the second abstract and a first indication information from the server, the first indication information being used to indicate whether there is the second retrieval information; displaying the second abstract and displaying the first control according to the first indication information.

4. The method according to claim 2 or 3, characterized in that, the responding to the operation and displaying the first abstract includes: responding to the operation, sending second indication information to the server; receiving the first abstract from the server; wherein, in the presence of the second retrieval information, the second indication information is used to indicate to update the abstract of the retrieval result of the first retrieval information; or, in the absence of the second retrieval information, the second indication information is used to indicate to subscribe to the retrieval result of the first retrieval information.

5. An information retrieval method, characterized in that, applied to a server, the method includes: receiving first retrieval information from a terminal; sending a first abstract corresponding to the first retrieval information to the terminal; Wherein, the first abstract is used to indicate the difference between the first retrieval result and the second retrieval result, and the first retrieval result is the retrieval result of the first retrieval information retrieved at a first time; wherein, in the case where there is second retrieval information, the second retrieval result is the retrieval result of the second retrieval information; or, in the case where there is no second retrieval information, the second retrieval result is the retrieval result of the first retrieval information retrieved at a second time, and the second time is later than the first time, and the second retrieval information is the retrieval information retrieved by the user before receiving the first retrieval information and having the same interest point as the first retrieval information.

6. The method according to claim 5, wherein, before sending the first abstract corresponding to the first retrieval information to the terminal, the method further includes: generating a second abstract according to the first retrieval information; the second abstract is the abstract of the first retrieval result; obtaining an interest point matching result according to the first retrieval information, and the interest point matching result is used to indicate whether there is second retrieval information having the same interest point as the first retrieval information; sending the second abstract to the terminal, and sending first indication information to the terminal according to the interest point matching result, and the first indication information is used to indicate whether there is the second retrieval information.

7. The method according to claim 5 or 6, wherein, before sending the first abstract corresponding to the first retrieval information to the terminal, the method further includes: obtaining a first retrieval result corresponding to the first retrieval information; receiving second indication information from the terminal; in the case where the second indication information is used to indicate updating the abstract of the retrieval result of the first retrieval information, obtaining a second retrieval result corresponding to the second retrieval information, and generating a first abstract according to the first retrieval result and the second retrieval result; or, in the case where the second indication information is used to indicate subscribing to the retrieval result of the first retrieval information, obtaining a second retrieval result corresponding to the first retrieval information after a preset time, and generating a first abstract according to the first retrieval result and the second retrieval result.

8. The method according to claim 7, wherein, the generating the first abstract according to the first retrieval result and the second retrieval result includes: taking the first retrieval result, the second retrieval result and first data as inputs, and generating the first abstract through an artificial intelligence model; wherein, the first data includes at least one of the following: the first retrieval information, the feedback information of the user, and the interest point. In the case where there is second retrieval information, the feedback information is used to indicate the interaction information of the user with the page corresponding to the second retrieval result; or, in the case where there is no second retrieval information, the feedback information is the interaction information of the user with the page corresponding to the first retrieval result.

9. An electronic device, wherein, Comprising a memory and a processor, the memory is used for storing a computer program, and the processor is used for executing the computer program to implement the information retrieval method according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and the computer program is used to implement the information retrieval method according to any one of claims 1-8.

Images