Information query method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention relates to an information query method and device, electronic equipment and a storage medium. The method comprises the steps that a query statement input by a user is received; vectorizing the query statement to obtain a semantic vector; similarity calculation is carried out on the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, and each hot topic vector in the at least two hot topic vectors represents a hot topic; based on the at least two similarities, determining a target hotspot theme corresponding to the query statement; based on the target hot topic, determining a tag set corresponding to the query statement; and based on the tag set, obtaining feedback information for feeding back to the user. According to the embodiment of the invention, the tag related to the hotspot theme can be adaptively determined, and the required information is provided for the user based on the tag, so that the pushing precision of the real-time hotspot information can be improved, and more efficient and accurate interaction with the user is realized.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to an information query method, device, electronic device and storage medium. Background Art

[0002] With the explosion of information and the rapid development of social media, hot topics are updated very frequently, and users' focus is constantly changing. Traditional question-answering systems often fail to respond in a timely manner when faced with real-time hot topics, resulting in delayed or irrelevant information obtained by users. Therefore, a technology that can dynamically adapt to changes in real-time hot topics is needed to improve the system's ability to identify and respond to the latest hot content. Existing hotspot detection methods usually rely on fixed keyword matching or historical data training, which makes it difficult to accurately capture emerging hot topics and efficiently associate them with user queries. In addition, the general label distribution mechanism cannot accurately distribute semantically consistent labels for user queries.

[0003] The current technical means for information query on hot topics mainly include:

[0004] Hotspot detection based on keyword matching: This technology detects real-time hotspots by presetting keywords or phrases. Although it is simple to implement, it is prone to omissions or misjudgments when facing emerging vocabulary or semantic changes, and it is difficult to adapt to the dynamically updated hotspot environment;

[0005] Hot spot identification model based on historical data training: By using a large amount of historical data to train the model, it predicts and identifies content that may become hot spots. This method works well in dealing with hot spots in known areas, but it responds slowly to emergencies or new topics and cannot reflect the latest trends in real time;

[0006] Fixed label distribution mechanism: Existing question-answering systems usually rely on fixed label distribution rules and assign labels based on predefined categories or topics. However, they lack dynamic perception of real-time hot spots and find it difficult to flexibly adjust the applicability of labels based on semantic similarity, especially in scenarios with high real-time requirements. Summary of the invention

[0007] In view of this, in order to solve some or all of the above-mentioned technical problems, the embodiments of the present application provide an information query method, device, electronic device and storage medium.

[0008] In a first aspect, an embodiment of the present application provides an information query method, the method comprising: receiving a query statement input by a user; vectorizing the query statement to obtain a semantic vector; calculating similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, wherein each of the at least two hot topic vectors represents a hot topic; based on the at least two similarities, determining a target hot topic corresponding to the query statement; based on the target hot topic, determining a tag set corresponding to the query statement; based on the tag set, obtaining feedback information for providing feedback to the user.

[0009] In one possible implementation, based on a tag set, obtaining feedback information for providing feedback to a user includes: determining at least two recommendation factor scores corresponding to each tag in the tag set, and a recommendation factor weight corresponding to each recommendation factor score; determining a recommendation value corresponding to each tag in the tag set based on the recommendation factor scores and the recommendation factor weights; determining a target tag from the tag set based on the recommendation value; and obtaining feedback information that matches the target tag.

[0010] In a possible implementation, obtaining feedback information matching the target tag includes: determining a data processing interface corresponding to the target tag; and obtaining the feedback information from a target database using the data processing interface.

[0011] In one possible implementation, determining at least two recommendation factor scores corresponding to each tag in a tag set and a recommendation factor weight corresponding to each recommendation factor score includes: for each tag in the tag set, determining a relevance value between the tag and the query statement, and / or determining a popularity value of the tag, and / or determining a user preference value for the tag based on the user's historical behavior data; determining at least two of the relevance value, the popularity value, and the preference value as at least two recommendation factor scores; and obtaining a pre-set recommendation factor weight corresponding to each of the at least two recommendation factor scores.

[0012] In one possible implementation, before receiving a query statement input by a user, the method further includes: obtaining a hot information set; extracting hot words from each hot information in the hot information set, and determining a hot word vector corresponding to the hot word; aggregating the generated hot word vectors to obtain at least two hot word vector sets, wherein each hot word vector set in the at least two hot word vector sets corresponds to a hot topic; performing a comprehensive vector calculation on each hot word vector set in the at least two hot word vector sets to obtain a hot topic vector representing each hot topic.

[0013] In one possible implementation, hot words are extracted from each hot information in the hot information set, including: filtering each hot information in the hot information set to obtain filtered sentences after filtering out non-key content; performing word segmentation, deduplication and stop word removal on the filtered sentences to obtain hot words.

[0014] In one possible implementation, obtaining the hot information set includes: in response to the current moment reaching the hot topic vector update time, obtaining the hot information set generated within a preset time period before the current moment.

[0015] In one possible implementation, a comprehensive vector calculation is performed on each hot word vector set in at least two hot word vector sets to obtain a hot topic vector representing each hot topic, including: for each hot word vector set in at least two hot word vector sets, determining the vector weight corresponding to each hot word vector in the hot word vector set; based on the vector weight, performing weighted summation of the hot word vectors in the hot word vector set to obtain a hot topic vector corresponding to the hot word vector set.

[0016] In one possible implementation, after obtaining feedback information for providing feedback to the user based on a tag set, the method further includes: obtaining user behavior data generated by the user performing corresponding operations on the feedback information; determining the user's acceptance level value for the feedback information based on the user behavior data; if the acceptance level value is less than a preset acceptance level threshold, obtaining improvement strategy information for the feedback information; based on the improvement strategy information, regenerating the feedback information, and outputting the regenerated feedback information.

[0017] In a second aspect, an embodiment of the present application provides an information query device, which includes: a receiving module for receiving a query statement input by a user; a processing module for vectorizing the query statement to obtain a semantic vector; a first calculation module for respectively calculating the similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, wherein each of the at least two hot topic vectors represents a hot topic; a first determination module for determining a target hot topic corresponding to the query statement based on at least two similarities; a second determination module for determining a tag set corresponding to the query statement based on the target hot topic; and a first acquisition module for acquiring feedback information for providing feedback to the user based on the tag set.

[0018] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a memory for storing a computer program; a processor for executing the computer program stored in the memory, and when the computer program is executed, the method of any embodiment of the information query method of the first aspect of the present application is implemented.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, a method as in any embodiment of the information query method of the first aspect described above is implemented.

[0020] In a fifth aspect, an embodiment of the present application provides a computer program, which includes a computer-readable code. When the computer-readable code runs on a device, the processor in the device implements a method as in any embodiment of the information query method of the first aspect mentioned above.

[0021] The information query method, device, electronic device and storage medium provided in the embodiments of the present application obtain a semantic vector by vectorizing the query statement input by the user, and respectively calculate the similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, and determine the target hot topic corresponding to the query statement based on the at least two similarities obtained, and then determine the tag set corresponding to the query statement based on the target hot topic, and obtain feedback information for providing feedback to the user based on the tag set. The embodiments of the present application achieve matching of the semantic vector and the hot topic vector of the user's query statement, and adaptively determine the tags related to the hot topic, and provide the user with the information he needs based on the tags, thereby improving the push accuracy of real-time hot information and achieving more efficient and accurate interaction with the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0023] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0024] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0025] Figure 1 A flowchart of an information query method provided in an embodiment of the present application;

[0026] Figure 2 A flowchart of another information query method provided in an embodiment of the present application;

[0027] Figure 3 A flowchart of another information query method provided in an embodiment of the present application;

[0028] Figure 4 A flowchart of another information query method provided in an embodiment of the present application;

[0029] Figure 5 A flowchart of another information query method provided in an embodiment of the present application;

[0030] Figure 6 A flowchart of another information query method provided in an embodiment of the present application;

[0031] Figure 7 A flowchart of another information query method provided in an embodiment of the present application;

[0032] Figure 8 A schematic diagram of the structure of an information query device provided in an embodiment of the present application;

[0033] Fig. 9 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. It should be noted that unless otherwise specifically stated, the relative arrangement of the components and steps, the numerical expressions and the numerical values ​​described in these embodiments do not limit the scope of the present application.

[0035] Those skilled in the art will understand that the terms "first" and "second" in the embodiments of the present application are only used to distinguish between different steps, devices, modules and other objects, and do not represent any specific technical meanings, nor do they indicate the logical order between them.

[0036] It should also be understood that in this embodiment, “plurality” may refer to two or more than two, and “at least one” may refer to one, two or more than two.

[0037] It should also be understood that any component, data or structure mentioned in the embodiments of the present application can generally be understood as one or more, unless explicitly limited or otherwise indicated in the context.

[0038] In addition, the term "and / or" in this application is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this application generally indicates that the associated objects before and after are in an "or" relationship.

[0039] It should also be understood that the description of the various embodiments in this application focuses on the differences between the various embodiments, and the same or similar aspects thereof can be referenced to each other, and for the sake of brevity, they will not be described one by one.

[0040] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.

[0041] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the above-mentioned technologies, methods, and equipment should be considered as part of the specification.

[0042] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0043] It should be noted that, in the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other. To facilitate the understanding of the embodiments of the present application, the present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0044] In order to solve the technical problems of low efficiency and poor accuracy in user query of hot information in the prior art, the present application provides an information query method, which can more accurately determine the relevant tags of the content required by users for hot topics, thereby improving the efficiency and accuracy of interaction with users.

[0045] Figure 1A schematic flow chart of an information query method provided for an embodiment of the present application. This method can be applied to one or more electronic devices such as vehicles (e.g., intelligent driving vehicles), smart phones, laptops, desktop computers, portable computers, servers, etc. In addition, the execution subject of this method can be hardware or software. When the above-mentioned execution subject is hardware, the execution subject can be one or more of the above-mentioned electronic devices. For example, a single electronic device can execute this method, or a plurality of electronic devices can cooperate with each other to execute this method. When the above-mentioned execution subject is software, this method can be implemented as multiple software or software modules, or as a single software or software module. It is not specifically limited here.

[0046] like Figure 1 As shown, the method specifically includes:

[0047] Step 101: receiving a query statement input by a user.

[0048] In this embodiment, the user can input query statements through various terminal devices, so as to interact with the electronic device executing this method. For example, the user can input the query statement "I want to listen to music" through the voice input function of the vehicle system, and wait for the vehicle system to feedback corresponding information.

[0049] Step 102: vectorize the query statement to obtain a semantic vector.

[0050] In this embodiment, the electronic device can vectorize the query statement through a text vectorization model. For example, the query statement can be vectorized using a pre-trained model such as Word2Vec or BERT, and the obtained semantic vector can represent the semantics of the query statement.

[0051] Step 103 , respectively calculating similarities between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities.

[0052] Each of the at least two hot topic vectors represents a hot topic. A hot topic can be a category name of a certain type of information. Usually, a hot topic can be determined in advance based on the number of views of the information, that is, a set of a certain type of information whose number of views in a certain time period reaches a certain threshold is determined from a preset database as information under a hot topic. A hot topic vector is a new vector obtained by comprehensive vectorization of information under a hot topic, and the vector can represent a hot topic.

[0053] The method for calculating the similarity between the semantic vector and the hot topic vector can be implemented by calculating the distance between the vectors. For example, the cosine distance, Euclidean distance, etc. between the two vectors are calculated, and the similarity is calculated based on the distance.

[0054] Step 104: determine the target hot topic corresponding to the query statement based on at least two similarities.

[0055] In this embodiment, the hot topic corresponding to the maximum similarity among the obtained similarities can be used as the target hot topic, or at least one hot topic with a similarity greater than a preset similarity threshold can be used as the target hot topic.

[0056] Step 105: Determine a tag set corresponding to the query statement based on the target hot topic.

[0057] In this embodiment, the target hot topic can correspond to at least one tag, and each tag represents a further classification of the information under the target hot topic. The tags corresponding to the target hot topic can be pre-stored in a tag database. The tag database contains multiple hot topics, and each hot topic corresponds to at least one tag. According to the corresponding relationship represented in the tag database, the tag set corresponding to the target hot topic can be determined, that is, the tag set corresponding to the query statement. For example, if the user wants to query recent hot foods, after the above steps, it is determined that the target hot topic is "popular foods", then the corresponding tags include "fruits", "vegetables", "fast food", etc. Different tags represent a type of food-related information.

[0058] Step 106: Based on the tag set, feedback information for providing feedback to the user is obtained.

[0059] In this embodiment, relevant information can be obtained from the corresponding data interface according to each tag in the tag set as feedback information to the user; or one or more tags can be selected from the tag set according to preset rules (for example, selecting the tag with the highest query volume, or selecting the tag with the highest relevance to the query statement), and feedback information can be obtained from the data interface corresponding to the selected tag.

[0060] As an example, if a user searches for "healthy food", they can select "fruits" and "vegetables" with higher health weights as the labels to be used from labels such as "fruits", "vegetables", and "fast food", and then obtain relevant feedback information from the data interface corresponding to these two labels.

[0061] The information query method provided by the embodiment of the present application obtains a semantic vector by vectorizing the query statement input by the user, and respectively calculates the similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, and determines the target hot topic corresponding to the query statement based on the at least two similarities obtained, and then determines the tag set corresponding to the query statement based on the target hot topic, and obtains feedback information for providing feedback to the user based on the tag set. The embodiment of the present application realizes matching the semantic vector of the user's query statement and the hot topic vector, and adaptively determines the tags related to the hot topic, and provides the user with the information he needs based on the tags, thereby improving the push accuracy of real-time hot information and achieving more efficient and accurate interaction with the user.

[0062] In some optional implementations of this embodiment, such as Figure 2 As shown, step 106 includes:

[0063] Step 1061 : Determine at least two recommendation factor scores corresponding to each tag in the tag set, and a recommendation factor weight corresponding to each recommendation factor score.

[0064] Among them, each tag can correspond to at least two recommendation factors, and the recommendation factors represent the dimensions referenced when providing feedback to users. For example, the recommendation factor can be the popularity of the tag, or the relevance of the tag to the query statement. Recommendation factors such as popularity and relevance can be represented by numerical values, which are the recommendation factor scores. The recommendation factor weights corresponding to different recommendation factors can be preset.

[0065] Step 1062: Determine the recommendation value corresponding to each tag in the tag set based on the recommendation factor score and the recommendation factor weight.

[0066] Specifically, for a label, the weighted sum of each recommendation factor score and the corresponding recommendation factor weight corresponding to the label can be performed to obtain the recommendation value corresponding to the label. Optionally, the value range can be converted or a preset offset can be added based on the weighted sum to obtain the recommendation value.

[0067] Step 1063: Determine a target tag from the tag set based on the recommendation value.

[0068] Specifically, a tag with the largest recommendation value may be selected from the tag set as the target tag, or at least one tag with a recommendation value greater than a preset recommendation value threshold may be selected as the target tag.

[0069] Step 1064, obtaining feedback information matching the target tag.

[0070] Specifically, the feedback information can be obtained from the data interface corresponding to the target tag.

[0071] This embodiment sets recommendation factors for tags and determines target tags according to recommendation factor scores and recommendation factor weights, thereby achieving quantifiable acquisition of tags with the highest relevance to the information that the user wants to query, thereby improving the accuracy of tag distribution and the accuracy of providing feedback information to users.

[0072] In some optional implementations of this embodiment, such as Figure 3 As shown, step 1064 includes:

[0073] Step 10641, determine the data processing interface corresponding to the target tag.

[0074] Different tags may correspond to different data processing interfaces, and the data processing interfaces may be connected to resources such as databases of corresponding types, data processing models, etc. After the target tag is determined, the corresponding data processing interface may be called.

[0075] Step 10642, using the data processing interface, obtain feedback information from the target database.

[0076] In one embodiment, the data processing interface can connect to a database corresponding to the target tag, and use a query method for the database and a preset query rule to obtain feedback information from the database. For example, if the user's query statement indicates that the user wants to query a certain song, the target tag can be a popular song, and then the song is searched from the corresponding song database.

[0077] In another embodiment, the data processing interface can be connected to a language model, which can generate a dialogue with the user and query the information the user wants. Through the data processing interface, the query statement is input into the model, and the model queries the data required by the user and generates a dialogue statement, and the dialogue statement is used as feedback information.

[0078] This embodiment calls the data processing interface corresponding to the target tag, so that after determining the tag corresponding to the query statement, the location of the data required by the user can be accurately located, thereby improving the efficiency and accuracy of the query.

[0079] In some optional implementations of this embodiment, such as Figure 4 As shown, step 1061 includes:

[0080] Step 10611, for each tag in the tag set, determine the relevance value of the tag to the query statement, and / or determine the popularity value of the tag, and / or determine the user's preference value for the tag based on the user's historical behavior data.

[0081] The above correlation value indicates the correlation between the query and the tag. The higher the value, the higher the correlation, that is, the better the matching between the tag and the query. For example, the similarity between the tag and the query can be determined as the correlation value.

[0082] The popularity value indicates the degree of attention paid to the information related to the tag. For example, the popularity value can be determined by counting the total number of views and the total number of plays of each piece of information corresponding to the tag.

[0083] The historical behavior data indicates the operation of the user on each piece of information corresponding to the tag. For example, the browsing time and browsing times of each piece of information corresponding to the tag can be counted to calculate the user's preference value for the tag.

[0084] Step 10612: Determine at least two of the relevance value, popularity value, and preference value as at least two recommendation factor scores.

[0085] Step 10613: Obtain a preset recommendation factor weight corresponding to each recommendation factor score of at least two recommendation factor scores.

[0086] As an example, suppose the query is "healthy food", and the following three tags are obtained: fruit, vegetable, fast food. If only the relevance value and popularity value (0-10 points) and popularity value (0-10 points) are considered, and the recommendation factor weight corresponding to the relevance value is 0.7, and the recommendation factor weight corresponding to the popularity value is 0.3. If the relevance value of fruit is 8 and the popularity value is 7; the relevance value of vegetable is 9 and the popularity value is 6; the relevance value of fast food is 3 and the popularity value is 9, the final score of each tag is calculated as follows:

[0087] Fruit: (8*0.7)+(7*0.3)=5.6+2.1=7.7

[0088] Vegetables: (9*0.7)+(6*0.3)=6.3+1.8=8.1

[0089] Fast food: (3*0.7)+(9*0.3)=2.1+2.7=4.8

[0090] The calculation results are ranked as follows: vegetables (8.1), fruits (7.7), and fast food (4.8). If the preset threshold is 6, the query statement "healthy food" can be assigned the two most relevant labels "vegetables" and "fruits" as target labels.

[0091] This embodiment can comprehensively determine the recommendation value corresponding to each tag from multiple aspects by determining multiple different recommendation factor scores, so that the assigned tags are more compatible with the user's query intention and the accuracy of the generated feedback information is improved.

[0092] In some optional implementations of this embodiment, such as Figure 5 As shown, before step 101, the method further includes:

[0093] Step 107, obtaining a hotspot information set.

[0094] The hot information set may be a set of information whose browsing volume is greater than a browsing volume threshold in a recent period, or a set of a preset number of information obtained after being sorted by browsing volume. As an example, the hot information set may be obtained from a recent hot list.

[0095] Step 108, extracting hot words from each hot information in the hot information set, and determining the hot word vector corresponding to the hot word.

[0096] Specifically, each hot information may be segmented, and words with specific meanings may be selected as hot words.

[0097] Optionally, you can extract hot words by following the steps below:

[0098] Each hot information in the hot information set is filtered to obtain a filtered sentence after filtering out non-key content; the filtered sentence is segmented, de-duplicated and screened to obtain hot words.

[0099] Among them, non-key content can be content that has no impact on the main content of the hot information. For example, if the hot information is an article on a web page, the content outside the main text is non-key content. Then, after the filtered sentences are segmented, each word is deduplicated and filtered. The filtering operation can be to remove some modal particles, meaningless words, stop words, etc., so as to obtain hot words.

[0100] By filtering, segmenting, deduplicating and screening hot information, hot words can more accurately represent some recent hot topics, thereby improving the efficiency and accuracy of user information query.

[0101] After obtaining the hot words, the hot words can be vectorized using pre-trained models such as Word2Vec and BERT to obtain the hot word vectors.

[0102] Step 109, aggregate the generated hot word vectors to obtain at least two hot word vector sets.

[0103] Among them, each hot word vector set in at least two hot word vector sets corresponds to a hot topic.

[0104] Through the above steps, multiple hot word vectors can be obtained. According to the relevant clustering algorithm in this field, these hot word vectors can be clustered, and the hot word vectors under each category can form a hot word vector set.

[0105] Step 110, performing comprehensive vector calculation on each of the at least two hot word vector sets to obtain a hot topic vector representing each hot topic.

[0106] Specifically, for a set of hot word vectors, we can take the average of the values ​​at the same position among the values ​​included in each hot word vector in the set, or determine the frequency of occurrence of each hot word through TF-IDF, determine the weight value according to the frequency, and then perform weighted summation on the values ​​at the same position in each hot word vector to obtain the hot topic vector corresponding to each hot word vector set.

[0107] This embodiment can realize timely processing of hot information by pre-processing a large number of hot information sets. The calculated hot topic vector can more accurately represent recent hot topics, thereby helping to push information to users in a more targeted manner and improving the accuracy of users obtaining the required information.

[0108] In some optional implementations of this embodiment, step 107 may be performed as follows:

[0109] In response to the current moment reaching the hot topic vector update time, a hot information set generated within a preset time period before the current moment is obtained.

[0110] The hot topic vector update time may be the time when a preset update cycle is reached, or the time when the hot topic vector is updated manually. The preset time period may be set arbitrarily, such as one day, one week, etc.

[0111] This embodiment obtains a set of hot information generated within a preset time period before the current moment. Its operation is similar to that of a sliding window, that is, the time period covered by the sliding window is fixed. As time goes by, the sliding window slides forward, thereby continuously updating the hot information set, ensuring that the user queries the latest hot information, and improving the ability to capture events and development trends.

[0112] In some optional implementations of this embodiment, such as Figure 6 As shown, step 110 includes:

[0113] Step 1101: for each hot word vector set in at least two hot word vector sets, determine a vector weight corresponding to each hot word vector in the hot word vector set.

[0114] The vector weight corresponding to each hot word vector can be pre-set. For example, for the hot topic "artificial intelligence", several related hot words and their hot word vectors (simplified into two-dimensional vectors) are as follows:

[0115] "machine learning": [0.8, 0.6];

[0116] “deeplearning”: [0.9, 0.7];

[0117] "neural network": [0.7, 0.8];

[0118] "big data": [0.6, 0.5].

[0119] The vector weights can be determined based on the importance or frequency of these words. For example, the vector weights are:

[0120] “Machine Learning”: 0.3;

[0121] “Deep Learning”: 0.3;

[0122] "neural network": 0.25;

[0123] “Big Data”: 0.15.

[0124] Step 1102, based on the vector weight, weighted sum is performed on the hot word vectors in the hot word vector set to obtain the hot topic vector corresponding to the hot word vector set.

[0125] The calculation process of weighted summation is as follows:

[0126] (0.3*[0.8,0.6])+(0.3*[0.9,0.7])+(0.25*[0.7,0.8])+(0.15*[0.6,0.5])=[0.24,0.18]+[0.27,0.21]+[0.175,0.2]+[0.09,0.075]=[0.775,0.665]. Therefore, the hot topic vector of the hot topic "artificial intelligence" is expressed as [0.775,0.665].

[0127] This embodiment sets vector weights and uses the vector weights to perform comprehensive calculations on each hot word vector set, so that the hot topic vector can more accurately represent a hot topic, which helps to more accurately match user query information.

[0128] In some optional implementations of this embodiment, such as Figure 7 As shown, after step 106, the method further includes:

[0129] Step 111, obtaining user behavior data generated by the user performing corresponding operations in response to the feedback information.

[0130] Among them, user behavior data can be data generated by the user performing corresponding operations according to the operation method indicated by the feedback information after obtaining feedback information. For example, the query sentence input by the user indicates that the user wants to play music. After the music is fed back to the user, the user's behavior in the subsequent period of time (for example, 30 seconds) can be recorded: if the user does not perform any operation, it is determined that the user is satisfied with the feedback information; if the user manually switches or stops the music, it is determined that the user is dissatisfied with the feedback information.

[0131] Step 112: Determine the user's acceptance level of the feedback information based on the user behavior data.

[0132] The acceptance value can be obtained by collecting statistics on user behavior data. For example, the number of times the user is satisfied or dissatisfied with the feedback information recently (for example, the last 100 times) is recorded, and the calculation formula of the acceptance value can be: number of satisfied times / total times*100%.

[0133] Step 113: If the acceptance level value is less than a preset acceptance level threshold, obtain improvement strategy information for the feedback information.

[0134] Specifically, if the acceptance value is less than the acceptance threshold, it means that the user's acceptance of the feedback information is low, and the interaction method needs to be improved. The above-mentioned improvement strategy information is used to instruct the electronic device how to improve the query method. For example, the improvement strategy information can be a pre-written script file.

[0135] Step 114: regenerate feedback information based on the improved strategy information, and output the regenerated feedback information.

[0136] The interaction method can be adjusted based on the improvement strategy information. For example, when a user searches for a song, the user's dissatisfaction often occurs when the electronic device fails to correctly understand the name of the song the user wants to listen to. Therefore, the improvement strategy information can be configured. After the feedback information is determined, a prompt message can be output to let the user confirm whether the feedback information is correct. That is, the prompt message is output first: "Will play 'xxx', okay?" The user can reply "yes" or "no" by voice, and the electronic device can re-query and obtain new feedback information. In this way, after a certain period of monitoring, if it is determined that the user's acceptance value reaches the preset acceptance threshold, it is determined that the improvement strategy information is valid.

[0137] This embodiment continuously records user behavior data and then adjusts the interaction process, thereby achieving continuous improvement of the query method and improving the accuracy of the query.

[0138] Figure 8 A schematic diagram of the structure of an information query device provided for an embodiment of the present application. Specifically comprising: a receiving module 801, for receiving a query statement input by a user; a processing module 802, for vectorizing the query statement to obtain a semantic vector; a first calculation module 803, for respectively calculating the similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, wherein each of the at least two hot topic vectors represents a hot topic; a first determination module 804, for determining a target hot topic corresponding to the query statement based on at least two similarities; a second determination module 805, for determining a tag set corresponding to the query statement based on the target hot topic; a first acquisition module 806, for obtaining feedback information for providing feedback to the user based on the tag set.

[0139] In one possible embodiment, the acquisition module includes: a first determination unit, used to determine at least two recommendation factor scores corresponding to each tag in the tag set, and a recommendation factor weight corresponding to each recommendation factor score; a second determination unit, used to determine the recommendation value corresponding to each tag in the tag set based on the recommendation factor score and the recommendation factor weight; a third determination unit, used to determine the target tag from the tag set based on the recommendation value; and an acquisition unit, used to obtain feedback information matching the target tag.

[0140] In a possible implementation, the acquisition unit includes: a first determination subunit, used to determine a data processing interface corresponding to the target tag; and a first acquisition subunit, used to acquire feedback information from a target database using the data processing interface.

[0141] In one possible embodiment, the first determination unit includes: a second determination subunit, which is used to determine, for each tag in the tag set, a relevance value between the tag and the query statement, and / or determine the popularity value of the tag, and / or determine the user's preference value for the tag based on the user's historical behavior data; a third determination subunit, which is used to determine at least two of the relevance value, the popularity value and the preference value as at least two recommendation factor scores; and a second acquisition subunit, which is used to obtain a pre-set recommendation factor weight corresponding to each of the at least two recommendation factor scores.

[0142] In one possible embodiment, the device also includes: a second acquisition module, used to acquire a hot information set; an extraction module, used to extract hot words from each hot information in the hot information set, and determine the hot word vector corresponding to the hot word; an aggregation module, used to aggregate the generated hot word vectors to obtain at least two hot word vector sets, wherein each hot word vector set in the at least two hot word vector sets corresponds to a hot topic; a second calculation module, used to perform comprehensive vector calculation on each hot word vector set in the at least two hot word vector sets to obtain a hot topic vector representing each hot topic.

[0143] In one possible implementation, the extraction module includes: a filtering unit, which is used to filter each hot information in the hot information set to obtain a filtered statement after filtering out non-key content; a processing unit, which is used to perform word segmentation, deduplication and stop word removal on the filtered statement to obtain hot words.

[0144] In a possible implementation, the second acquisition module is further used to: in response to the current moment reaching the hot topic vector update time, acquire a set of hot information generated within a preset time period before the current moment.

[0145] In one possible implementation, the second calculation module includes: a fourth determination unit, used to determine the vector weight corresponding to each hot word vector in at least two hot word vector sets for each hot word vector set; a calculation unit, used to perform weighted summation of the hot word vectors in the hot word vector set based on the vector weight to obtain a hot topic vector corresponding to the hot word vector set.

[0146] In one possible embodiment, the device also includes: a third acquisition module, used to obtain user behavior data generated by the user's corresponding operations on the feedback information; a third determination module, used to determine the user's acceptance level value for the feedback information based on the user behavior data; a fourth acquisition module, used to obtain improvement strategy information for the feedback information if the acceptance level value is less than a preset acceptance level threshold; a generation module, used to regenerate the feedback information based on the improvement strategy information, and output the regenerated feedback information.

[0147] The information query device provided in this embodiment can be as follows Figure 8 The information query device shown in can execute all the steps of the above information query methods, thereby achieving the technical effects of the above information query methods. Please refer to the above related description for details. For the sake of brevity, it will not be repeated here.

[0148] Fig. 9 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application, Fig. 9The electronic device 900 shown includes: at least one processor 901, a memory 902, at least one network interface 904 and other user interfaces 903. The various components in the electronic device 900 are coupled together via a bus system 905. It is understood that the bus system 905 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 905 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, the bus system 905 is not described in detail. Fig. 9 Various buses are labeled as bus system 905.

[0149] The user interface 903 may include a display, a keyboard, or a pointing device (eg, a mouse, a trackball, a touch pad, or a touch screen).

[0150] It can be understood that the memory 902 in the embodiment of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct RAM bus random access memory (DRRAM). The memory 902 described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0151] In some implementations, the memory 902 stores the following elements, executable units or data structures, or a subset thereof, or an extended set thereof: an operating system 9021 and an application program 9022 .

[0152] The operating system 9021 includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks. The application 9022 includes various application programs, such as a media player (Media Player), a browser (Browser), etc., for implementing various application services. The program for implementing the method of the embodiment of the present application can be included in the application 9022.

[0153] In this embodiment, by calling the program or instruction stored in the memory 902, specifically, the program or instruction stored in the application 9022, the processor 901 is used to execute the method steps provided by each method embodiment, for example, including:

[0154] Receive a query statement input by a user; vectorize the query statement to obtain a semantic vector; calculate the similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, wherein each of the at least two hot topic vectors represents a hot topic; based on the at least two similarities, determine a target hot topic corresponding to the query statement; based on the target hot topic, determine a tag set corresponding to the query statement; based on the tag set, obtain feedback information for providing feedback to the user.

[0155] The method disclosed in the above embodiment of the present application can be applied to the processor 901, or implemented by the processor 901. The processor 901 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 901. The above processor 901 can be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to execute, or the hardware and software units in the decoding processor can be executed. The software unit can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 902, and the processor 901 reads the information in the memory 902 and completes the steps of the above method in combination with its hardware.

[0156] It is understood that the embodiments described herein may be implemented in hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementation, the processing unit may be implemented in one or more application specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPDevice, DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), general purpose processors, controllers, microcontrollers, microprocessors, other electronic units for performing the above functions of the present application, or a combination thereof.

[0157] For software implementation, the technology described above in this article can be implemented by a unit that performs the functions described above in this article. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or outside the processor.

[0158] The electronic device provided in this embodiment may be Fig. 9 The electronic device shown in can execute all the steps of the information query methods described above, thereby achieving the technical effects of the information query methods described above. Please refer to the above related description for details. For the sake of brevity, it will not be repeated here.

[0159] The embodiment of the present application also provides a storage medium (computer-readable storage medium). The storage medium here stores one or more programs. The storage medium may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a read-only memory, a flash memory, a hard disk or a solid-state drive; the memory may also include a combination of the above-mentioned types of memory.

[0160] When one or more programs in the storage medium can be executed by one or more processors, the above-mentioned information query method executed on the electronic device side can be implemented.

[0161] The processor is used to execute the program stored in the memory to implement the following steps of the information query method performed on the electronic device side:

[0162] Receive a query statement input by a user; vectorize the query statement to obtain a semantic vector; calculate the similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, wherein each of the at least two hot topic vectors represents a hot topic; based on the at least two similarities, determine a target hot topic corresponding to the query statement; based on the target hot topic, determine a tag set corresponding to the query statement; based on the tag set, obtain feedback information for providing feedback to the user.

[0163] The professionals should also be further aware that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0164] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0165] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0166] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. An information query method, characterized in that: The method comprises: Receive query statements input by users; Vectorizing the query statement to obtain a semantic vector; Calculating similarities between the semantic vector and at least two preset hot topic vectors respectively to obtain at least two similarities, wherein each of the at least two hot topic vectors represents a hot topic; Based on the at least two similarities, determining a target hot topic corresponding to the query statement; Based on the target hot topic, determine a tag set corresponding to the query statement; Based on the tag set, feedback information for providing feedback to the user is obtained.

2. The method according to claim 1, characterized in that The obtaining, based on the tag set, feedback information for providing feedback to the user includes: Determine at least two recommendation factor scores corresponding to each tag in the tag set, and a recommendation factor weight corresponding to each recommendation factor score; Determining a recommendation value corresponding to each tag in the tag set based on the recommendation factor score and the recommendation factor weight; Based on the recommendation value, determining a target tag from the tag set; Obtain feedback information matching the target label.

3. The method according to claim 2, characterized in that The obtaining of feedback information matching the target tag includes: Determine a data processing interface corresponding to the target tag; The feedback information is obtained from the target database using the data processing interface.

4. The method according to claim 2, characterized in that: The determining of at least two recommendation factor scores corresponding to each tag in the tag set and a recommendation factor weight corresponding to each recommendation factor score includes: For each tag in the tag set, determining a relevance value of the tag to the query statement, and / or determining a popularity value of the tag, and / or determining a preference value of the user for the tag based on the historical behavior data of the user; Determining at least two of the relevance value, the popularity value, and the preference value as the at least two recommendation factor scores; A preset recommendation factor weight corresponding to each recommendation factor score of the at least two recommendation factor scores is obtained.

5. The method according to claim 1, characterized in that Before receiving the query statement input by the user, the method further includes: Get hotspot information collection; Extracting a hot word from each hot information in the hot information set, and determining a hot word vector corresponding to the hot word; Aggregating the generated hot word vectors to obtain at least two hot word vector sets, wherein each hot word vector set in the at least two hot word vector sets corresponds to a hot topic; A comprehensive vector calculation is performed on each of the at least two hot word vector sets to obtain a hot topic vector representing each hot topic.

6. The method according to claim 5, characterized in that The step of extracting hot words from each hot information in the hot information set includes: Filtering each hot information in the hot information set to obtain a filtered statement after filtering out non-key content; The filtered sentences are segmented, duplicated and stop words are removed to obtain the hot words.

7. The method according to claim 5, characterized in that The step of obtaining a hotspot information set includes: In response to the current moment reaching the hot topic vector update time, a hot information set generated within a preset time period before the current moment is obtained.

8. The method according to claim 5, characterized in that The step of performing comprehensive vector calculation on each of the at least two hot word vector sets to obtain a hot topic vector representing each hot topic includes: For each hot word vector set in the at least two hot word vector sets, determining a vector weight corresponding to each hot word vector in the hot word vector set; Based on the vector weights, the hot word vectors in the hot word vector set are weighted summed to obtain the hot topic vector corresponding to the hot word vector set.

9. The method according to claim 1, characterized in that: After obtaining feedback information for providing feedback to the user based on the tag set, the method further includes: Acquiring user behavior data generated by the user performing corresponding operations in response to the feedback information; Determining, based on the user behavior data, a value of the user's acceptance level of the feedback information; If the acceptance level value is less than a preset acceptance level threshold, obtaining improvement strategy information for the feedback information; Based on the improvement strategy information, feedback information is regenerated, and the regenerated feedback information is output.

10. An information query device, characterized in that: The device comprises: A receiving module, used for receiving a query statement input by a user; A processing module, used for vectorizing the query statement to obtain a semantic vector; A first calculation module is used to calculate similarity between the semantic vector and at least two preset hot topic vectors to obtain at least two similarities, wherein each of the at least two hot topic vectors represents a hot topic; A first determination module, configured to determine a target hot topic corresponding to the query statement based on the at least two similarities; A second determination module is used to determine a tag set corresponding to the query statement based on the target hot topic; The first acquisition module is used to acquire feedback information for providing feedback to the user based on the tag set.

11. An electronic device, characterized in that: include: Memory for storing computer programs; A processor is used to execute the computer program stored in the memory, and when the computer program is executed, the information query method described in any one of claims 1 to 9 is implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the information query method described in any one of claims 1 to 9 is implemented.

Images