CN110929176B - Information recommendation method, device and electronic equipment - Google Patents

Abstract

The embodiment of the invention discloses an information recommendation method and device and electronic equipment. According to the method, keywords input in a retrieval interface are firstly obtained, functional words to be retrieved, which can represent categories or characteristics of POIs, are extracted from the keywords, then the POIs corresponding to the functional words to be retrieved are obtained according to the corresponding relation between the functional words and the POIs, in order to accurately recommend information, the POIs extracted according to the functional words to be retrieved are set as candidate POIs, then the association degree between the functional words to be retrieved and each candidate POI is calculated, finally the candidate POIs are selected and recommended according to the size of the association degree, the technical problem that the fit degree between the recommendation of the POIs and the retrieval intention of a user is not high in the prior art is solved, and the accuracy of the recommendation of the POIs is improved.

Description

Information recommendation method, device and electronic equipment

Technical Field

The present invention relates to the field of information technology, and in particular to an information recommendation method, device and electronic equipment.

Background Art

With the continuous development of network technology, more and more people are searching for network information through the Internet. For example, people use electronic maps to search for geographic location information, search for web pages containing information corresponding to keywords in browsers, and so on.

At present, when searching for corresponding web page information through keywords, relevant information is usually recommended based on information such as the popularity of the web page search, clicks, whether it participates in promotion, and the number of times the entered keywords appear in the web page. When searching for points of interest (POI) through keywords, the retrieved POIs are recommended based on the distance between the POI and the current location, the popularity of the POI, whether it participates in promotion, etc. Regardless of whether the POIs are recommended based on distance, popularity, or promotion, many times users still cannot obtain the desired POIs, or cannot obtain the desired POIs from the recommended homepage, that is, there is a technical problem in the prior art that the recommendation of POIs is not highly consistent with the user's search intention.

Summary of the invention

The embodiments of the present invention provide an information recommendation method, device and electronic device, which are used to solve the technical problem in the prior art that the recommendation of points of interest (POI) is not highly consistent with the search intention of the user.

In a first aspect, an embodiment of the present invention provides an information recommendation method, the method comprising:

Get the keywords entered in the search interface;

Extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2;

Acquire the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and acquire the association degree between the function word to be retrieved and each candidate point of interest POI;

The candidate point of interest (POI) is selected and recommended according to the magnitude of the association degree.

Optionally, extracting the function words to be searched from the keywords includes:

Extracting a target keyword from the keywords, wherein the target keyword can represent the search intent of the keyword;

The target keyword is used as the function word to be searched.

Optionally, the method further includes:

Establishing a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI, or establishing a correspondence between each point of interest POI and the function word according to the search log of each point of interest POI;

The step of obtaining the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI includes:

According to the correspondence between each point of interest POI and the function word, the point of interest POI corresponding to the function word to be retrieved is obtained as a candidate point of interest POI.

Optionally, the establishing a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI includes:

Extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI;

The category word and/or feature word of each point of interest POI is used as the function word of each point of interest POI, and a corresponding relationship between each point of interest POI and the function word is established.

Optionally, the establishing a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI includes:

Obtain the name, category, features and/or consumption information of each point of interest (POI) contained in the page content corresponding to each point of interest (POI), wherein the consumption information is used to reflect the consumption of the user at the point of interest (POI);

Cluster the name, category, features and/or consumption information of each POI to obtain the functional words of each POI;

The function words of each point of interest (POI) are input into a convolutional neural network, and the convolutional neural network is trained. The corresponding relationship between each point of interest (POI) and the function words is represented by the trained convolutional neural network, wherein the trained convolutional neural network can convert the input function words to be retrieved into candidate points of interest (POI).

Optionally, the establishing a correspondence between each point of interest POI and a function word according to the search log of each point of interest POI includes:

Obtain target keywords contained in the search log of each point of interest (POI);

The target keyword is used as a function word of the point of interest (POI) corresponding to the search log.

Optionally, obtaining the correlation between the function word to be retrieved and each candidate point of interest POI includes:

According to all function words corresponding to each candidate point of interest POI, each candidate point of interest POI is converted into a first word vector, and the function words to be retrieved are converted into a second word vector;

The similarity between the first word vector and the second word vector is calculated, and the similarity is used as the association between the function word to be retrieved and each candidate point of interest POI.

Optionally, the selecting and recommending the candidate point of interest POI according to the magnitude of the association degree includes:

Acquire the candidate points of interest POIs whose relevance is greater than a set threshold as target points of interest POIs, and recommend the target points of interest POIs; or

The first m candidate points of interest (POIs) with the greatest correlation are obtained as target points of interest (POIs), and the target points of interest (POIs) are recommended, where m≥1.

In a second aspect, an embodiment of the present invention further provides an information recommendation device, the device comprising:

The keyword acquisition module is used to obtain the keywords entered in the search interface;

A function word extraction module is used to extract the function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest POI, and one point of interest POI corresponds to n function words, where n≥2;

A POI acquisition module is used to acquire the POI corresponding to the function word to be retrieved as a candidate POI, and acquire the correlation between the function word to be retrieved and each candidate POI;

The POI recommendation module is used to select and recommend the candidate POI according to the magnitude of the association.

Optionally, the function word extraction module is used to:

Extracting a target keyword from the keywords, wherein the target keyword can represent the search intent of the keyword;

The target keyword is used as the function word to be searched.

Optionally, the device further comprises:

A corresponding relationship establishing module is used to establish a corresponding relationship between each point of interest POI and a function word according to the function word contained in the page content corresponding to each point of interest POI, or to establish a corresponding relationship between each point of interest POI and a function word according to the search log of each point of interest POI;

The module for obtaining points of interest (POI) is used for:

According to the correspondence between each point of interest POI and the function word, the point of interest POI corresponding to the function word to be retrieved is obtained as a candidate point of interest POI.

Optionally, the corresponding relationship establishing module is used for:

Extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI;

The category word and/or feature word of each point of interest POI is used as the function word of each point of interest POI, and a corresponding relationship between each point of interest POI and the function word is established.

Optionally, the corresponding relationship establishing module is further used for:

Obtain the name, category, features and/or consumption information of each point of interest (POI) contained in the page content corresponding to each point of interest (POI), wherein the consumption information is used to reflect the consumption of the user at the point of interest (POI);

Cluster the name, category, features and/or consumption information of each POI to obtain the functional words of each POI;

The function words of each point of interest (POI) are input into a convolutional neural network, and the convolutional neural network is trained. The corresponding relationship between each point of interest (POI) and the function words is represented by the trained convolutional neural network, wherein the trained convolutional neural network can convert the input function words to be retrieved into candidate points of interest (POI).

Optionally, the corresponding relationship establishing module is further used for:

Obtain target keywords contained in the search log of each point of interest (POI);

The target keyword is used as a function word of the point of interest (POI) corresponding to the search log.

Optionally, the module for obtaining points of interest (POI) is further used for:

According to all function words corresponding to each candidate point of interest POI, each candidate point of interest POI is converted into a first vector word, and the function words to be retrieved are converted into a second vector word;

The similarity between the first vector word and the second vector word is calculated, and the similarity is used as the association between the function word to be retrieved and each candidate point of interest POI.

Optionally, the POI recommendation module is used to:

Acquire the candidate points of interest (POI) whose relevance is greater than a set threshold as target points of interest (POI), and recommend the target points of interest (POI);

The first m candidate points of interest (POIs) with the greatest correlation are obtained as target points of interest (POIs), and the target points of interest (POIs) are recommended, where m≥1.

In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which performs the following steps when executed by a processor:

Get the keywords entered in the search interface;

Extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2;

Acquire the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and acquire the association degree between the function word to be retrieved and each candidate point of interest POI;

The candidate point of interest (POI) is selected and recommended according to the magnitude of the association degree.

In a fourth aspect, an embodiment of the present invention provides an electronic device, comprising a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, wherein the one or more programs include instructions for performing the following operations:

Get the keywords entered in the search interface;

Extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2;

Acquire the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and acquire the association degree between the function word to be retrieved and each candidate point of interest POI;

The candidate POI is selected and recommended according to the magnitude of the correlation. The above one or more technical solutions in the embodiments of the present application have at least the following technical effects:

The present invention provides an information recommendation method, device and electronic device, the method comprising: obtaining keywords input in a search interface; extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of a point of interest (POI), and one point of interest (POI) corresponds to n function words, n≥2; obtaining the point of interest (POI) corresponding to the function words to be searched as candidate point of interest (POI), and obtaining the correlation between the function words to be searched and each candidate point of interest (POI); and selecting and recommending the candidate point of interest (POI) according to the size of the correlation. Since one point of interest POI corresponds to multiple function words, and one function word can also correspond to multiple points of interest POIs, the points of interest POIs extracted according to the function words to be retrieved are complicated. In order to accurately recommend POIs, the points of interest POIs extracted according to the function words to be retrieved are set as candidate points of interest POIs, and then the correlation between the function words to be retrieved and each candidate point of interest POI is obtained, and the candidate points of interest POIs are selected and recommended according to the size of the correlation, so that the points of interest POIs with a large correlation with the function words to be retrieved can be recommended first, and the function words to be retrieved are extracted from the search keywords, which are in line with the user's search intent, that is, the points of interest POIs with a large correlation with the function words to be retrieved are more in line with the user's search intent than the POIs with a small correlation. Therefore, recommending POIs based on the size of the correlation between the function words to be retrieved and the points of interest POIs solves the technical problem in the prior art that the recommendation of the points of interest POIs is not highly consistent with the user's search intent, and improves the accuracy of the recommendation of the points of interest POIs.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 shows a flow chart of an information recommendation method provided by an embodiment of the present invention.

FIG. 2 shows a schematic diagram of an information recommendation device provided by an embodiment of the present invention.

Fig. 3 is a block diagram of an electronic device according to an exemplary embodiment.

FIG. 4 is a schematic diagram of the structure of a server in an embodiment of the present invention.

DETAILED DESCRIPTION

The main implementation principles, specific implementation methods and corresponding beneficial effects of the technical solutions of the embodiments of the present application are described in detail below with reference to the accompanying drawings.

Example

Please refer to FIG1 , which shows a flow chart of an information recommendation method provided by an embodiment of the present invention. An embodiment of the present invention provides an information recommendation method, which can be applied to a client and a server. The information recommendation method includes the following steps:

S100: Acquire keywords input in the search interface;

S200: extracting function words to be searched from the keywords, wherein the function words are used to represent the category or feature of the point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2;

S300: obtaining a point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and obtaining a correlation between the function word to be retrieved and each candidate point of interest POI;

S400: Select and recommend the candidate point of interest (POI) according to the magnitude of the association.

In S100, the search interface may be a search interface corresponding to a search box of an application such as a browser or a map, or a search interface provided by a system or application of an electronic device when some information is selected. In addition, the keywords input into the search interface may be input by a user, or may be automatically generated or selected by a system or application of an electronic device.

Keywords refer specifically to the words used by a single media when creating and using an index. You can enter multiple keywords here.

Generally, the accuracy of extracting search keywords has a great impact on the accuracy of search output results. Extracting an accurate keyword can fundamentally improve the accuracy of the search. For the embodiment of the present invention, accurate function words to be searched are extracted from the keywords input in the search interface, and the POI corresponding to the function words to be searched can be obtained. The POI obtained can better meet the user's search intention.

In order to obtain accurate function words to be searched, the embodiment of the present invention adopts the scheme described in the following step S200 to obtain the function words to be searched.

S200: extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of a point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2.

Point of Interest (POI) is also called Point of Information (Point of Information). POI is generally represented by a bubble icon on electronic maps. Tourist attractions, government agencies, companies, shopping malls, restaurants, etc. on electronic maps are all POIs. POI is the core data of location-based services and is widely used in electronic maps, such as selecting a destination before navigation and viewing nearby restaurants.

Function words are used to characterize the category or characteristics of a POI. The category or characteristics of a POI can be described by multiple functions. POIs of different categories can have the same characteristics, and POIs with different characteristics can also have the same or similar categories. Therefore, it can be understood that a POI can correspond to multiple function words, and a function word can also correspond to multiple POIs.

For example, the function word "roast duck" can correspond to the store corresponding to the brand "Quanjude", that is, the POI "Quanjude" can correspond to the function word "roast duck" or the function word "Beijing roast duck". The function word corresponding to the POI "Shangri-La Hotel" can be "accommodation", "hotel", etc.

Extracting the function words to be searched from the keywords can be obtained by the following steps: first, extracting target keywords from the keywords, wherein the target keywords can represent the search intent of the keywords; then, using the target keywords as the function words to be searched. It should be noted that the acquisition of target keywords is not limited to the number thereof, and can be one or more.

There may be multiple keywords obtained from the search interface, and one or more keywords can be extracted from them as target keywords. As an implementation method, the step of extracting the target keywords from the keywords can be specifically: input the keywords into a pre-trained CNN (Convolutional Neural Network), and the CNN outputs the target keywords corresponding to the keywords, and the target keywords can represent the search intention of the keywords, wherein the CNN is pre-trained with a sample set consisting of a keyword and the corresponding target keyword, and the trained CNN can convert the input keywords into output target keywords. Among them, CNN is a feedforward neural network, and artificial neurons can respond to surrounding units, can perform large-scale information processing, and feedforward correction of the input weights in the network is performed according to the input samples. The final CNN can reflect one or more outputs corresponding to the input information.

As another implementation method, the step of extracting target keywords from the keywords is specifically as follows: the keywords are input into a pre-trained decision tree model, and the decision tree model outputs target keywords corresponding to the keywords, and the target keywords can represent the retrieval intent of the keywords, wherein the decision tree model is pre-trained using a sample set consisting of a keyword and the corresponding target keyword, and the trained decision tree model can convert the input keywords into output target keywords.

The above two implementations are both based on machine learning to extract target keywords from keywords. As another implementation, nouns or verbs in the keywords can be extracted from the input keywords and the nouns or verbs can be used as target keywords.

There is a corresponding relationship between function words and POIs. The target keyword can represent the search intent of the input keyword. Therefore, the target keyword obtained from the keyword is used as the function word to be searched, and then the POIs are recommended based on the function word to be searched. The recommended POIs can be more in line with the user's search intent, reducing errors from the root. The recommended POIs are more in line with the user's search intent.

In the embodiment of the present invention, before S300, the information recommendation method further includes establishing a correspondence between each point of interest POI and a function word. The establishment of a correspondence between each point of interest POI and a function word can be achieved in a variety of ways, specifically, it can be achieved in the following ways:

As an implementation method of establishing the correspondence between each point of interest POI and a function word, the correspondence between each point of interest POI and a function word may be established according to the function word contained in the page content corresponding to each point of interest POI.

The page content corresponding to the point of interest POI refers to the content contained in the page where the web page or map interface corresponding to the point of interest POI is located. The content may include text, pictures, videos, audio, etc. Specifically, it may include registration information, classification information, comment information, etc. of the point of interest POI.

In order to establish a corresponding relationship between each point of interest POI and a function word according to the function word contained in the page content corresponding to each point of interest POI, specifically, the following method can be used to implement it:

As a first implementation, the corresponding relationship between each point of interest POI and the function word can be established in the following manner:

Extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI; use the category words and/or feature words of each point of interest POI as the function words of each point of interest POI, and establish a corresponding relationship between each point of interest POI and the function words. Specifically, the method of extracting the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI can be as follows: the features in the classification information, registration information, and comment information can be used as the category words and/or feature words of the point of interest POI, or the method described in step 210 can be used to extract the category words and/or feature words of the point of interest POI from the registration information and comment information.

As a second implementation, the corresponding relationship between each point of interest POI and the function word may be established in the following manner:

First, the name, category, features and/or consumption information of each POI contained in the page content corresponding to the POI is obtained, wherein the consumption information is used to reflect the consumption of the user at the POI. The consumption situation includes the per capita consumption price, the number of dishes ordered, the number of netizen comments, the number of likes, etc.

In an embodiment of the present invention, the name, category, feature and/or consumption information of the POI can be obtained by searching on the Internet, or the name, category, feature and/or consumption information of the POI can be obtained from a pre-established database containing the name, category, feature and/or consumption information of the POI, wherein the pre-established database can be composed of the name, category, feature and/or consumption information of the POI obtained from the data interface of a partner, or can be composed of the name, category, feature and/or consumption information of the POI collected in advance on the Internet.

Secondly, the name, category, feature and/or consumption information of each POI are clustered to obtain the functional words of each POI.

A decision tree model, a convolutional neural network, a SVM (Support Vector Machine) model, or other machine learning models can be used to learn and cluster the names, categories, features, and/or consumption information of the POIs to obtain the functional words of each POI. Specifically, a sample set consisting of the names, categories, features, and/or consumption information of the POIs is collected in advance, and the sample set is input into the machine learning model, and the machine learning model is trained. The trained machine learning model can convert the names, categories, features, and/or consumption information of the input POIs into functional words corresponding to the POIs.

Then, after obtaining the function words of each POI, the corresponding relationship between the function words and the POI is established. Specifically, the function words of each POI are input into the convolutional neural network, and the convolutional neural network is trained. The trained convolutional neural network can characterize the corresponding relationship between each POI and the function words, that is, the trained convolutional neural network can convert the input function words to be retrieved and output candidate POIs. For example, input keywords such as "ice cream, fried chicken, hamburger, cola" into the convolutional neural network, and the convolutional neural network can output the POI corresponding to "KFC".

As another implementation method of establishing the correspondence between each point of interest POI and a function word, the correspondence between each point of interest POI and a function word may be established according to the search log of each point of interest POI. Specifically, it may be implemented in the following manner:

First, the target keywords contained in the search log of each point of interest POI are obtained. The target keywords obtained may be one or more, and the present embodiment does not limit the number of target keywords. One target keyword corresponds to one function word, and one function word may correspond to multiple points of interest POIs.

Here, the search log of each POI includes the keywords entered in the previous search history, the target keywords corresponding to the keywords, and the user selection results corresponding to the target keywords. For example, the user clicks on a search result, and the user selection result corresponds to a POI. For example, the search log records the following format: target keyword <-> user selection result <-> POI.

Next, after obtaining the search log of the POI, the target keyword can be found along the path of point of interest POI <-> user selection result <-> target keyword.

Then, the acquired target keyword is used as a function word of the point of interest POI corresponding to the search log, and a corresponding relationship between the point of interest POI and the function word is established.

Of course, the correspondence between the POI and the function words can also be established by training a classification model. A large number of POIs and corresponding target keywords are used as training samples to train the classification model, so that the classification model can classify the target keywords and the corresponding POIs well. When the target keyword is obtained, the target keyword is input into the trained classification model, and the POI corresponding to the target keyword can be output.

S300: Acquire a point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and acquire a correlation between the function word to be retrieved and each candidate point of interest POI.

Specifically, the candidate POIs may be obtained by obtaining the POIs corresponding to the function words to be retrieved as the candidate POIs according to the correspondence between each POI and the function words.

The acquisition of the correlation between the function word to be retrieved and each candidate point of interest POI can be achieved in the following way:

According to all the function words corresponding to each candidate point of interest POI, each candidate point of interest POI is converted into a first word vector, and the function words to be retrieved are converted into a second word vector; the similarity between the first word vector and the second word vector is calculated, and the similarity is used as the association between the function words to be retrieved and each candidate point of interest POI. The angle between the first word vector and the second word vector can be calculated, and then the cosine value of the angle is calculated, and the similarity between the first word vector and the second word vector is measured by the cosine value. The larger the cosine value, the higher the similarity between the first word vector and the second word vector, and the smaller the cosine value, the lower the similarity between the first word vector and the second word vector.

Step S400: selecting and recommending the candidate point of interest (POI) according to the magnitude of the association degree.

When the number of obtained candidate points of interest POIs reaches a certain value, it is unrealistic and unnecessary to recommend these candidate points of interest POIs. Therefore, it is necessary to selectively recommend candidate points of interest POIs.

Selecting and recommending the candidate POI according to the magnitude of the association can be specifically achieved in the following manner:

As an implementation manner, the candidate points of interest POIs whose correlation is greater than a set threshold are obtained as target points of interest POIs, and the target points of interest POIs are recommended.

As another implementation manner, the first m candidate points of interest POIs with the greatest correlation are obtained as target points of interest POIs, and the target points of interest POIs are recommended, where m≥1.

In this way, candidate POIs with relatively large correlation can be recommended preferentially, and the recommended POIs are more in line with the user's search intention. On the one hand, this meets the needs of users, and on the other hand, there is no need to recommend all the acquired POIs, thus saving resources.

By adopting the above scheme, firstly, by obtaining the keywords input in the search interface, the function words to be retrieved that can represent the category or characteristics of the POI are extracted from the keywords, and then the POI corresponding to the function words to be retrieved is obtained according to the corresponding relationship between the function words and the POI. Since one POI corresponds to multiple function words, and one function word can also correspond to multiple POIs, the POIs extracted according to the function words to be retrieved are complicated. In order to accurately recommend information, the POIs extracted according to the function words to be retrieved are set as candidate POIs. Then, the association between the function words to be retrieved and each candidate POI is obtained. In order to intuitively and accurately obtain the association between the function words to be retrieved and each candidate POI, each candidate POI is converted into a first word vector, and the function words to be retrieved are converted into a second word vector, and the similarity between the first word vector and the second word vector is calculated, and the similarity is used as the association between the function words to be retrieved and each candidate POI. Finally, the candidate POIs are selected for recommendation based on the degree of correlation, so that the POIs with a high degree of correlation with the function words to be retrieved can be recommended first, and the function words to be retrieved are extracted from the search keywords, which conform to the user's search intention, that is, the POIs with a high degree of correlation with the function words to be retrieved are more in line with the user's search intention than the POIs with a low degree of correlation. Therefore, the POIs are recommended based on the degree of correlation between the function words to be retrieved and the POIs, which solves the technical problem that the recommendation of the POIs in the prior art is not highly consistent with the user's search intention, and improves the accuracy of the POI recommendation. In view of the above-mentioned embodiment, an information recommendation method is provided, and the embodiment of the present application also provides an information recommendation device 200, which is applied to a client or server for information recommendation. Please refer to Figure 2, the device includes: a keyword acquisition module 210, a function word extraction module 220, a POI acquisition module 230, a POI recommendation module 240, and a corresponding relationship establishment module 250. The keyword acquisition module 210, the function word extraction module 220, the POI acquisition module 230, the POI recommendation module 240 and the corresponding relationship establishment module 250 may be connected via a bus. The keyword acquisition module 210 is used to acquire keywords input in the search interface.

The function word extraction module 220 is used to extract the function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest POI, and one point of interest POI corresponds to n function words, n≥2.

The POI acquisition module 230 is used to acquire the POI corresponding to the function word to be retrieved as the candidate POI, and acquire the association between the function word to be retrieved and each candidate POI.

The POI recommendation module 240 is used to select the recommended candidate POIs according to the degree of association.

As an optional implementation, the function word extraction module 220 is used to extract a target keyword from the keywords, wherein the target keyword can represent the search intent of the keyword, and use the target keyword as the function word to be searched.

As an optional implementation, the device also includes: a correspondence establishment module 250, which is used to establish a correspondence between each point of interest POI and a function word according to the function word contained in the page content corresponding to each point of interest POI, or to establish a correspondence between each point of interest POI and a function word according to the retrieval log of each point of interest POI. The point of interest POI acquisition module 230 is used to: according to the correspondence between each point of interest POI and the function word, acquire the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI.

As an optional implementation, the corresponding relationship establishing module 250 is used to: extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI; use the category words and/or feature words of each point of interest POI as the function words of each point of interest POI, and establish a corresponding relationship between each point of interest POI and the function words.

As an optional implementation manner, the corresponding relationship establishing module 250 is further used for:

The name, category, feature and/or consumption information of the point of interest (POI) contained in the page content corresponding to each point of interest (POI), wherein the consumption information is used to reflect the consumption of the user at the point of interest (POI), the name, category, feature and/or consumption information of each point of interest (POI) are clustered to obtain the function words of each point of interest (POI), the function words of each point of interest (POI) are input into a convolutional neural network, the convolutional neural network is trained, and the correspondence between each point of interest (POI) and the function words is represented by the trained convolutional neural network, wherein the trained convolutional neural network can convert the input function words to be retrieved into candidate points of interest (POI).

As an optional implementation, the corresponding relationship establishing module 250 is further used to: obtain a target keyword contained in a search log of each point of interest POI; and use the target keyword as a function word of the point of interest POI corresponding to the search log.

As an optional implementation, the POI acquisition module 230 is also used to: convert each candidate point of interest POI into a first vector word according to all function words corresponding to each candidate point of interest POI, and convert the function words to be retrieved into a second vector word; calculate the similarity between the first vector word and the second vector word, and use the similarity as the correlation between the function words to be retrieved and each candidate point of interest POI.

As an optional implementation, the POI recommendation module 240 is used to: obtain the candidate POIs with a correlation greater than a set threshold as the target POI, recommend the target POI, obtain the first m candidate POIs with the largest correlation as the target POI, and recommend the target POI, where m≥1. Regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

Fig. 3 is a block diagram of an electronic device 800 for recommending points of interest (POI) according to an exemplary embodiment. For example, the electronic device 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.

3 , the electronic device 800 may include one or more of the following components: a processing component 802 , a memory 804 , a power component 806 , a multimedia component 808 , an audio component 810 , an input/output (I/O) interface 812 , a sensor component 814 , and a communication component 816 .

The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

The memory 804 is configured to store various types of data to support operations on the device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phone book data, messages, pictures, videos, etc. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

The power supply component 806 provides power to the various components of the electronic device 800. The power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the electronic device 800.

The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front camera and/or a rear camera. When the device 800 is in an operating mode, such as a shooting mode or a video mode, the front camera and/or the rear camera may receive external multimedia data. Each front camera and the rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

The audio component 810 is configured to output and/or input audio signals. For example, the audio component 810 includes a microphone (MIC), and when the electronic device 800 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 804 or sent via the communication component 816. In some embodiments, the audio component 810 also includes a speaker for outputting audio signals.

I/O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: home button, volume button, start button, and lock button.

The sensor assembly 814 includes one or more sensors for providing various aspects of status assessment for the electronic device 800. For example, the sensor assembly 814 can detect the open/closed state of the device 800, the relative positioning of components, such as the display and keypad of the electronic device 800, and the sensor assembly 814 can also detect the position change of the electronic device 800 or a component of the electronic device 800, the presence or absence of contact between the user and the electronic device 800, the orientation or acceleration/deceleration of the electronic device 800, and the temperature change of the electronic device 800. The sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 814 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above methods.

In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, and the instructions can be executed by a processor 820 of an electronic device 800 to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform an information recommendation method, the method comprising:

Obtain keywords input in a search interface; extract function words to be retrieved from the keywords, wherein the function words are used to characterize categories or features of POIs, and one POI corresponds to n function words, where n≥2; obtain the POIs corresponding to the function words to be retrieved as candidate POIs, and obtain the degree of association between the function words to be retrieved and each candidate POI; and select and recommend the candidate POIs according to the degree of association.

FIG4 is a schematic diagram of the structure of a server in an embodiment of the present invention. The server 1900 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 1922 (for example, one or more processors) and memory 1932, and one or more storage media 1930 (for example, one or more mass storage devices) storing application programs 1942 or data 1944. Among them, the memory 1932 and the storage medium 1930 may be short-term storage or permanent storage. The program stored in the storage medium 1930 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the server. Furthermore, the central processing unit 1922 may be configured to communicate with the storage medium 1930 and execute a series of instruction operations in the storage medium 1930 on the server 1900.

The server 1900 may also include one or more power supplies 1926, one or more wired or wireless network interfaces 1950, one or more input and output interfaces 1958, one or more keyboards 1956, and/or, one or more operating systems 1941, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not disclosed in this disclosure. The specification and examples are to be considered exemplary only, and the true scope and spirit of the present invention are indicated by the following claims.

It should be understood that the present invention is not limited to the precise structure described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (12)

1. An information recommendation method, characterized in that the method comprises: Get the keywords entered in the search interface; Extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2; Acquire the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and acquire the association degree between the function word to be retrieved and each candidate point of interest POI; Select and recommend the candidate point of interest POI according to the magnitude of the association; The method further comprises: Establishing a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI, or establishing a correspondence between each point of interest POI and the function word according to the search log of each point of interest POI; the search log of each point of interest POI includes the keyword input in the search history, the target keyword corresponding to the keyword, and the user selection result corresponding to the target keyword; The establishing of a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI includes: Extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI; use the category words and/or feature words of each point of interest POI as the function words of each point of interest POI, and establish a corresponding relationship between each point of interest POI and the function words; Alternatively, the name, category, feature and/or consumption information of the point of interest (POI) contained in the page content corresponding to each point of interest (POI) is obtained, wherein the consumption information is used to reflect the consumption of the user at the point of interest (POI); the name, category, feature and/or consumption information of each point of interest (POI) is clustered to obtain the function word of each point of interest (POI); the function word of each point of interest (POI) is input into a convolutional neural network, and the convolutional neural network is trained, and the corresponding relationship between each point of interest (POI) and the function word is represented by the trained convolutional neural network, wherein the trained convolutional neural network can convert the input function word to be retrieved into a candidate point of interest (POI); The obtaining of the correlation between the function word to be retrieved and each candidate point of interest POI includes: According to all function words corresponding to each candidate point of interest POI, each candidate point of interest POI is converted into a first word vector, and the function words to be retrieved are converted into a second word vector; Calculate the similarity between the first word vector and the second word vector, and use the similarity as the association between the function word to be retrieved and each candidate point of interest POI, including: calculating the angle between the first word vector and the second word vector, and then calculating the cosine value of the angle, using the cosine value to measure the similarity between the first word vector and the second word vector, the larger the cosine value, the higher the similarity between the first word vector and the second word vector, and the smaller the cosine value, the lower the similarity between the first word vector and the second word vector. 2. The information recommendation method according to claim 1, characterized in that the step of extracting the function words to be searched from the keywords comprises: Extracting a target keyword from the keywords, wherein the target keyword can represent the search intent of the keyword; The target keyword is used as the function word to be searched. 3. The information recommendation method according to claim 1, characterized in that the step of obtaining the POI corresponding to the function word to be retrieved as the candidate POI comprises: According to the correspondence between each point of interest POI and the function word, the point of interest POI corresponding to the function word to be retrieved is obtained as a candidate point of interest POI. 4. The information recommendation method according to claim 1, characterized in that the step of establishing a correspondence between each point of interest (POI) and a function word according to the search log of each point of interest (POI) comprises: Obtain target keywords contained in the search log of each point of interest (POI); The target keyword is used as a function word of the point of interest (POI) corresponding to the search log. 5. The information recommendation method according to claim 1, characterized in that the step of selecting and recommending the candidate POI according to the magnitude of the association degree comprises: Acquire the candidate points of interest POIs whose relevance is greater than a set threshold as target points of interest POIs, and recommend the target points of interest POIs; or The first m candidate points of interest (POIs) with the greatest correlation are obtained as target points of interest (POIs), and the target points of interest (POIs) are recommended, where m≥1. 6. An information recommendation device, characterized in that the device comprises: The keyword acquisition module is used to obtain the keywords entered in the search interface; A function word extraction module is used to extract the function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest POI, and one point of interest POI corresponds to n function words, where n≥2; A POI acquisition module is used to acquire the POI corresponding to the function word to be retrieved as a candidate POI, and acquire the correlation between the function word to be retrieved and each candidate POI; A POI recommendation module is used to select and recommend the candidate POI according to the degree of association; The device also includes: A corresponding relationship establishing module is used to establish a corresponding relationship between each point of interest POI and a function word according to the function word contained in the page content corresponding to each point of interest POI, or to establish a corresponding relationship between each point of interest POI and a function word according to a search log of each point of interest POI; the search log of each point of interest POI includes a keyword input in the search history, a target keyword corresponding to the keyword, and a user selection result corresponding to the target keyword; The corresponding relationship establishment module is used for: Extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI; use the category words and/or feature words of each point of interest POI as the function words of each point of interest POI, and establish a corresponding relationship between each point of interest POI and the function words; Alternatively, the name, category, feature and/or consumption information of the point of interest (POI) contained in the page content corresponding to each point of interest (POI) is obtained, wherein the consumption information is used to reflect the consumption of the user at the point of interest (POI); the name, category, feature and/or consumption information of each point of interest (POI) is clustered to obtain the function word of each point of interest (POI); the function word of each point of interest (POI) is input into a convolutional neural network, and the convolutional neural network is trained, and the corresponding relationship between each point of interest (POI) and the function word is represented by the trained convolutional neural network, wherein the trained convolutional neural network can convert the input function word to be retrieved into a candidate point of interest (POI); The module for obtaining points of interest (POI) is also used for: According to all function words corresponding to each candidate point of interest POI, each candidate point of interest POI is converted into a first vector word, and the function words to be retrieved are converted into a second vector word; Calculate the similarity between the first vector word and the second vector word, and use the similarity as the association between the function word to be retrieved and each candidate point of interest POI, including: calculating the angle between the first word vector and the second word vector, and then calculating the cosine value of the angle, using the cosine value to measure the similarity between the first word vector and the second word vector, the larger the cosine value, the higher the similarity between the first word vector and the second word vector, and the smaller the cosine value, the lower the similarity between the first word vector and the second word vector. 7. The information recommendation device according to claim 6, characterized in that the function word extraction module is used to: Extracting a target keyword from the keywords, wherein the target keyword can represent the search intent of the keyword; The target keyword is used as the function word to be searched. 8. The information recommendation device according to claim 6, wherein the module for obtaining points of interest (POI) is used for: According to the correspondence between each point of interest POI and the function word, the point of interest POI corresponding to the function word to be retrieved is obtained as a candidate point of interest POI. 9. The information recommendation device according to claim 6, characterized in that the corresponding relationship establishment module is further used for: Obtain target keywords contained in the search log of each point of interest (POI); The target keyword is used as a function word of the point of interest (POI) corresponding to the search log. 10. The information recommendation device according to claim 6, wherein the POI recommendation module is used to: Acquire the candidate points of interest (POI) whose relevance is greater than a set threshold as target points of interest (POI), and recommend the target points of interest (POI); The first m candidate points of interest (POIs) with the greatest correlation are obtained as target points of interest (POIs), and the target points of interest (POIs) are recommended, where m≥1. 11. A computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the following steps are performed: Get the keywords entered in the search interface; Extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2; Acquire the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and acquire the association degree between the function word to be retrieved and each candidate point of interest POI; Select and recommend the candidate point of interest POI according to the magnitude of the association; The following steps are also implemented: Establishing a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI, or establishing a correspondence between each point of interest POI and the function word according to the search log of each point of interest POI; the search log of each point of interest POI includes the keyword input in the search history, the target keyword corresponding to the keyword, and the user selection result corresponding to the target keyword; The establishing of a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI includes: Extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI; use the category words and/or feature words of each point of interest POI as the function words of each point of interest POI, and establish a corresponding relationship between each point of interest POI and the function words; Alternatively, the name, category, feature and/or consumption information of the point of interest (POI) contained in the page content corresponding to each point of interest (POI) is obtained, wherein the consumption information is used to reflect the consumption of the user at the point of interest (POI); the name, category, feature and/or consumption information of each point of interest (POI) is clustered to obtain the function word of each point of interest (POI); the function word of each point of interest (POI) is input into a convolutional neural network, and the convolutional neural network is trained, and the corresponding relationship between each point of interest (POI) and the function word is represented by the trained convolutional neural network, wherein the trained convolutional neural network can convert the input function word to be retrieved into a candidate point of interest (POI); The obtaining of the correlation between the function word to be retrieved and each candidate point of interest POI includes: According to all function words corresponding to each candidate point of interest POI, each candidate point of interest POI is converted into a first word vector, and the function words to be retrieved are converted into a second word vector; Calculate the similarity between the first word vector and the second word vector, and use the similarity as the association between the function word to be retrieved and each candidate point of interest POI, including: calculating the angle between the first word vector and the second word vector, and then calculating the cosine value of the angle, using the cosine value to measure the similarity between the first word vector and the second word vector, the larger the cosine value, the higher the similarity between the first word vector and the second word vector, and the smaller the cosine value, the lower the similarity between the first word vector and the second word vector. 12. An electronic device, characterized in that it comprises a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, and the one or more programs include instructions for performing the following operations: Get the keywords entered in the search interface; Extracting function words to be searched from the keywords, wherein the function words are used to characterize the category or feature of the point of interest (POI), and one point of interest (POI) corresponds to n function words, where n≥2; Acquire the point of interest POI corresponding to the function word to be retrieved as a candidate point of interest POI, and acquire the association degree between the function word to be retrieved and each candidate point of interest POI; Select and recommend the candidate point of interest POI according to the magnitude of the association; Also contains instructions for: Establishing a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI, or establishing a correspondence between each point of interest POI and the function word according to the search log of each point of interest POI; the search log of each point of interest POI includes the keyword input in the search history, the target keyword corresponding to the keyword, and the user selection result corresponding to the target keyword; The establishing of a correspondence between each point of interest POI and the function word according to the function word contained in the page content corresponding to each point of interest POI includes: Extract the category words and/or feature words of each point of interest POI from the registration information, classification information, and comment information in the page content corresponding to each point of interest POI; use the category words and/or feature words of each point of interest POI as the function words of each point of interest POI, and establish a corresponding relationship between each point of interest POI and the function words; Alternatively, the name, category, feature and/or consumption information of the point of interest (POI) contained in the page content corresponding to each point of interest (POI) is obtained, wherein the consumption information is used to reflect the consumption of the user at the point of interest (POI); the name, category, feature and/or consumption information of each point of interest (POI) is clustered to obtain the function word of each point of interest (POI); the function word of each point of interest (POI) is input into a convolutional neural network, and the convolutional neural network is trained, and the corresponding relationship between each point of interest (POI) and the function word is represented by the trained convolutional neural network, wherein the trained convolutional neural network can convert the input function word to be retrieved to output a candidate point of interest (POI); The obtaining of the correlation between the function word to be retrieved and each candidate point of interest POI includes: According to all function words corresponding to each candidate point of interest POI, each candidate point of interest POI is converted into a first word vector, and the function words to be retrieved are converted into a second word vector; Calculate the similarity between the first word vector and the second word vector, and use the similarity as the association between the function word to be retrieved and each candidate point of interest POI, including: calculating the angle between the first word vector and the second word vector, and then calculating the cosine value of the angle, using the cosine value to measure the similarity between the first word vector and the second word vector, the larger the cosine value, the higher the similarity between the first word vector and the second word vector, and the smaller the cosine value, the lower the similarity between the first word vector and the second word vector.