CN116561263A

CN116561263A - Searching method, searching device, searching equipment and storage medium

Info

Publication number: CN116561263A
Application number: CN202210114605.8A
Authority: CN
Inventors: 刘叶舟
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2022-01-30
Filing date: 2022-01-30
Publication date: 2023-08-08

Abstract

The embodiment of the application provides a searching method, device, equipment and storage medium, which can be applied to various scenes such as cloud technology, artificial intelligence, intelligent traffic, vehicle-mounted and the like, and the method comprises the following steps: in response to a search operation triggered in the search interface for an original search condition, displaying a search recommendation interface comprising: target search intention and at least one recommended search condition corresponding to the target search intention. In response to the scrolling operation triggered by the target search intention in the search recommendation interface, other recommended search conditions corresponding to the target search intention are displayed in the search recommendation interface, namely, the target object is guided to pass through the scrolling operation in the search process, so that the search recommendation interface displays more recommended search conditions associated with the target search intention, the target object can obtain the recommended search conditions which are most matched with the target search intention to search, the accuracy and the efficiency of the search are improved, and the search experience of the target object is greatly improved.

Description

Searching method, searching device, searching equipment and storage medium

Technical Field

The embodiment of the application relates to the technical field of artificial intelligence, in particular to a searching method, device, equipment and storage medium.

Background

With the development of internet technology, information in the internet is increasing, and in order to facilitate a target object to quickly obtain required information from massive information, an information search function becomes an indispensable part of many applications.

When searching, the related technology firstly performs word segmentation on the search words input by the target object to obtain key words in the search words, and then searches based on the obtained key words to obtain search results corresponding to the search words.

However, in the above scheme, when the target object cannot accurately describe the content to be searched, there is a deviation in the search word input by the target object, resulting in lower accuracy of the output search result.

Disclosure of Invention

The embodiment of the application provides a search method, a search device, search equipment and a storage medium, which are used for improving the accuracy of search results.

In one aspect, an embodiment of the present application provides a search method, including:

in response to a search operation triggered in the search interface for an original search condition, displaying a search recommendation interface comprising: a target search intention corresponding to the original search condition, and at least one recommended search condition corresponding to the target search intention;

Responding to a scrolling operation triggered by aiming at the target search intention in the search recommendation interface, and displaying other recommended search conditions corresponding to the target search intention in the search recommendation interface;

and responding to search condition selection operation triggered by each displayed recommended search condition in the search recommendation interface, and displaying target search results corresponding to the selected target recommended search condition.

In one aspect, an embodiment of the present application provides a search apparatus, including:

the recommendation module is used for responding to the search operation triggered based on the original search condition in the search interface, and displaying a search recommendation interface, wherein the search recommendation interface comprises: a target search intention corresponding to the original search condition, and at least one recommended search condition corresponding to the target search intention;

the recommending module is further used for responding to a scrolling operation triggered by aiming at the target searching intention in the searching recommending interface, and displaying other recommending searching conditions corresponding to the target searching intention in the searching recommending interface;

and the search module is used for responding to search condition selection operation triggered by each displayed recommended search condition in the search recommendation interface and displaying target search results corresponding to the selected target recommended search condition.

Optionally, the recommendation module is specifically configured to:

and in the search recommendation interface, displaying other recommended search conditions corresponding to the target search intention according to the recommendation sequences corresponding to the other recommended search conditions.

Optionally, the search module is specifically configured to:

responsive to a search condition selection operation triggered for the at least one recommended search condition in the search recommendation interface, displaying a search result interface comprising: and selecting target search results corresponding to the target search conditions from the at least one recommended search condition.

Optionally, the search module is specifically configured to:

in response to search condition selection operations triggered for the other recommended search conditions in the search recommendation interface, displaying a search result interface comprising: and selecting target search results corresponding to the target search conditions from the other recommended search conditions.

Optionally, the search recommendation interface further includes: the original search condition and a connecting line between the original search condition and the target search intention are used for representing the association relation between the original search condition and the target search intention.

Optionally, the recommendation module is further configured to:

before a search recommendation interface is displayed, determining a corresponding target search intention based on a preset knowledge graph according to the original search condition, and acquiring at least one recommendation search condition corresponding to the target search intention.

Optionally, the recommendation module is specifically configured to:

inquiring a preset knowledge graph based on the original search condition;

when it is determined that a reference search condition matching the original search condition exists in the knowledge graph, taking a reference search intention corresponding to the reference search condition in the knowledge graph as the target search intention, and taking at least one associated search condition corresponding to the reference search intention as the at least one recommended search condition.

Optionally, at least one recommended search condition in the search recommendation interface is displayed according to a recommendation sequence corresponding to the at least one recommended search condition;

the recommendation module is further configured to:

before a search recommendation interface is displayed, a recommendation sequence corresponding to the at least one recommendation search condition is obtained from the knowledge graph; or alternatively, the process may be performed,

determining search preference characteristics of the target object based on a first historical search record of the target object associated with the original search condition, and performing recommendation ordering on the at least one recommendation search condition based on the search preference characteristics to obtain a recommendation sequence corresponding to the at least one recommendation search condition.

Optionally, the recommendation module is further configured to:

before the original search condition is acquired, the following steps are respectively executed for a plurality of reference search conditions: predicting search intention of a reference search condition by adopting a target recommendation model to obtain corresponding reference search intention;

and constructing a knowledge graph based on the plurality of reference search conditions, the plurality of obtained reference search intents and at least one associated search condition corresponding to each of the plurality of reference search intents.

Optionally, the recommendation module is specifically configured to:

for the plurality of reference search intents, the following steps are respectively performed: based on a second historical search record associated with one reference search intention, sorting at least one associated search condition corresponding to the one reference search intention to obtain a recommendation sequence corresponding to the at least one associated search condition;

and constructing the knowledge graph based on the plurality of reference search conditions, the plurality of obtained reference search intents, at least one associated search condition corresponding to each of the plurality of reference search intents and the corresponding recommendation sequence.

Optionally, the recommendation module is further configured to:

acquiring a plurality of sample search conditions and corresponding mark search intents;

Performing iterative training on the recommended model to be trained by adopting the plurality of sample search conditions and the corresponding mark search intention until the iteration stopping condition is met, and obtaining a target recommended model, wherein each iterative process comprises the following steps:

determining a corresponding predicted search intent based on the sample search conditions;

and determining a target loss value based on the predicted search intention and the marked search intention, and carrying out parameter adjustment on the recommendation model to be trained through the target loss value.

In one aspect, embodiments of the present application provide a computer program comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the search method described above when the program is executed.

In one aspect, embodiments of the present application provide a computer-readable storage medium storing a computer program executable by a computer device, which when run on the computer device, causes the computer device to perform the steps of the search method described above.

In one aspect, embodiments of the present application provide a computer program product comprising a computer program stored on a computer readable storage medium, the computer program comprising program instructions which, when executed by a computer device, cause the computer device to perform the above-described search method steps.

In the embodiment of the application, in response to the search operation triggered by the original search condition in the search interface, the target search intention corresponding to the original search condition and at least one recommended search condition corresponding to the target search intention are displayed to the target object, and compared with the original search condition, the recommended search condition is more matched with the real search intention of the target object, so that the accuracy of the search result can be effectively improved when searching is performed based on the recommended search condition. And secondly, responding to the scrolling operation triggered by the target searching intention in the searching recommendation interface, and displaying other recommended searching conditions corresponding to the target searching intention in the searching recommendation interface, namely, in the searching process, guiding the target object to display more other recommended searching conditions associated with the target searching intention through the scrolling operation, and providing more selectable recommended searching conditions for the target object, so that the target object can acquire the target recommended searching condition which is most matched with the target searching intention from a plurality of recommended searching conditions to search, and acquire target searching results, thereby improving the searching accuracy and efficiency of the target object and greatly improving the searching experience of the target object.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, it will be apparent that the drawings in the following description are only some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic structural diagram of a system architecture according to an embodiment of the present application;

FIG. 2 is a schematic diagram of a search interface according to an embodiment of the present disclosure;

FIG. 3 is a schematic diagram I of a search recommendation interface according to an embodiment of the present application;

FIG. 4 is a second schematic diagram of a search recommendation interface according to an embodiment of the present disclosure;

FIG. 5 is a third schematic diagram of a search recommendation interface according to an embodiment of the present disclosure;

FIG. 6 is a schematic diagram IV of a search recommendation interface according to an embodiment of the present application;

FIG. 7 is a schematic diagram of a search results interface provided by an embodiment of the present application;

fig. 8 is a schematic flow chart of a search method according to an embodiment of the present application;

fig. 9 is a schematic diagram fifth of a search recommendation interface provided in an embodiment of the present application;

FIG. 10a is a diagram sixth of a search recommendation interface according to an embodiment of the present disclosure;

FIG. 10b is a diagram seventh of a search recommendation interface according to an embodiment of the present disclosure;

fig. 11 is a schematic structural diagram of a knowledge graph according to an embodiment of the present application;

FIG. 12 is a flowchart of a method for obtaining a second history search record according to an embodiment of the present disclosure;

fig. 13 is a second schematic structural diagram of a knowledge graph according to an embodiment of the present application;

fig. 14 is a second flowchart of a search method according to an embodiment of the present application;

fig. 15 is a schematic structural diagram of a search device according to an embodiment of the present application;

fig. 16 is a schematic structural diagram of a computer device according to an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantageous effects of the present invention more apparent, the present invention will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

For ease of understanding, the terms involved in the embodiments of the present invention are explained below.

Artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use the knowledge to obtain optimal results. In other words, artificial intelligence is an integrated technology of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence, i.e. research on design principles and implementation methods of various intelligent machines, enables the machines to have functions of sensing, reasoning and decision.

The artificial intelligence technology is a comprehensive subject, and relates to the technology with wide fields, namely the technology with a hardware level and the technology with a software level. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.

Machine Learning (ML) is a multi-domain interdisciplinary, involving multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, etc. It is specially studied how a computer simulates or implements learning behavior of a human to acquire new knowledge or skills, and reorganizes existing knowledge structures to continuously improve own performance. Machine learning is the core of artificial intelligence, a fundamental approach to letting computers have intelligence, which is applied throughout various areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, confidence networks, reinforcement learning, transfer learning, induction learning, teaching learning, and the like. For example, in the embodiment of the application, a machine learning technology is adopted to respectively predict reference search intentions corresponding to a plurality of reference search conditions, and a knowledge graph is constructed based on the plurality of reference search conditions, the obtained plurality of reference search intentions, and at least one associated search condition corresponding to the plurality of reference search intentions.

And (3) data marking: data annotation is the process of annotating metadata such as text, video, images, etc., and the marked data will be used to train a machine-learned model.

Semantic vector: converting symbolic representations of text into vector representations in semantic space is now a common practice to quantify comparative semantics, which is usually based on the distributed assumption of Harris that words in similar contexts usually have similar semantics.

Data enhancement: (Data Augmentation), a technique for artificially expanding a training dataset by letting limited data produce more equivalent data.

A Long short-term memory (LSTM) is a special recurrent neural network (Recurrent Neural Network, RNN) that can learn Long-term dependency information.

Deep neural network (Deep Neural Networks, DNN for short): a framework for deep learning, which is a neural network with at least one hidden layer. Similar to the shallow neural network, the deep neural network can also provide modeling for a complex nonlinear system, but the extra level provides a higher level of abstraction for the model, thus increasing the model's ability.

Knowledge graph: a semantic network whose nodes represent entities (entities) or concepts (concepts) and edges represent various semantic relationships between entities/concepts.

Search intention: "requirements" behind a particular search query.

The following describes the design ideas of the embodiments of the present application.

When searching, the related technology firstly performs word segmentation on the search words input by the target object to obtain key words in the search words, and then searches based on the obtained key words to obtain search results corresponding to the search words. However, in the above scheme, when the target object cannot accurately describe the content to be searched, there is a deviation in the search word input by the target object, resulting in lower accuracy of the output search result.

It was found by analysis that when the target object does not know how clearly to describe the content that needs to be searched, although the target object enters a search term, the target object may want to obtain content that is not related to the search term, but there are other search intents. For example, when the target object searches for the "weight loss" search term, the target object may want to obtain not the content related to the "weight loss" search term, but the content related to the "weight loss", that is, the actual search intention of the target object is "weight loss". At this time, if the search word related to "thin display" is displayed in the search recommendation interface, and the target object is guided to continuously scroll and update the displayed search word in the search recommendation interface, so that the target object obtains the search word which is most matched with the actual search intention of the target object to perform the search, the accuracy and the efficiency of the search are effectively improved, and the search experience of the target object is greatly improved.

In view of this, the embodiment of the application provides a search method, which includes: in response to a search operation triggered in the search interface for an original search condition, displaying a search recommendation interface, wherein the search recommendation interface comprises: target search intents corresponding to the original search conditions, and at least one recommended search condition corresponding to the target search intents. And responding to the scrolling operation triggered by the target search intention in the search recommendation interface, and displaying other recommended search conditions corresponding to the target search intention in the search recommendation interface. And then, responding to search condition selection operation triggered by each displayed recommended search condition in the search recommendation interface, and displaying target search results corresponding to the selected target recommended search condition.

Referring to fig. 1, a system architecture diagram applicable to the embodiments of the present application, where the system architecture at least includes a terminal device 101 and a search server 102, the number of the terminal devices 101 may be one or more, and the number of the search servers 102 may be one or more, and the number of the terminal devices 101 and the search servers 102 is not specifically limited in this application.

The terminal device 101 is pre-installed with a target application having a search function, where the target application may be a client application, a web page application, an applet application, or the like. The terminal device 101 may be, but is not limited to, a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart home appliance, a smart voice interaction device, a smart car-mounted device, and the like.

The search server 102 is a background server of the target application, and the search server 102 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, a content delivery network (Content Delivery Network, CDN), basic cloud computing services such as big data and an artificial intelligence platform. The terminal device 101 and the search server 102 may be directly or indirectly connected through wired or wireless communication, which is not limited herein.

The searching method in the embodiment of the present application may be executed by the terminal device 101, may be executed by the search server 102, or may be executed by the terminal device 101 and the search server 102 interactively.

The search method in the embodiment of the present application is executed by the terminal device 101, for example, including the steps of:

the terminal device 101 responds to a search operation triggered in the search interface for the original search condition, and displays a search recommendation interface, wherein the search recommendation interface comprises: target search intents corresponding to the original search conditions, and at least one recommended search condition corresponding to the target search intents. And responding to the scrolling operation triggered by the target search intention in the search recommendation interface, and displaying other recommended search conditions corresponding to the target search intention in the search recommendation interface. And then, responding to search condition selection operation triggered by each displayed recommended search condition in the search recommendation interface, and displaying target search results corresponding to the selected target recommended search condition.

In practical applications, the search method in the embodiments of the present application may be applied to search for any media content, including but not limited to articles, images, and videos.

The following are article searches by way of example:

the terminal device starts a pre-installed search application, and the search application displays a search interface, as shown in fig. 2, wherein the search interface comprises a search box and a 'determination' button. After entering the original search term "lose weight" in the search box, click the ok button.

The terminal equipment responds to a search operation triggered by the original search word weight reduction in the search interface, and displays a search recommendation interface, as shown in fig. 3, wherein the search recommendation interface comprises an association relation between the original search word weight reduction and the target search intention weight reduction, namely the original search word weight reduction is connected to a recommendation area 301 of the target search intention weight reduction in a connecting line mode, and the recommendation area 301 represents a search star corresponding to the target search intention weight reduction.

The terminal equipment displays recommended search words corresponding to the target search intention 'lean display' in the search recommendation interface according to the recommendation sequence at the same time: "lean wearing skill", "lean dress", "lean pants", as shown in fig. 4, recommended search terms: "lean wearing skill", "lean dress", "lean pants" are connected by means of a wire to the recommended area 301 of the target search intention "lean".

The terminal equipment responds to the scrolling operation triggered by the recommendation area 301 in the search recommendation interface, and updates the recommended search words displayed on the search recommendation interface, namely, the display: "lean dress", "lean pants", "lean make-up skills", as shown in FIG. 5, recommended search terms: the "lean dress", "lean trousers", "lean cosmetic skills" are connected to the recommended region 301 of the target search intention "lean", by means of a wire, and the recommended search word "lean wearing skills" is in a hidden state because the size of the search recommended interface is limited. The search recommendation interface may also display four recommended search terms, i.e., a "lean wearing skill", "a lean dress", "a lean trousers", "a lean makeup skill", at the same time, which is not particularly limited in this application.

Clicking a target object on a target search word 'lean-showing cosmetic skill' displayed on a search recommendation interface, and as shown in fig. 6, responding to a clicking operation triggered on the search recommendation interface, recommending the search word by the terminal equipment: the target search word 'lean making-up skill' is selected from 'lean dress', 'lean trousers', 'lean making-up skill', and then a search request carrying the target search word is sent to a search server.

The search server acquires a plurality of articles associated with the target search word 'lean make-up skill' from the search library, and then transmits related information of the plurality of articles to the terminal device. The terminal device displays a search result interface, as shown in fig. 7, where the search result interface includes an article information display area 701 and an article information display area 702, where the article information display area 701 includes an article title "display thin makeup course" and a corresponding article thumbnail, and the article information display area 702 includes an article title "5 minutes to draw a display thin makeup look" and a corresponding article thumbnail.

It should be noted that, the search method in the embodiment of the present application is not limited to be applied to the above application scenario, but may be a commodity search scenario, a take-out search scenario, a merchant information search scenario, an audio/video search scenario, and the like.

It will be appreciated that in the specific embodiments of the present application, related data of users such as historical search records, search preference features, etc., when the embodiments of the present application are applied to specific products or technologies, user permissions or consents need to be obtained, and the collection, use and processing of related data need to comply with related laws and regulations and standards of related countries and regions.

Based on the system architecture diagram shown in fig. 1, the embodiment of the present application provides a flow of a search method, as shown in fig. 8, where the flow of the method is performed by a computer device, which may be the terminal device 101 and/or the search server 102 shown in fig. 1, and includes the following steps:

step S801, in response to a search operation triggered in the search interface for the original search condition, displays a search recommendation interface.

In particular, the search criteria may be text (e.g., search terms), images (e.g., search pictures), audio-visual (e.g., search speech), and the like. The original search condition is a search condition input by a target object in an application with a search function, and the target object includes but is not limited to an account number and a device number.

The search recommendation interface includes: target search intents corresponding to the original search conditions, and at least one recommended search condition corresponding to the target search intents. Target search intent refers to the "need" behind a particular search query, which may be characterized in terms of text, images, audio-video, and the like. The target search intent itself may also be presented as a recommended search criteria in the search recommendation interface. When the original search condition corresponds to a plurality of target search intents, at least one recommended search condition corresponding to each target search intention may be acquired, respectively.

When the original search condition is text, words that are useless or have interference with the search, such as words of "o", etc., may be included in the original search condition. In order to improve the searching efficiency and accuracy, after the original searching condition is obtained, the original searching condition is segmented, and then keywords are obtained from the segmentation result. And determining a corresponding target search intention based on the obtained keywords, and obtaining at least one recommended search condition corresponding to the target search intention.

Optionally, the search recommendation interface further comprises: the search system comprises an original search condition and a connecting line between the original search condition and a target search intention, wherein the connecting line is used for representing the association relation between the original search condition and the target search intention.

Specifically, in the search recommendation interface, a connecting line mode is adopted to connect the area where the original search condition is located to the area where the target search intention is located. In the same way, in the search recommendation interface, the area where the recommended search condition is located is connected to the area where the target search intention is located in a connecting line mode, so that the association relationship between the recommended search condition and the target search intention is represented.

For example, referring to fig. 9, a schematic diagram of a search recommendation interface provided in an embodiment of the present application includes an original search term "chicken breast", a target search intention "exercise", and a recommended search term: "exercise teaching", "classical exercise", "exercise", wherein the original search word "chicken breast" is connected by wire to the recommended area 1201 of the target search intention "chicken breast", while the search word is recommended: the "exercise teaching", "classical exercise", "exercise" are connected to the recommended area 1201 of the target search intention "chicken breast", respectively, by means of a wire.

In the embodiment of the application, the original search condition is connected to the target search intention in the search recommendation interface in a connecting line manner, so that the target object can clearly acquire the association relationship between the original search condition and the target search intention, and further, a proper recommended search condition is selected from the recommended search conditions corresponding to the target search intention for searching, thereby improving the search efficiency.

Step S802, responding to a scrolling operation triggered by aiming at a target search intention in a search recommendation interface, and displaying other recommended search conditions corresponding to the target search intention in the search recommendation interface.

Specifically, since the screen size of the terminal device is limited and the recommended search conditions corresponding to the target search intention are many, the terminal device cannot simultaneously display all the recommended search conditions corresponding to the target search intention. When all the recommended search conditions cannot be displayed at the same time in the search recommendation interface, a part of the recommended search conditions can be displayed first, then in response to a scrolling operation triggered by the target search intention in the search recommendation interface, other recommended search conditions are not displayed before the progressive display, wherein the scrolling operation can rotate clockwise, anticlockwise and the like, and all the recommended search conditions corresponding to the target search intention can be displayed in sequence according to the recommendation sequence.

When other recommended search conditions are displayed in the search recommendation interface, the recommended search conditions which have been displayed before can be displayed at the same time, and the display size of each recommended search condition displayed in the search recommendation interface (the size of the recommended search condition which has been displayed before can also be reduced) can be reduced so as to adapt to the screen size of the terminal device.

When other recommended search conditions are displayed in the search recommendation interface, the recommended search conditions which are displayed before can be hidden, so that the display size of each recommended search condition displayed in the search recommendation interface does not need to be adjusted.

In some embodiments, the number of recommended search conditions per update in the search recommendation interface is determined based on the scroll angle of the scroll operation. If the search recommendation interface displays all recommended search conditions after the scrolling operation for a plurality of times, the terminal device can respond to the rollback operation triggered by the target search intention in the search recommendation interface to display the recommended search conditions which are displayed before. The terminal device may also continue to present the respective recommended search conditions from scratch in response to a scroll operation triggered in the search recommendation interface for the target search intention.

For example, the recommended region 1201 of the target search intention "chicken breast" is a scrollable circular region, and on the basis of the search recommendation interface shown in fig. 9, the terminal device updates the recommended search word presented in the search recommendation interface, that is, presents the recommended search word, in response to a clockwise scrolling operation for the recommended region 1201 triggered in the search recommendation interface: "classical exercise", "exercise machine", "exercise meal practice", as shown in fig. 10a, wherein "exercise meal practice" is a newly presented recommended search term and conceals the recommended search term: "exercise action teaching".

For example, the recommended region 1201 of the target search intention "chicken breast" is a scrollable circular region, and on the basis of the search recommendation interface shown in fig. 9, the terminal device updates the recommended search word presented in the search recommendation interface, that is, presents the recommended search word, in response to a clockwise scrolling operation for the recommended region 1201 triggered in the search recommendation interface: "exercise teaching", "classical exercise", "exercise machine", "exercise meal practice", as shown in fig. 10b, wherein "exercise meal practice" is a newly presented recommended search term.

Step S803, in response to a search condition selection operation triggered in the search recommendation interface for each presented recommended search condition, presenting a target search result corresponding to the selected target recommended search condition.

In specific implementation, the search condition selection operation may be a click operation, a double click operation, a long press operation, or the like. In response to search condition selection operations in the search recommendation interface triggered for each presented recommended search condition, selecting a target search condition from each recommended search condition, and then obtaining target search results matching the target search condition from a search library, the target search results including but not limited to articles, images, audio, video. The search repository may be located at the terminal device, may be located at the search server, or may be independent of the terminal device and the search server.

In one possible implementation, in response to a search condition selection operation triggered for at least one recommended search condition in a search recommendation interface, displaying a search result interface, the search result interface includes: and selecting target search results corresponding to the target search conditions from the at least one recommended search condition.

For example, referring to fig. 10a, for a schematic diagram of a search recommendation interface provided in an embodiment of the present application, a target object clicks a recommended search word "classical exercise" in the search recommendation interface, and a terminal device sends a search request carrying the target search word "classical exercise" to a search server in response to a search condition selection operation triggered in the search recommendation interface.

The search server acquires a plurality of articles associated with the target search word 'classical exercise action' from the search library, and then sends relevant information of the articles to the terminal equipment. And the terminal equipment displays the related information of a plurality of articles in the search result interface.

In another possible implementation manner, in response to a search condition selection operation triggered for other recommended search conditions in the search recommendation interface, a search result interface is displayed, where the search result interface includes: and selecting target search results corresponding to the target search conditions from other recommended search conditions.

For example, referring to fig. 10a, for a schematic diagram of a search recommendation interface provided in the embodiments of the present application, a "gym practice" is a newly displayed recommended search term, a target object clicks the recommended search term "gym practice" in the search recommendation interface, and a terminal device sends a search request carrying the target search term "gym practice" to a search server in response to a search condition selection operation triggered in the search recommendation interface.

The search server acquires a plurality of articles associated with the target search word 'body building meal' from the search library, and then sends related information of the articles to the terminal equipment. And the terminal equipment displays the related information of a plurality of articles in the search result interface.

In the embodiment of the application, more recommended search conditions associated with the target search intention are displayed in response to the scrolling operation triggered by the target object aiming at the target search intention in the search recommendation interface, so that the target object can select the target search condition which is most matched with the self requirement in the more recommended search conditions, and the accuracy of the search result is improved.

Optionally, before displaying the search recommendation interface, determining a corresponding target search intention according to the original search condition based on a preset knowledge graph, and acquiring at least one recommendation search condition corresponding to the target search intention.

Specifically, the knowledge graph may be located at the terminal device, may be located at the search server, or may be independent of the terminal device and the search server. Before a knowledge graph is adopted, a corresponding target search intention is determined according to an original search condition, and at least one recommended search condition corresponding to the target search intention is obtained, the embodiment of the application adopts at least the following embodiments to construct the knowledge graph:

in the first embodiment, a plurality of reference search conditions are acquired, and the following steps are executed for each of the plurality of reference search conditions: and predicting the search intention of a reference search condition by adopting the target recommendation model to obtain a corresponding reference search intention. A knowledge graph is constructed based on the plurality of reference search conditions, the plurality of obtained reference search intents, and at least one associated search condition to which the plurality of reference search intents each corresponds.

Specifically, the target recommendation model may be an LSTM model, a DNN model, or the like. For each reference search condition, the reference search intention corresponding to the reference search condition may be at least one reference search condition among other reference search conditions, or may be a search condition other than the other reference search conditions.

At least one associated search condition is configured for each reference search intention, and the associated search condition corresponding to each reference search intention is updated periodically or in real time. For each reference search condition, at least one associated search condition can be configured, and the associated search condition corresponding to each reference search condition is updated periodically or in real time.

Taking the reference search condition and the reference search intention as nodes, taking the associated search condition corresponding to the reference search condition and the reference search intention as attribute information of the nodes, and then connecting the nodes corresponding to the plurality of reference search conditions and the plurality of reference search intentions based on the association relation between the reference search condition and the reference search intention which are predicted and output by the target recommendation model to obtain a knowledge graph.

For example, as shown in fig. 11, setting a plurality of reference search conditions includes: reference search word 1, reference search word 2, reference search word 3, reference search word 4 and reference search word 5, wherein reference search word 1 corresponds to associated search word A1 and associated search word B1; the reference search word 2 corresponds to the associated search word A2 and the associated search word B2; the reference search word 3 corresponds to the associated search word A3, the associated search word B3 and the associated search word C3; the reference search term 4 corresponds to the associated search term A4; the reference search term 5 corresponds to the associated search term A5, the associated search term B5, and the associated search term C5.

Inputting the reference search word 1 into the target recommendation model, and determining the reference search intention corresponding to the reference search word 1 as the reference search word 4. And inputting the reference search word 2 into the target recommendation model, and determining the reference search intention corresponding to the reference search word 2 as the reference search word 3. And inputting the reference search word 3 into a target recommendation model, and determining the reference search intention corresponding to the reference search word 3 as the reference search word 2. And inputting the reference search word 4 into the target recommendation model, and determining the reference search intention corresponding to the reference search word 4 as the reference search word 2. Inputting the reference search word 5 into the target recommendation model, and determining the reference search intention corresponding to the reference search word 5 as the reference search word 1.

The reference search word 1, the reference search word 2, the reference search word 3, the reference search word 4 and the reference search word 5 are used as nodes in the knowledge graph, and the associated search words corresponding to the reference search word 1, the reference search word 2, the reference search word 3, the reference search word 4 and the reference search word 5 are used as attribute information of the nodes. And then connecting the nodes corresponding to each reference search condition according to the association relation among each reference search condition output by the target recommendation model to obtain a knowledge graph.

According to the embodiment of the application, the reference search intention corresponding to each reference search condition is predicted through the target recommendation model obtained through training, and the reference search intention corresponding to each reference search condition does not need to be marked manually, so that the efficiency of constructing the knowledge graph is improved. And secondly, determining the target searching intention corresponding to the original searching condition input by the target object by constructing the knowledge graph, and improving the understanding of the searching intention of the target object, thereby improving the accuracy of the searching result.

In the second embodiment, a plurality of reference search conditions are acquired, and the following steps are executed for each of the plurality of reference search conditions: and predicting the search intention of a reference search condition by adopting the target recommendation model to obtain a corresponding reference search intention.

For a plurality of reference search intents, the following steps are respectively performed: and sorting at least one associated search condition corresponding to the reference search intention based on a second historical search record associated with the reference search intention, and obtaining a recommendation sequence corresponding to the at least one associated search condition.

And then constructing a knowledge graph based on the plurality of reference search conditions, the plurality of obtained reference search intents, at least one associated search condition corresponding to each of the plurality of reference search intents and the corresponding recommendation sequence.

Specifically, the second historical search record includes a search process record and a search result view record for the reference search intent, as shown in fig. 12, the search process record including: in the process of searching for the reference search intention in the terminal device, search conditions, search time, position information, account information and other search data recorded by the search server are searched. The search result review record includes: in the process of viewing search results referencing search intents in the terminal device, behavior data recorded by the search server includes, but is not limited to, article clicking operation, clicking time, position information, article titles, article summaries and the like.

Based on the second historical search record, the probability that each associated search condition is searched can be obtained, and then the associated search conditions are ordered according to the order of the probabilities from large to small, so that the recommended order of the associated search conditions is obtained.

Taking the reference search condition and the reference search intention as nodes, taking the associated search condition corresponding to the reference search condition and the reference search intention and the recommendation sequence of the associated search condition as attribute information of the nodes, and then, based on the association relation between the reference search condition and the reference search intention which are predicted and output by the target recommendation model, connecting the nodes corresponding to the plurality of reference search conditions and the plurality of reference search intentions to obtain a knowledge graph.

For example, as shown in fig. 13, setting a plurality of reference search conditions includes: reference search word 1, reference search word 2, reference search word 3, reference search word 4 and reference search word 5, wherein reference search word 1 corresponds to associated search word A1 and associated search word B1, and the recommendation sequence of the associated search words is as follows: associated search word A1, associated search word B1; the reference search word 2 corresponds to the associated search word A2 and the associated search word B2, and the recommendation sequence of the associated search word is as follows: associated search word B2, associated search word A2; the reference search word 3 corresponds to the associated search word A3, the associated search word B3 and the associated search word C3, and the recommendation sequence of the associated search word is as follows: associated search word B3, associated search word A3, associated search word C3; the reference search term 4 corresponds to the associated search term A4; the reference search word 5 corresponds to the associated search word A5, the associated search word B5 and the associated search word C5, and the recommendation sequence of the associated search word is as follows: associated search term B5, associated search term A5, and associated search term C5.

The reference search word 1, the reference search word 2, the reference search word 3, the reference search word 4 and the reference search word 5 are used as nodes in the knowledge graph, and the associated search word corresponding to each of the reference search word 1, the reference search word 2, the reference search word 3, the reference search word 4 and the reference search word 5 and the recommendation sequence of the associated search word are used as attribute information of the nodes. And then connecting the nodes corresponding to each reference search condition according to the association relation among each reference search condition output by the target recommendation model to obtain a knowledge graph.

In the embodiment of the application, based on the second historical search record associated with the reference search intention, at least one associated search condition corresponding to the reference search intention is sequenced to obtain the recommended sequence corresponding to the at least one associated search condition, and the recommended sequence is added to the knowledge graph as the attribute information of the node, so that when the associated search condition is recommended to the target object according to the knowledge graph, the associated search condition suitable for the target object can be recommended preferentially, and the accuracy and efficiency of searching are improved.

Optionally, before the knowledge graph is constructed, training is required to be performed on the target object model, and the training process includes the following steps:

A plurality of sample search conditions and corresponding tag search intents are obtained. And then adopting a plurality of sample search conditions and corresponding mark search intents to carry out iterative training on the recommended model to be trained until the iteration stopping condition is met, and obtaining a target recommended model, wherein each iterative process comprises the following steps:

based on the sample search conditions, a corresponding predicted search intent is determined. And determining a target loss value based on the predicted search intention and the marked search intention, and carrying out parameter adjustment on the recommendation model to be trained through the target loss value.

Specifically, the sample search conditions are subjected to data marking in advance, and the marked search intention corresponding to the sample search conditions is obtained. If the label search intention obtained by labeling is partial or concentrated, the sample search conditions can be clustered based on semantic vectors of the sample search conditions, so that the repeated labeling quantity of similar data is reduced. After a certain number of tagged search intents are obtained, the tagged data may be sample augmented in a nearest neighbor manner or a data enhancement manner.

In the model training, some critical parameters need to be set, and the effect of the model is directly affected by the good and bad setting of the parameters, so in the embodiment of the application, the values of the critical parameters are obtained by adopting a grid searching mode. In addition, when model training is carried out, the regular term coefficients are set to prevent overfitting, so that the generalization capability of the model is improved.

The target loss value is used to characterize the difference between the predicted and tagged search intents, and model training is to minimize the difference between the true predicted and tagged search intents. And determining whether the recommendation model is converged according to the target loss value. And when the recommendation model is determined not to be converged, adjusting model parameters of the recommendation model according to the target loss value. And training the next round through the recommendation model after parameter adjustment. And when the recommendation model is determined to be converged, finishing training, and outputting a trained target recommendation model. Of course, in the embodiment of the present application, when the number of iterative training times reaches the preset number of times, training may be ended, and a trained target recommendation model may be output.

According to the embodiment of the application, the reference search intention corresponding to each reference search condition is predicted through the target recommendation model obtained through training, and the reference search intention corresponding to each reference search condition does not need to be marked manually, so that the efficiency and the accuracy of constructing the knowledge graph are improved.

Optionally, after the knowledge graph is constructed, the embodiment of the present application determines the corresponding target search intention according to the original search condition at least in the following manner, and obtains at least one recommended search condition corresponding to the target search intention:

And inquiring a preset knowledge graph based on the original search condition. When it is determined that a reference search condition matching the original search condition exists in the knowledge graph, taking a reference search intention corresponding to the reference search condition in the knowledge graph as a target search intention, and taking at least one associated search condition corresponding to the reference search intention as at least one recommended search condition.

For example, based on the knowledge graph shown in fig. 11 of the original search term query, the query results are: the original search word is matched with the reference search word 1 in the knowledge graph, the reference search intention corresponding to the reference search word 1 in the knowledge graph (namely the reference search word 4) is used as a target search intention, and the associated search word A4 corresponding to the reference search word 4 is used as a recommended search word of the target search intention.

According to the method and the device, the search conditions and the corresponding search intents are mapped through the knowledge graph, so that when the original search conditions input by the target object are received, the corresponding target search intents can be obtained through the knowledge graph, at least one recommended search condition corresponding to the target search intents is obtained from the knowledge graph, the target object can be searched based on the recommended search conditions, and the accuracy and the efficiency of searching are improved.

Optionally, at least one recommended search condition in the search recommendation interface is displayed according to a recommendation sequence corresponding to the at least one recommended search condition. In the search recommendation interface, other recommendation search conditions corresponding to the target search intention are displayed according to respective recommendation orders of the other recommendation search conditions, wherein a method for obtaining the recommendation order corresponding to at least one recommendation search condition is the same as a method for obtaining the recommendation order corresponding to the other recommendation search conditions, and for obtaining respective recommendation order expansion of the at least one recommendation search condition, the present application provides at least the following embodiments for obtaining respective recommendation orders of the at least one recommendation search condition:

in the first embodiment, a recommendation sequence corresponding to at least one recommendation search condition is obtained from the knowledge graph.

For example, based on the knowledge graph shown in the original search term query fig. 13, the query results are: the original search word is matched with the reference search word 2 in the knowledge graph, the reference search intention corresponding to the reference search word 2 in the knowledge graph (namely the reference search word 3) is taken as a target search intention, and the associated search word A3, the associated search word B3 and the associated search word C3 corresponding to the reference search word 3 are taken as recommended search words of the target search intention. Meanwhile, the recommendation sequence corresponding to each recommendation search word can be obtained: associated search term B3, associated search term A3, and associated search term C3.

In the second embodiment, based on a first historical search record of a target object associated with an original search condition, search preference characteristics of the target object are determined, and based on the search preference characteristics, at least one recommended search condition is subjected to recommendation ordering, so that a recommendation sequence corresponding to the at least one recommended search condition is obtained.

Specifically, the first historical search record includes a search process record and a search result view record of the target object, wherein the search process record includes: search data such as search words, search time, position information, account information and the like recorded by the target object in each search process. The search result review record includes: behavior data recorded by the target object during each browsing of the search results, including but not limited to article clicking operation, click time, location information, article titles, article summaries, and the like.

For example, setting recommended search terms includes: "lean wearing skill", "lean makeup skill", "lean dress", "lean pants". Based on the first historical search record of the target object, the search preference characteristic of the target object is content related to 'wearing', the recommended search words with strong relevance to 'wearing' are arranged at the front position, the recommended search words with weak relevance to 'wearing' are arranged at the rear position, and the obtained recommendation sequence is as follows: "lean wearing skill", "lean dress", "lean pants", "lean make-up skill".

In the embodiment of the application, after determining the target search intention and at least one recommended search condition corresponding to the target search intention, acquiring the recommended sequence of the at least one recommended search condition from the knowledge graph, or determining the recommended sequence of the at least one recommended search condition based on the search preference characteristics of the target object, and displaying the at least one recommended search condition according to the recommended sequence so as to recommend the search condition which is more matched with the target object to the target object preferentially, thereby improving the accuracy and efficiency of the search.

In order to better explain the embodiments of the present application, a search method provided by the embodiments of the present application is described below in connection with a specific implementation scenario, where a flow of the method may be performed by the terminal device 101 shown in fig. 1 and the search server 102, and the method includes the following steps, as shown in fig. 14:

in step S1401, the terminal device acquires the original search term in response to the search operation triggered for the original search condition in the search interface.

Specifically, the search application displays a search interface, as shown in FIG. 2, which includes a search box, a "ok" button. After the original search word weight is input in the search box, the determination button is clicked, and the terminal equipment responds to the search operation triggered in the search application to acquire the original search word weight.

Step S1402, the terminal device sends the original search word to the search server.

Step S1403, the search server obtains a target search intention based on the original search term query knowledge graph.

Specifically, the search server obtains the target search intention of "lean display" based on the original search term of "lean" query knowledge graph.

In step S1404, the search server obtains all the recommended search words corresponding to the target search intention and the recommendation sequence of each recommended search word from the knowledge graph.

Specifically, the search server acquires each recommended search word corresponding to the target search intention 'lean' from the knowledge graph: "lean wearing skill", "lean makeup skill", "lean dress", "lean pants", recommended order of individual recommended search terms: "lean wearing skill", "lean dress", "lean pants", "lean make-up skill".

In step S1405, the search server transmits the target search intention, each recommended search term, and the corresponding recommendation order to the terminal device.

In step S1406, the terminal device displays the association between the original search term and the target search intention in the search recommendation interface, and displays part of the recommended search terms according to the recommendation order.

Specifically, as shown in fig. 4, the search recommendation interface includes an association relationship between an original search term "weight-losing" and a target search intention "lean", and recommended search terms preferentially displayed according to a recommendation sequence: "lean wearing skill", "lean dress", "lean pants". The original search term "weight-loss" is connected to the recommended region of the target search intention "lean" by means of a wire. Recommending search terms: the thin wearing skill, the thin dress and the thin trousers are connected to the recommended area of the target search intention, namely the thin area by a connecting line.

In step S1407, the terminal device updates the presented recommended search term in response to the scrolling operation triggered for the target search intention in the search recommendation interface.

Specifically, the terminal device responds to a scrolling operation triggered by a recommendation area aiming at a target search intention in a search recommendation interface, and updates recommended search words displayed on the search recommendation interface, namely, the recommended search words are displayed: "lean dress", "lean pants", "lean make-up skills", as shown in FIG. 5, recommended search terms: the "lean dress", "lean trousers", "lean cosmetic skills" are connected by means of a wire to the recommended area of the target search intention "lean".

In step S1408, the terminal device selects a target search term from the presented recommended search terms in response to the search condition selection operation triggered in the search recommendation interface.

Specifically, the terminal device selects a target search word "lean cosmetic skills" from recommended search words "lean dress", "lean pants", "lean cosmetic skills" in response to a click operation triggered at the search recommendation interface.

In step S1409, the terminal device sends the target search word to the search server.

In step S1410, the search server obtains the target search result from the search library based on the target search word.

In step S1411, the search server transmits the target search result to the terminal device.

Specifically, the search server acquires a plurality of articles associated with the target search word "lean make-up skill" from the search library, and then transmits related information of the plurality of articles to the terminal device.

In step S1412, the terminal device displays the target search result in the search result interface.

Specifically, the terminal device displays a search result interface, and specifically as shown in fig. 7, the search result interface includes an article information display area 701 and an article information display area 702, where the article information display area 701 includes an article title of "displaying a thin makeup course" and a corresponding article thumbnail, and the article information display area 702 includes an article title of "5 minutes to draw a thin makeup look" and a corresponding article thumbnail.

Based on the same technical concept, the embodiment of the present application provides a schematic structural diagram of a search apparatus, as shown in fig. 15, the search apparatus 1500 includes:

a recommendation module 1501 for displaying a search recommendation interface in response to a search operation triggered based on an original search condition in the search interface, the search recommendation interface comprising: a target search intention corresponding to the original search condition, and at least one recommended search condition corresponding to the target search intention;

the recommendation module 1501 is further configured to respond to a scrolling operation triggered by the target search intention in the search recommendation interface, and display other recommended search conditions corresponding to the target search intention in the search recommendation interface;

the search module 1502 is configured to respond to a search condition selection operation triggered in the search recommendation interface for each presented recommended search condition, and present a target search result corresponding to the selected target recommended search condition.

Optionally, the recommendation module 1501 is specifically configured to:

Optionally, the search module 1502 is specifically configured to:

Optionally, the recommendation module 1501 is further configured to:

Optionally, the recommendation module 1501 is specifically configured to:

inquiring a preset knowledge graph based on the original search condition;

the recommendation module 1501 is further configured to:

Optionally, the recommendation module 1501 is further configured to:

Optionally, the recommendation module 1501 is specifically configured to:

Optionally, the recommendation module 1501 is further configured to:

Based on the same technical concept, the embodiment of the present application provides a computer device, which may be the terminal device and/or the search server shown in fig. 1, as shown in fig. 16, including at least one processor 1601, and a memory 1602 connected to the at least one processor, where a specific connection medium between the processor 1601 and the memory 1602 is not limited in the embodiment of the present application, and in fig. 16, the connection between the processor 1601 and the memory 1602 is exemplified by a bus. The buses may be divided into address buses, data buses, control buses, etc.

In the embodiment of the present application, the memory 1602 stores instructions executable by the at least one processor 1601, and the at least one processor 1601 may perform the steps of the search method described above by executing the instructions stored in the memory 1602.

Wherein the processor 1601 is a control center of the computer device, various interfaces and lines may be utilized to connect various portions of the computer device, and to perform information searching by executing or executing instructions stored in the memory 1602 and invoking data stored in the memory 1602. Alternatively, the processor 1601 may include one or more processing units, and the processor 1601 may integrate an application processor primarily handling operating systems, user interfaces, application programs, etc., with a modem processor primarily handling wireless communications. It will be appreciated that the modem processor described above may not be integrated into the processor 1601. In some embodiments, the processor 1601 and the memory 1602 may be implemented on the same chip, and in some embodiments they may be implemented separately on separate chips.

The processor 1601 may be a general purpose processor such as a Central Processing Unit (CPU), digital signal processor, application specific integrated circuit (Application Specific Integrated Circuit, ASIC), field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, and may implement or perform the methods, steps, and logic blocks disclosed in embodiments of the present application. The general purpose processor may be a microprocessor or any conventional processor or the like. The steps of a method disclosed in connection with the embodiments of the present application may be embodied directly in a hardware processor for execution, or in a combination of hardware and software modules in the processor for execution.

Memory 1602 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. Memory 1602 may include at least one type of storage medium, which may include, for example, flash Memory, hard disk, multimedia card, card Memory, random access Memory (Random Access Memory, RAM), static random access Memory (Static Random Access Memory, SRAM), programmable Read-Only Memory (Programmable Read Only Memory, PROM), read-Only Memory (ROM), electrically erasable programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory), magnetic Memory, magnetic disk, optical disk, and the like. Memory 1602 is any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer device, but is not limited to such. The memory 1602 in the present embodiment may also be a circuit or any other device capable of implementing a memory function for storing program instructions and/or data.

Based on the same inventive concept, embodiments of the present application provide a computer-readable storage medium storing a computer program executable by a computer device, which when run on the computer device, causes the computer device to perform the steps of the above-described search method.

Based on the same inventive concept, embodiments of the present application provide a computer program product comprising a computer program stored on a computer readable storage medium, the computer program comprising program instructions which, when executed by a computer device, cause the computer device to perform the steps of the above-described search method.

It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, or as a computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.

It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.

Claims

1. A search method, comprising:

2. The method of claim 1, wherein the presenting, in the search recommendation interface, other recommended search conditions corresponding to the target search intent comprises:

3. The method of claim 1, wherein the presenting the target search results corresponding to the selected target recommended search criteria in response to a search criteria selection operation triggered in the search recommendation interface for each presented recommended search criteria comprises:

4. The method of claim 1, wherein the presenting the target search results corresponding to the selected target recommended search criteria in response to a search criteria selection operation triggered in the search recommendation interface for each presented recommended search criteria comprises:

5. The method of claim 1, wherein the search recommendation interface further comprises: the original search condition and a connecting line between the original search condition and the target search intention are used for representing the association relation between the original search condition and the target search intention.

6. The method of claim 1, wherein prior to displaying the search recommendation interface, further comprising:

based on a preset knowledge graph, determining a corresponding target search intention according to the original search condition, and acquiring at least one recommended search condition corresponding to the target search intention.

7. The method of claim 6, wherein the determining the corresponding target search intention based on the original search condition and obtaining at least one recommended search condition corresponding to the target search intention based on the preset knowledge graph comprises:

inquiring a preset knowledge graph based on the original search condition;

8. The method of claim 6, wherein at least one recommended search condition in the search recommendation interface is presented in a recommended order corresponding to the at least one recommended search condition;

before the search recommendation interface is displayed, the method further comprises the following steps:

acquiring a recommendation sequence corresponding to the at least one recommendation search condition from the knowledge graph; or alternatively, the process may be performed,

9. The method of claim 1, wherein prior to the obtaining the original search criteria, further comprising:

for a plurality of reference search conditions, the following steps are respectively executed: predicting search intention of a reference search condition by adopting a target recommendation model to obtain corresponding reference search intention;

10. The method of claim 9, wherein the constructing the knowledge graph based on the plurality of reference search conditions, the plurality of obtained reference search intents, and at least one reference search condition to which the plurality of reference search intents each corresponds comprises:

11. The method of claim 10, wherein the target recommendation model is obtained by training in the following manner:

12. A search apparatus, comprising:

13. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any of claims 1-11 when the program is executed.

14. A computer readable storage medium, characterized in that it stores a computer program executable by a computer device, which program, when run on the computer device, causes the computer device to perform the steps of the method according to any one of claims 1-11.

15. A computer program product comprising a computer program stored on a computer readable storage medium, the computer program comprising program instructions which, when executed by a computer device, cause the computer device to perform the method steps of any of claims 1-11.