CN114329212A

CN114329212A - Information recommendation method and device and electronic equipment

Info

Publication number: CN114329212A
Application number: CN202111652770.0A
Authority: CN
Inventors: 李书伟; 姚珺珺; 刘晓庆
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2021-12-30
Filing date: 2021-12-30
Publication date: 2022-04-12

Abstract

The disclosure provides an information recommendation method and device and electronic equipment, relates to the technical field of data processing, and particularly relates to the technical field of search recommendation. The specific implementation scheme is as follows: acquiring N pieces of historical search information matched with target search information input by a user from log data; determining a similar meaning word list of words in the N pieces of historical search information; under the condition that the N pieces of historical search information comprise first historical search information, deleting the first historical search information in the N pieces of historical search information to obtain first candidate recommendation information, wherein the first historical search information comprises a first word, an intersection exists between a near meaning word list of the first word and a near meaning word list of a second word, and the second word comprises at least one of the following items: the word of the first historical search information, the word of the second historical search information in the N pieces of historical search information, which corresponds to the first word, and the word of the target search information, which corresponds to the first word; and recommending information based on the first candidate recommendation information.

Description

Information recommendation method and device and electronic equipment

Technical Field

The present disclosure relates to the field of data processing technologies, and in particular, to a search recommendation technology, and more particularly, to an information recommendation method and apparatus, and an electronic device.

Background

With the development of scientific technology and internet technology, the e-commerce platform rises rapidly, and the e-commerce platform can provide various network services for users, so that great convenience is brought to production and life.

On an e-commerce platform, when a user enters a website to search for a commodity, search information is generally required to be input on a search bar, and correspondingly, the platform can recommend information according to the input condition of the search information of the user.

Currently, information recommendation is usually performed by extracting historical search information related to search information input by a user from log data.

Disclosure of Invention

The disclosure provides an information recommendation method and device and electronic equipment.

According to a first aspect of the present disclosure, there is provided an information recommendation method including:

acquiring N pieces of historical search information matched with target search information input by a user from log data, wherein N is a positive integer;

determining a list of similar words of the words in the N pieces of historical search information;

deleting first historical search information in the N pieces of historical search information to obtain first candidate recommendation information under the condition that the N pieces of historical search information include the first historical search information, wherein the first historical search information includes a first word, a synonym list of the first word and a synonym list of a second word have intersection, and the second word includes at least one of the following items: a word of the first historical search information, a word of second historical search information of the N pieces of historical search information corresponding to the first word, and a word of the target search information corresponding to the first word;

and recommending information based on the first candidate recommending information.

According to a second aspect of the present disclosure, there is provided an information recommendation apparatus including:

the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring N pieces of historical search information matched with target search information input by a user from log data, and N is a positive integer;

the first determining module is used for determining a near meaning word list of words in the N pieces of historical search information;

a deleting module, configured to delete, when the N pieces of historical search information include first historical search information, the first historical search information in the N pieces of historical search information to obtain first candidate recommendation information, where the first historical search information includes a first word, where a near word list of the first word and a near word list of a second word have an intersection, and the second word includes at least one of the following: a word of the first historical search information, a word of second historical search information of the N pieces of historical search information corresponding to the first word, and a word of the target search information corresponding to the first word;

and the recommending module is used for recommending information based on the first candidate recommending information.

According to a third aspect of the present disclosure, there is provided an electronic device comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform any one of the methods of the first aspect.

According to a fourth aspect of the present disclosure, there is provided a non-transitory computer readable storage medium having stored thereon computer instructions for causing a computer to perform any one of the methods of the first aspect.

According to a fifth aspect of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, implements any of the methods of the first aspect.

According to the technology disclosed by the invention, the problem of poor information recommendation effect is solved, and the information recommendation effect is improved.

It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.

Drawings

The drawings are included to provide a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

fig. 1 is a flowchart illustrating an information recommendation method according to a first embodiment of the present disclosure;

fig. 2 is a schematic configuration diagram of an information recommendation device according to a second embodiment of the present disclosure;

FIG. 3 is a schematic block diagram of an example electronic device used to implement embodiments of the present disclosure.

Detailed Description

Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, in which various details of the embodiments of the disclosure are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

First embodiment

As shown in fig. 1, the present disclosure provides an information recommendation method, including the steps of:

step S101: n pieces of historical search information matched with target search information input by a user are obtained from the log data.

Wherein N is a positive integer.

In the embodiment, the information recommendation method relates to the technical field of data processing, in particular to the technical field of search recommendation, and can be widely applied to a commodity purchasing scene. The information recommendation method of the embodiment of the disclosure may be executed by the information recommendation apparatus of the embodiment of the disclosure. The information recommendation device of the embodiment of the disclosure can be configured in any electronic equipment to execute the information recommendation method of the embodiment of the disclosure. The electronic device may be a server or a terminal device, and is not limited specifically here.

The present embodiment may be applied to a toB procurement scenario and may also be applied to an toC procurement scenario, which are not specifically limited herein. The toB procurement scenario and toC procurement scenario differ among them in that the e-commerce platform is oriented to different users, for the toB procurement scenario it is generally oriented to a group, such as a community, group or organization, and for the toC procurement scenario it is generally oriented to an individual.

The target search information may be information entered by the user in a search input box, such as a "notebook" of target search information entered by the user on the e-commerce platform.

When the information recommendation device detects that the user is performing input in the search input box, the target search information input in the search input box by the user can be acquired in real time or periodically, and information recommendation is performed based on the target search information. The information recommendation device can predict the search information which is possibly needed by the user based on the target search information, and can display the predicted search information in a pull-down menu mode in the search input box to perform information recommendation, so that more accurate demand expression can be provided for the user, the user is helped to quickly locate the content which the user wants to search, and the typing time of the user is saved.

In the case where the target search information input by the user is acquired, the information recommendation device may match the target search information with the historical search information in the log data to acquire N pieces of historical search information that match the target search information from the log data, where N is a positive integer.

Before matching, the historical search information in the log data can be filtered, or after matching, a plurality of historical search information matched with the target search information can be filtered, so that the search information is subjected to quality control before information recommendation, and N pieces of historical search information are obtained.

The quality control may include: 1) filtering historical search information by utilizing a pre-established search information blacklist, namely filtering the historical search information in the search information blacklist and filtering the historical search information containing words in the search information blacklist; 2) filtering out historical search information with too short word number and too long word number; 3) and filtering out historical search information with too few retrieval results or low correlation with commodity objects.

The matching between the historical search information and the target search information may refer to semantic matching between the historical search information and the target search information, may also refer to matching between a category corresponding to the historical search information and a category corresponding to the target search information, and may also refer to matching between a core word in the historical search information and a core word in the target search information, which is not specifically limited herein.

Accordingly, the ways of matching the target search information with the historical search information in the log data include, but are not limited to, semantic matching, category matching, core word matching, and the like.

Step S102: determining a list of synonyms for a term in the N historical search information.

In this step, each piece of historical search information in the N pieces of historical search information includes at least one word, and the historical search information may include a noun, an adjective, a verb, or the like, for example, if the historical search information is a "small notebook," the historical search information may include an adjective and a noun, and for example, if the historical search information is a "folding airplane," the historical search information may include a verb and a noun.

A list of similar words of the words in the N pieces of historical search information may be determined, and in an alternative embodiment, the historical search information may be subjected to word segmentation and query processing based on a pre-trained model for each piece of historical search information, so as to obtain a target result of the historical search information, where the target result may include at least one word of the historical search information and the list of similar words of each word.

The synonym list of terms may include terms that have a synonymous relationship with the term, e.g., the synonym list of the term "custom" may include terms such as "custom", "production", etc.

Step S103: deleting first historical search information in the N pieces of historical search information to obtain first candidate recommendation information under the condition that the N pieces of historical search information include the first historical search information, wherein the first historical search information includes a first word, a synonym list of the first word and a synonym list of a second word have intersection, and the second word includes at least one of the following items: a word of the first historical search information, a word of a second historical search information of the N historical search information corresponding to the first word, and a word of the target search information corresponding to the first word.

The first history search information may be history search information in which semantic duplication exists inside the information, for example, the history search information "notebook small", and the two words "notebook" and "notebook" in the information are duplicated.

The first historical search information may also be historical search information that has semantic duplication with other historical search information in the N pieces of historical search information, for example, semantic duplication between the first historical search information "notebook customization" and other historical search information "notebook customization" exists.

The first historical search information may also be historical search information that is semantically repeated with the target search information presence information.

Whether the first history search information with repeated semantics exists in the N history search information can be determined by carrying out cross validation on the similar meaning word lists of the two words. When determining history search information with repeated semantics in the information, cross-validation may be performed on the near-synonym lists of the two terms in the history search information, and if the near-synonym lists of the two terms have intersection, the repeated semantics in the history search information is the first history search information. At this time, the first word and the second word are both two words in one piece of history search information.

For example, the history search information "notebook small", the list of synonyms of the first word "notebook" and the list of synonyms of the second word "notebook" have an intersection, which is the first history search information.

When history search information with repeated semantics exists between information is determined, the similar meaning word lists of two words corresponding to each other in the two history search information can be subjected to cross validation, if the similar meaning word lists of every two words corresponding to each other have intersection, the two history search information have repeated semantics, and one of the two history search information is the first history search information. At this time, the first word and the second word are two words corresponding to each other in the two pieces of history search information, respectively.

For example, if one piece of history search information is "notebook custom" and the other piece of history search information is "notebook custom", there is an intersection between the list of near-synonyms of the first word "custom" and the list of near-synonyms of the second word "custom" corresponding to the first word, and there is semantic duplication between the two pieces of history search information.

In addition, whether semantic duplication exists between the target search information and the historical search information or not can be determined in the same way as the historical search information with semantic duplication between the information, and the repeated description is omitted here. At this time, the first word may be a word in the history search information, and the second word may be a word corresponding to the first word in the target search information.

Wherein, two word correspondences in different search information may include, but are not limited to, a location correspondence, a part-of-speech correspondence, and the like.

When the N pieces of history search information include the first history search information, the first history search information in the N pieces of history search information may be deleted to obtain first candidate recommendation information, and the first candidate recommendation information may include history search information in the N pieces of history search information from which the first history search information is deleted. The historical search information in the first candidate recommendation information can be used as the search information to be recommended to perform information recommendation.

Step S104: and recommending information based on the first candidate recommending information.

In this step, the first candidate recommendation information may include one, two, or more pieces of search information to be recommended, and the recommendation weight of each piece of search information to be recommended may be determined.

In the embodiment, when the to-be-recommended search information is determined based on the historical search information, whether the historical search information with repeated semantics exists can be determined by performing cross validation on the similar meaning word lists of the two words, and the historical search information with repeated semantics is deleted, so that the to-be-recommended search information can be optimized, and the information recommendation effect can be improved.

Optionally, before the step S103, the method further includes at least one of:

for each historical search information in the N pieces of historical search information, determining the historical search information as the first historical search information under the condition that the synonym lists of any two terms in the historical search information have an intersection;

determining that first target historical search information is the first historical search information when the first target historical search information exists in the N pieces of historical search information, wherein for each word in the first target historical search information, a word corresponding to the word exists in the target search information, and an intersection exists between the similar meaning word lists of every two corresponding words in the first target historical search information and the target search information;

and under the condition that N is greater than 1, determining that the historical search information of one of the two pieces of historical search information is the first historical search information if the synonym lists of every two corresponding words in the two pieces of historical search information have intersection, wherein the two corresponding words in the two pieces of historical search information are respectively from the two pieces of historical search information.

In this embodiment, it may be detected that semantic duplication exists inside the information for each piece of history search information in the N pieces of history search information, and when the history search information includes at least two words, cross-validation may be performed on the synonym lists of every two words in the history search information, and if there is an intersection between the synonym lists of any two words, the semantic duplication inside the history search information exists, which is the first history search information.

The semantic duplication between the target search information and each of the N pieces of historical search information may be detected, and in the case of semantic duplication between the historical search information and the target search information, the number of words in the target search information is usually greater than or equal to the number of words in the historical search information.

Therefore, whether a word corresponding to the word exists in the target search information can be determined for each word in the historical search information, if so, whether an intersection exists in the similar word lists of every two words corresponding to each other is determined, and if so, semantic duplication between the historical search information and the target search information can be determined.

When N is greater than 1, semantic duplication between information may be detected for every two pieces of history search information in the N pieces of history search information, cross-validation may be performed on the synonym lists of two words corresponding to each other in the two pieces of history search information, if an intersection exists between the synonym lists of every two words corresponding to each other, semantic duplication between the two pieces of history search information exists, and one of the two pieces of history search information is the first history search information.

When semantic duplication is detected between pieces of information, if duplication is removed only by using the inclusion relationship of the similar meaning word list of any two words corresponding to each other, a plurality of pieces of search information may be considered to be synonymous relationships including the core word. For example, the target search information "notebook", one historical search information "notebook customization", and the other historical search information "notebook customization" may all be considered to be synonymous relationships since they all include the core word "notebook".

Therefore, in practical application, the words in the historical search information, which have inclusion relation with the similar meaning word list of the core word of the target search information, can be removed, the remaining words are used for cross validation of the similar meaning word list, and if the similar meaning word lists of every two corresponding words have intersection, semantic duplication between the two search information is determined. This can reduce the amount of calculation.

For example, the target search information is "notebook", one piece of history search information is "notebook custom" and the other piece of history search information is "notebook custom", the word "notebook" in the two pieces of history search information may be removed, the remaining two word "custom" and "custom" synonym lists corresponding to each other are cross-verified, and if there is an intersection between the two word synonym lists, it is determined that the semantics of the two pieces of history search information are repeated.

In this embodiment, the first history search information in which semantic duplication exists inside information and semantic duplication exists between information among N pieces of history search information can be detected by performing cross validation on the similar word lists of two words.

Optionally, before the step S103, the method further includes:

and under the condition that second target historical search information exists in the N pieces of historical search information, determining the second target historical search information as the first historical search information, wherein the target search information comprises the second target historical search information.

In this embodiment, it may be determined whether or not the target search information and the history search information are in an information inclusion relationship for each of the N pieces of history search information, and in a case where the target search information includes the history search information, it may be determined that the history search information is the first history search information.

The target search information including the history search information may mean that the same content as the history search information exists in the target search information. For example, if the target search information is "hard-shell notebook", and one of the history search information is "notebook", the target search information includes the history search information.

The target search information including the historical search information may also mean that the target search information has content with the same semantic meaning as the historical search information. For example, if the target search information is "large desktop computer" and a piece of history search information is "computer", the target search information includes the history search information.

In this embodiment, by detecting the information inclusion relationship between the target search information and the history search information, it is possible to detect the first history search information having semantic overlap with the target search information among the N pieces of history search information.

Optionally, the target search information includes M words, and before the step S104, the method further includes:

determining a first core word of the M words;

obtaining a descriptor associated with the first core word in a database, wherein the database is stored with the core word and the descriptor in an associated manner;

splicing the first core word and the descriptor associated with the first core word to obtain second candidate recommendation information;

the recommending information based on the first candidate recommending information comprises:

and recommending information based on the first candidate recommending information and the second candidate recommending information.

In this embodiment, the target search information may include M words, where M is a positive integer. The M words may include nouns, adjectives, verbs, etc., for example, if the target search information is "small notebook", the target search information may include the adjectives and the nouns, and if the target search information is "folding airplane", the target search information may include the verbs and the nouns.

In an alternative embodiment, if the user inputs the target search information while being separated by spaces between different words, the information recommendation apparatus may divide the words by detecting the spaces to obtain M words.

In another alternative embodiment, the information recommendation device may perform word segmentation on the target search information through a pre-trained word segmentation model, such as a jieba word segmentation tool, to obtain M words and a word segmentation weight of each word. Before word segmentation, useless characters of target search information can be filtered out by using words in a prestored stop word list, and then word segmentation is carried out, so that the word segmentation accuracy can be improved.

The first core word in the M words may refer to a keyword in the M words, where the keyword may be a noun and refers to a word in the target search information that best expresses the search requirement of the user, for example, if the target search information is a "hard-shell notebook," the keyword of the target search information is a notebook.

The determination method of the first core word may also include multiple manners, for example, it may be determined whether a word in the M words matches a word in the keyword library, and a word in the M words that matches the keyword library is determined as the first core word.

For another example, category analysis may be performed on each of the M words to obtain category information corresponding to the word; determining at least one candidate word from the M words, wherein the category information corresponding to each candidate word is intersected with the category information obtained by category analysis of the target search information; a first core word is determined from the at least one candidate word.

In this embodiment, a database stores core words and descriptors in an associated manner, one core word may store one, two, or more descriptors in an associated manner, a descriptor associated with a first core word in the database may be obtained, and the first core word is spliced with descriptors associated with the first core word, so as to obtain second candidate recommendation information.

For example, the first core word is a notebook, the descriptors associated with the first core word in the database include a hard shell, a small size, an ultra-thick type and the like, and the first core word is spliced with each descriptor respectively to obtain second candidate recommendation information, which includes search information "hard shell notebook", "small notebook" and "ultra-thick notebook" obtained by splicing.

Before information recommendation, a core word in the history search information in the log data may be predetermined, words except the core word in the history search information may be determined as descriptors, and the core word and the descriptors are stored in the database in an associated manner, which will be described in detail below.

After the second candidate recommendation information is obtained, information recommendation can be performed based on the first candidate recommendation information and the second candidate recommendation information. In an optional implementation manner, the first candidate recommendation information and the second candidate recommendation information may be summarized, after the summary, since there may be an intersection between the first candidate recommendation information and the second candidate recommendation information, that is, there may be repeated search information to be recommended, in this case, deduplication may be performed, and one of the repeated search information to be recommended may be stored.

In this embodiment, the recommendation weight of each piece of search information to be recommended may be determined, the pieces of search information to be recommended are ranked according to the recommendation weight from large to small, and the pieces of search information to be recommended with the highest ranking recommendation weight are recommended to the user.

If the search information to be recommended is the search information to be recommended in the first candidate recommendation information, the recommendation weight of the search information to be recommended may be determined based on the matching degree of the search information to be recommended and the target search information.

If the to-be-recommended search information is to-be-recommended search information in the second candidate recommendation information, the weight corresponding to the descriptor in the to-be-recommended search information may be determined based on the weight corresponding to the descriptor in the to-be-recommended search information, for example, the weight corresponding to the descriptor in the to-be-recommended search information may be determined as the recommendation weight of the to-be-recommended search information. Each descriptor associated with a core word in the database can correspond to a weight, and the greater the weight is, the more closely the association degree of the descriptor and the core word is.

If the to-be-recommended search information is to-be-recommended search information in an intersection of the first candidate recommendation information and the second candidate recommendation information, in this case, the weight determination corresponding to the to-be-recommended search information and the weight comprehensive determination corresponding to the descriptor in the to-be-recommended search information may be based on, for example, the weight determination corresponding to the to-be-recommended search information and the average value of the weights corresponding to the descriptor in the to-be-recommended search information may be determined as the recommendation weight of the to-be-recommended search information.

In another optional embodiment, a first recommendation weight of historical search information in the first candidate recommendation information may be determined based on a preset first channel weight; determining a second recommendation weight of information obtained by splicing the first core word and the descriptor associated with the first core word in the second candidate recommendation information based on a preset second channel weight; determining target recommendation information from the first candidate recommendation information and the second candidate recommendation information based on the first recommendation weight and the second recommendation weight; and recommending information based on the target recommendation information. Therefore, the two channels of the log data and the core words can be combined to respectively recommend search guidance, and the information recommendation effect can be further improved.

In this embodiment, the first channel weight may be a channel for information recommendation based on log data, the second channel weight may be a channel for information recommendation based on core words, and the first channel weight and the second channel weight may be set in advance, for example, the first channel weight may be set to 0.7 and the second channel weight may be set to 0.3.

Accordingly, the first recommendation weight may be determined based on the first channel weight and a weight corresponding to the historical search information in the first candidate recommendation information, that is, the search information to be recommended, in a comprehensive manner, and the second recommendation weight may be determined based on the second channel weight and a weight corresponding to the descriptor in the search information to be recommended in the second candidate recommendation information in a comprehensive manner. For example, the two weights may be multiplied to obtain a recommendation weight of the search information to be recommended.

In a possible implementation manner, the search information to be recommended in the first candidate recommendation information may be ranked from large to small according to the first recommendation weight, and the search information to be recommended ranked before the first recommendation weight is determined as the target recommendation information. The search information to be recommended in the second candidate recommendation information may also be ranked from large to small according to the second recommendation weight, and the search information to be recommended with the second recommendation weight ranked in front may be determined as the target recommendation information.

When information recommendation is performed based on the determined target recommendation information, if repeated search information exists in the determined target recommendation information, deduplication processing may be performed, and then the target recommendation information is recommended to the user.

In the embodiment, the information recommendation is performed by combining the first candidate recommendation information and the second candidate recommendation information, so that the recommendation information can be effectively expanded and enriched, and the information recommendation effect is further improved.

Optionally, M is greater than 1, and the determining a first core word in the M words includes:

analyzing the category of each word in the M words to obtain category information corresponding to the word;

determining at least one candidate word from the M words, wherein the category information corresponding to each candidate word is intersected with the category information obtained by category analysis of the target search information;

determining the first core word from the at least one candidate word.

In this embodiment, a pre-trained category analysis model may be used to perform category analysis on the target search information, so as to obtain category information corresponding to the target search information. And aiming at each word in the M words, the category analysis model can also be adopted to carry out category analysis on the word so as to obtain category information corresponding to the word.

The category analysis model may use a third-level category as an analysis target, that is, a category in the category information obtained through analysis is the third-level category, and the category information obtained through analysis may include at least one category and a weight corresponding to each category.

The category information corresponding to the word may include at least one category and a weight corresponding to each category, and the at least one candidate word may be determined from the M words, where the determination may be that, for each word in the M words, it may be determined whether the category information corresponding to the word intersects with the category information corresponding to the target search information, and if the intersection exists, the word may be determined as the candidate word.

The first core word may then be determined from at least one candidate word. Specifically, the target weight of each candidate word may be determined, and in an optional implementation, the participle weight of the candidate word may be multiplied by the position weight to obtain the target weight of the candidate word, where the position weight of the candidate word may be weighted according to the position of the candidate word in the target search information, and the position weight may be larger the farther the position is, the larger the position is. Accordingly, the candidate word with the largest target weight among the at least one candidate word may be determined as the first core word.

In the embodiment, the first core word in the target search information is determined by combining the category mode, so that the determination accuracy of the core word can be improved, and the information recommendation effect can be further improved.

It should be noted that, for the core words in the database, the determination manner of the first core word may also be used for determining, specifically, log data may be obtained, where the log data may include historical search information, historical search information in the log data may be filtered by using a pre-established search information blacklist, and historical search information in the log data may be filtered by using a pre-established category blacklist, for example, historical search information corresponding to category information in the log data where an intersection exists with the category blacklist may be filtered. Meanwhile, useless characters of historical search information can be filtered out by utilizing words in a prestored stop word list.

For the filtered log data, word segmentation processing can be performed on historical search information in the log data by using a word segmentation model, so that word segmentation results and word segmentation weights are obtained. If the word segmentation result is only one, determining the word segmentation result as a core word, and if the word segmentation result is multiple, performing category cross check on category information of each word segmentation result and category information of historical search information to obtain candidate words. If the candidate word is only one, the candidate word is a core word, if the candidate word is multiple, the candidate word is weighted according to the position sequence appearing in the historical search information, the target weight of the candidate word is the position proportion of the participle weight, the candidate word with the maximum target weight is taken as the core word, and the rest participle results of the historical search information are taken as the descriptor words.

Then, if the core words of different historical search information in the log data are the same, the descriptors associated with the core words can be clustered, and finally, a core word associated with a plurality of descriptors can be obtained, and the core word and the descriptors are stored in the database in an associated manner. Furthermore, the weight of the descriptor can be determined according to the participle weight and the position proportion of the descriptor, and the weight can be correspondingly stored in a database.

In addition, the information recommendation device may also perform information recommendation in combination with third candidate recommendation information, where the third candidate recommendation information may include historical search information searched by the user, and embody personalized preferences of the user. The personalized preference of the user can be mined based on the current-day historical search data and the past-day historical search data of the user, specifically, the historical search data searched by the user on the current day can most express the current possible search preference of the user, so if the target search information is contained in the current-day historical search data, the latest historical search information is taken as third candidate recommendation information according to the time sequence, and the historical search information is preferentially recommended to the user. And if the target search information is contained in the historical search data of the past day, sorting according to the historical click times of the target search information, and supplementing the target search information into the third candidate recommendation information.

Second embodiment

As shown in fig. 2, the present disclosure provides an information recommendation device 200 including:

a first obtaining module 201, configured to obtain N pieces of historical search information that match target search information input by a user from log data, where N is a positive integer;

a first determining module 202, configured to determine a list of synonyms of terms in the N pieces of historical search information;

a deleting module 203, configured to delete, when the N pieces of historical search information include first historical search information, the first historical search information in the N pieces of historical search information to obtain first candidate recommendation information, where the first historical search information includes a first word, where a near word list of the first word and a near word list of a second word have an intersection, and the second word includes at least one of the following: a word of the first historical search information, a word of second historical search information of the N pieces of historical search information corresponding to the first word, and a word of the target search information corresponding to the first word;

and the recommending module 204 is configured to recommend information based on the first candidate recommending information.

Optionally, the apparatus further comprises:

a second determining module, configured to determine, for each piece of historical search information in the N pieces of historical search information, that the piece of historical search information is the first piece of historical search information when an intersection exists between the synonym lists of any two terms in the piece of historical search information;

a third determining module, configured to determine, when first target historical search information exists in the N pieces of historical search information, that the first target historical search information is the first historical search information, for each word in the first target historical search information, a word corresponding to the word exists in the target search information, and in the first target historical search information and the target search information, an intersection exists between near-sense word lists of every two corresponding words;

a fourth determining module, configured to, for every two pieces of historical search information in the N pieces of historical search information, determine that the historical search information of one of the two pieces of historical search information is the first historical search information if an intersection exists between synonym lists of every two corresponding words in the two pieces of historical search information, where two corresponding words in the two pieces of historical search information are from the two pieces of historical search information, respectively, when N is greater than 1.

Optionally, the apparatus further comprises:

a fifth determining module, configured to determine, when second target historical search information exists in the N pieces of historical search information, that the second target historical search information is the first historical search information, where the target search information includes the second target historical search information.

Optionally, the target search information includes M words, and the apparatus further includes:

a sixth determining module, configured to determine a first core word in the M words;

the second acquisition module is used for acquiring the descriptors associated with the first core word in a database, and the core word and the descriptors are stored in the database in an associated manner;

the splicing module is used for splicing the first core word and the descriptor associated with the first core word to obtain second candidate recommendation information;

the recommending module 204 is specifically configured to recommend information based on the first candidate recommending information and the second candidate recommending information.

Optionally, M is greater than 1, and the sixth determining module is specifically configured to:

determining the first core word from the at least one candidate word.

Optionally, the recommending module 204 is specifically configured to:

determining a first recommendation weight of historical search information in the first candidate recommendation information based on a preset first channel weight;

determining a second recommendation weight of information obtained by splicing the first core word and the descriptor associated with the first core word in the second candidate recommendation information based on a preset second channel weight;

determining target recommendation information from the first candidate recommendation information and the second candidate recommendation information based on the first recommendation weight and the second recommendation weight;

and recommending information based on the target recommendation information.

The information recommendation device 200 provided by the present disclosure can implement each process implemented by the information recommendation method embodiment, and can achieve the same beneficial effects, and for avoiding repetition, the details are not repeated here.

In the technical scheme of the disclosure, the collection, storage, use, processing, transmission, provision, disclosure and other processing of the personal information of the related user are all in accordance with the regulations of related laws and regulations and do not violate the good customs of the public order.

The present disclosure also provides an electronic device, a readable storage medium, and a computer program product according to embodiments of the present disclosure.

FIG. 3 illustrates a schematic block diagram of an example electronic device that can be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 3, the apparatus 300 includes a computing unit 301 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM)302 or a computer program loaded from a storage unit 308 into a Random Access Memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the device 300 can also be stored. The calculation unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input/output (I/O) interface 305 is also connected to bus 304.

Various components in device 300 are connected to I/O interface 305, including: an input unit 306 such as a keyboard, a mouse, or the like; an output unit 307 such as various types of displays, speakers, and the like; a storage unit 308 such as a magnetic disk, optical disk, or the like; and a communication unit 309 such as a network card, modem, wireless communication transceiver, etc. The communication unit 309 allows the device 300 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.

The computing unit 301 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various dedicated Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, and so forth. The calculation unit 301 executes the respective methods and processes described above, such as the information recommendation method. For example, in some embodiments, the information recommendation method may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as storage unit 308. In some embodiments, part or all of the computer program may be loaded and/or installed onto device 300 via ROM 302 and/or communication unit 309. When the computer program is loaded into RAM 303 and executed by the computing unit 301, one or more steps of the information recommendation method described above may be performed. Alternatively, in other embodiments, the computing unit 301 may be configured to perform the information recommendation method in any other suitable way (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), system on a chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server with a combined blockchain.

It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present disclosure may be executed in parallel or sequentially or in different orders, and are not limited herein as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved.

The above detailed description should not be construed as limiting the scope of the disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. An information recommendation method, comprising:

2. The method according to claim 1, wherein in a case that first historical search information is included in the N pieces of historical search information, before deleting the first historical search information from the N pieces of historical search information and obtaining first candidate recommendation information, the method further includes at least one of:

3. The method according to claim 1, wherein in a case that first historical search information is included in the N pieces of historical search information, before deleting the first historical search information from the N pieces of historical search information and obtaining first candidate recommendation information, the method further includes:

4. The method of claim 1, wherein the target search information comprises M words, and before the recommending information based on the first candidate recommendation information, the method further comprises:

determining a first core word of the M words;

5. The method of claim 4, wherein M is greater than 1, said determining a first core word of the M words comprising:

determining the first core word from the at least one candidate word.

6. The method of claim 4, wherein the recommending information based on the first candidate recommendation information and the second candidate recommendation information comprises:

and recommending information based on the target recommendation information.

7. An information recommendation apparatus comprising:

8. The apparatus of claim 7, further comprising:

9. The apparatus of claim 7, further comprising:

10. The apparatus of claim 7, wherein the target search information comprises M terms, the apparatus further comprising:

the recommending module is specifically configured to recommend information based on the first candidate recommending information and the second candidate recommending information.

11. The apparatus of claim 10, wherein M is greater than 1, and the sixth determining module is specifically configured to:

determining the first core word from the at least one candidate word.

12. The apparatus according to claim 10, wherein the recommendation module is specifically configured to:

and recommending information based on the target recommendation information.

13. An electronic device, comprising:

at least one processor; and

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

14. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1-6.

15. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1-6.