CN113254587A - Search text recognition method and device, computer equipment and storage medium - Google Patents

Search text recognition method and device, computer equipment and storage medium Download PDF

Info

Publication number
CN113254587A
CN113254587A CN202110605909.XA CN202110605909A CN113254587A CN 113254587 A CN113254587 A CN 113254587A CN 202110605909 A CN202110605909 A CN 202110605909A CN 113254587 A CN113254587 A CN 113254587A
Authority
CN
China
Prior art keywords
search text
input
output result
submodel
word
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110605909.XA
Other languages
Chinese (zh)
Other versions
CN113254587B (en
Inventor
赵海林
魏强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing QIYI Century Science and Technology Co Ltd
Original Assignee
Beijing QIYI Century Science and Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing QIYI Century Science and Technology Co Ltd filed Critical Beijing QIYI Century Science and Technology Co Ltd
Priority to CN202110605909.XA priority Critical patent/CN113254587B/en
Publication of CN113254587A publication Critical patent/CN113254587A/en
Application granted granted Critical
Publication of CN113254587B publication Critical patent/CN113254587B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3344—Query execution using natural language analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/237—Lexical tools
    • G06F40/247—Thesauruses; Synonyms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/30—Semantic analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

The embodiment of the invention relates to a method and a device for identifying a search text, computer equipment and a storage medium, wherein the method comprises the following steps: determining a participle set and a similar meaning word set corresponding to a search text to be identified; performing multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set; taking the participle set and the similar meaning word set as a second input set; inputting the third input set into a second submodel to obtain a third output result; and inputting the first input set, the second input set and the third output result into the first submodel so that the first submodel outputs the identification result of the search text, and performing word segmentation and word expansion processing on the search text so that the semantics of the search text are not deficient, the semantic range of the search text is expanded, and accurate identification is realized on the short and deficient search text as a whole.

Description

Search text recognition method and device, computer equipment and storage medium
Technical Field
The embodiment of the invention relates to the field of text classification, in particular to a method and a device for identifying a search text, computer equipment and a storage medium.
Background
With the development of the internet, people's lives, works, and the like are closely related to the internet, so that people can generate a large amount of data on the internet, such as texts, voices, images, videos, and the like, information search becomes one of the most important requirements of internet users, and a corresponding search operation is performed by inputting a search text in a search engine.
The identification supervision of the search words is beneficial to purifying the network environment, improving the user experience, and avoiding related policy and law risks, the identification of the search text is equivalent to a text classification task, and the identification task of the search words cannot be effectively solved by a common long text classification means due to the characteristics of relatively short search words, insufficient semantic information and the like, so that accurate judgment cannot be realized in the judgment of whether the search text carries low-custom attributes.
Disclosure of Invention
In view of the above, to solve the technical problems or some technical problems, embodiments of the present invention provide a method and an apparatus for identifying a search text, a computer device, and a storage medium.
In a first aspect, an embodiment of the present invention provides a method for identifying a search text, where the search text identification model includes a first sub-model and a second sub-model, and the method includes:
determining a participle set and a similar meaning word set corresponding to a search text to be identified;
performing multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set;
taking the word segmentation set and the similar meaning word set as a second input set;
inputting the third input set into a second submodel to obtain a third output result;
inputting the first input set, the second input set and the third output result into the first submodel to cause the first submodel to output the recognition result of the search text.
In one possible implementation, the determining a set of segmented words and a set of similar words corresponding to the search text to be recognized includes:
performing word segmentation processing on the search text to obtain a word segmentation set containing a plurality of words and/or phrases; and performing semantic matching on each word and/or phrase in the word segmentation set, and taking one or more similar words with the similarity greater than a set threshold value with the word meaning and/or one or more similar phrases with the similarity greater than the set threshold value with the phrase meaning as a similar word set corresponding to the word segmentation set.
In one possible embodiment, the first set of inputs includes: a first vector set and a second vector set;
the third set of inputs includes at least two of: a first set of vectors, a second set of vectors, a third set of vectors, a fourth set of vectors, or a fifth set of vectors;
the first set of vectors comprises: vector representation of a word dimension corresponding to each participle in the participle set;
the second set of vectors includes: vector representation of character dimension corresponding to each participle in the participle set;
the third set of vectors comprises: vector representation of a text dimension corresponding to the search text;
the fourth set of vectors comprises: vector representation of a text distribution dimension corresponding to the search text;
the fifth set of vectors comprises: each participle in the participle set corresponds to vector representation of participle semantics.
In one possible embodiment, the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
the inputting the first input set, the second input set, and the third output result into the first submodel to cause the first submodel to output the recognition result of the search text includes:
inputting the first input set and the third output result into the neural network portion to cause the neural network portion to output the first output result corresponding to the search text;
inputting the second input set and the third output result into the linear part to enable the linear part to output the second output result corresponding to the search text;
and taking the first output result and the second output result as the recognition result of the search text.
In one possible embodiment, the second submodel is: a third output result output by the Xgboost model comprises a probability value corresponding to the search text and leaf node number characteristics corresponding to the word segmentation set and the similar word set;
the inputting the first input set and the third output result to the neural network portion to cause the neural network portion to output the first output result corresponding to the search text includes:
inputting the first input set into the neural network part, so that the neural network part extracts local features of the word segmentation set from the first input set, fuses the local features to obtain global features of the word segmentation set, and obtains the first output result of the search text by combining the global features with the probability value in the third output result.
In one possible embodiment, the inputting the second input set and the third output result into the linear part to cause the linear part to output the second output result corresponding to the search text includes:
and inputting the second input set into the linear part, so that the linear part obtains the second output result of the search text according to the participle set and the similar meaning word set and by combining the leaf node number characteristics in the third output result.
In one possible embodiment, the search text recognition model is constructed by:
performing word segmentation on the search text aiming at any search text in a training sample to obtain a word segmentation set corresponding to the search text, and performing word expansion processing on each word segmentation in the word segmentation set to obtain a near meaning word set corresponding to the word segmentation set; determining a first input set of the first submodel according to the word segmentation set, determining a second input set of the first submodel according to the word segmentation set and the word similarity set, and determining a third input set of the second submodel according to the word segmentation set; inputting the third input set into the second submodel to enable the second submodel to output a corresponding third output result; and training the first sub-model by using the first input set, the second input set and the third output result, and determining that the training of the search text recognition model is finished when the training of the first sub-model is finished.
In one possible embodiment, the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
the training the first submodel using the first set of inputs, the second set of inputs, and the third output results includes:
inputting the first input set and the third output result into the neural network part, and training the neural network part; and inputting the second input set and the third output result into the linear part, and training the linear part.
In one possible embodiment, a loss function is constructed from a first output of the neural network portion and a second output of the linear portion; using the loss function to assist the first sub-model to train;
wherein, the condition that the training of the first sub-model is completed comprises: the loss function satisfies a preset convergence condition.
In a second aspect, an embodiment of the present invention provides an apparatus for identifying a search text, where the search text identification model includes a first sub-model and a second sub-model, and the apparatus includes:
the set determining module is used for determining a participle set and a similar meaning word set corresponding to the search text to be identified;
the extraction module is used for carrying out multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set;
the set determination module is further configured to use the participle set and the synonym set as a second input set;
the output determining module is used for inputting the third input set into a second submodel to obtain a third output result;
a result determination module, configured to input the first input set, the second input set, and the third output result into the first submodel, so that the first submodel outputs the recognition result of the search text.
In a third aspect, an embodiment of the present invention provides a computer device, including: a processor and a memory, wherein the processor is configured to execute a building program of a search text recognition model stored in the memory to implement the method for recognizing a search text according to any one of the first aspect.
In a fourth aspect, an embodiment of the present invention provides a storage medium, where the storage medium stores one or more programs, and the one or more programs are executable by one or more processors to implement the method for identifying a search text according to any one of the above first aspects.
According to the identification scheme of the search text provided by the embodiment of the invention, the segmentation and word expansion processing are carried out on the search text, so that the semantic meaning of the search text is not deficient, the semantic scope of the search text is expanded, multi-dimensional feature extraction is carried out on a segmentation set and a similar word set, the model carries out classification and identification on the extracted feature vectors, the identification result of the search text is obtained, and the short and semantic deficient search text is accurately identified on the whole.
Drawings
Fig. 1 is a schematic flowchart of a method for constructing a search text recognition model according to an embodiment of the present invention;
fig. 2 is a schematic flow chart illustrating a process of performing word segmentation and expansion processing on the search text to obtain a word segmentation set and a similar word set corresponding to the search text according to an embodiment of the present invention;
FIG. 3 is a flow chart illustrating a process for determining an input set according to an embodiment of the present invention;
FIG. 4 is a schematic flow chart illustrating obtaining a third output result according to an embodiment of the present invention;
FIG. 5 is a schematic flow chart illustrating the training of a first sub-model according to an embodiment of the present invention;
FIG. 6 is a schematic structural diagram of a search text recognition model according to an embodiment of the present invention;
FIG. 7 is a flowchart illustrating an identification method for a search text according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of an apparatus for recognizing search texts according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
For the convenience of understanding of the embodiments of the present invention, the following description will be further explained with reference to specific embodiments, which are not to be construed as limiting the embodiments of the present invention.
Fig. 1 is a schematic flowchart of a method for constructing a search text recognition model according to an embodiment of the present invention, and as shown in fig. 1, the method specifically includes:
s11, performing word segmentation and expansion processing on the search text aiming at any search text in the training sample to obtain a word segmentation set and a similar word set corresponding to the search text.
The construction method of the search text recognition model provided by the embodiment of the invention is applied to training of the search text recognition model, the search text recognition model comprises a first submodel and a second submodel, the first submodel comprises a neural network part and a linear part, for example, the first submodel can be: the Wide & Deep model, the second submodel may be: the Xgboost model, the neural network portion, may be: deep side part in Wide & Deep model, linear part can be: wide side portion in Wide & Deep model.
Further, the second sub-model may be a pre-trained model, the first sub-model is a model to be trained, and the construction of the search text recognition model is completed by training the first sub-model.
The search text related to this embodiment may be a search word input to a search engine, and has features of short length, poor semantics, and the like, and the search text recognition model mainly recognizes some search words carrying special meanings, for example, the search text recognition model may recognize search words carrying "soft pornography" information, and correspondingly, the training sample used for training the search text recognition model may be a search text which has been labeled "soft pornography" in advance.
Further, for any search text in the training sample, performing word segmentation processing on the search text, specifically, performing word segmentation operation on the search text by using a word segmentation tool (such as jieba, SnowNLP, THULAC, NLPIR, etc.), and obtaining a word segmentation set after performing word segmentation processing on the search text, where the word segmentation set includes a plurality of words and/or word groups.
And carrying out expansion processing on each word and phrase after word segmentation processing of the search text, wherein the expansion processing aims at expanding the semantics of the search text, and a plurality of near-meaning words corresponding to each word and a plurality of near-meaning phrases corresponding to each phrase are obtained by carrying out near-meaning word matching on each word and phrase so as to obtain a near-meaning word set.
S12, determining a first input set of the first sub-model according to the word segmentation set, determining a second input set of the first sub-model according to the word segmentation set and the similar meaning word set, and determining a third input set of the second sub-model according to the word segmentation set.
In this embodiment, according to the word segmentation set and the word similarity set determined by searching the text in the above steps, the inputs of the first sub-model and the second sub-model are sequentially determined, and the first sub-model is divided into a neural network part and a linear part, so that the first sub-model includes two inputs.
Further, the set of tokens is determined as a first input set (input of the neural network part) for the first submodel, the set of tokens and the set of near-sense words are determined as a second input set (input of the linear part) for the first submodel, and the set of tokens is determined as a third input set for the second submodel.
And S13, inputting the third input set into the second submodel to enable the second submodel to output a corresponding third output result.
S14, training the first sub-model by using the first input set, the second input set and the third output result, and determining that the training of the search text recognition model is completed when the training of the first sub-model is completed.
Specifically, the training process for the first submodel may include: inputting the first input set and the third output result into the neural network part, and training the neural network part; and inputting the second input set and the third output result into the linear part, and training the linear part.
In an alternative of the embodiment of the invention, a loss function is constructed from the first output result of the neural network portion and the second output result of the linear portion; and utilizing the loss function to assist the first sub-model to train.
Accordingly, S14 may specifically include the following sub-steps:
s141 (not shown in the figure), training the first sub-model by using the first input set and the third output result as the input of the neural network part, and using the second input set and the third output result as the input of the linear part.
And inputting the third input set into the second submodel to enable the second submodel to output a corresponding third output result, namely outputting a prediction probability value corresponding to the search text and leaf node number characteristics corresponding to the word segmentation set and the similar meaning word set by using the Xgboost model through the third input set.
And taking the first input set and the third output result as the input of a neural network part in the first submodel, taking the second input set and the third output result as the input of a linear part in the first submodel, and training the first submodel.
And S142 (not shown in the figure), in the process of training the first submodel, constructing a loss function according to the first output result of the neural network part and the second output result of the linear part in the first submodel, wherein the loss function is used for representing the training result of the search text recognition model.
S143 (not shown in the figure), training the search text recognition model based on the loss function, until the loss function meets a preset convergence condition, and determining that the search text recognition model is trained.
In this embodiment, the training of the search text recognition model is completed by using the training first submodel, so that the training result of the first submodel is used as the training result of the search text recognition model, and in the training process of the first submodel, a loss function for representing the training result of the first submodel is constructed according to the first output result of the neural network part and the second output result of the linear part, and the loss function can also represent the training result of the search text recognition model.
And when the loss function does not meet the preset convergence condition, continuing to train the search text recognition model by adjusting the operation parameters of the first sub-model until the loss function meets the preset convergence condition.
According to the construction method of the search text recognition model provided by the embodiment of the invention, the segmentation and word expansion processing are carried out on the search text, so that the semantics of the search text are not deficient any more, the semantic range of the search text is expanded, the multi-dimensional feature extraction is carried out on the segmentation set and the similar word set, the model carries out classification recognition on the extracted feature vectors, the recognition result of the search text is obtained, and the short and semantically deficient search text is accurately recognized on the whole.
Fig. 2 is a schematic flow chart of performing word segmentation and expansion processing on the search text to obtain a word segmentation set and a similar word set corresponding to the search text according to the embodiment of the present invention, as shown in fig. 2, specifically including:
s21, performing word segmentation processing on the search text to obtain a word segmentation set containing a plurality of words and/or phrases.
S22, performing semantic matching on each word and/or phrase in the word segmentation set, and taking one or more similar words with the word semantic similarity larger than a set threshold and/or one or more similar phrases with the phrase semantic similarity larger than the set threshold as the similar word set corresponding to the word segmentation set.
In this embodiment, a segmentation tool is used to perform segmentation processing on a search text to obtain a segmentation set of a plurality of words and/or phrases (hereinafter, the words and phrases are collectively referred to as token sequences), and a rule of segmentation may be that the obtained words or phrases are all minimum and cannot be split again, for example, the search text is: the corresponding word segmentation set is as follows: "Yanxi", "attack and conquer" and "big ending".
Further, for each word and/or phrase in the word segmentation set, performing near-meaning word matching from a word stock, and taking one or more near-meaning words with similarity greater than a set threshold and/or one or more near-meaning phrases with similarity greater than a set threshold as a near-meaning word set corresponding to the word segmentation set.
In the process of matching near-sense words, the maximum number of near-sense words of each word or phrase may also be set, and the size of the set threshold is adjusted according to the matching result, for example, the maximum number of near-sense words of each word or phrase is set to 100, the set threshold is set to 80%, and for a word or phrase, the matching result of near-sense words may be: "big end" - "last episode", and so on.
Fig. 3 is a schematic flowchart of determining an input set according to an embodiment of the present invention, and as shown in fig. 3, the method specifically includes:
in this embodiment, the first input set includes two preset feature vector sets, and the two preset feature vectors may be: a first set of vectors and a second set of vectors, the third set of inputs including at least the first set of inputs.
The set of vectors in the input set may be extracted using a model, in an example, the first set of vectors and the second set of vectors may be extracted using a first model, the third set of vectors and the fourth set of vectors may be extracted using a second model, and the fifth set of vectors may be extracted using a third model, wherein the first set of vectors includes: vector representation of a word dimension corresponding to each participle in the participle set; the second set of vectors includes: vector representation of character dimension corresponding to each participle in the participle set; the third set of vectors comprises: vector representation of a text dimension corresponding to the search text; the fourth set of vectors comprises: vector representation of a text distribution dimension corresponding to the search text; the fifth set of vectors comprises: each participle in the participle set corresponds to vector representation of participle semantics.
In an example, the first set of vectors may be: a set of token level embeddin vectors, where the second set of vectors may be: a set of char level embedding vectors, and the third set of vectors may be: the set of query-level embedding vectors, and the fourth set of vectors may be: a set of query-channel distribution vectors, and a fifth set of vectors may be: the first model of the set of DSSM token embedding vectors may be: the ngram2vec model, the second model may be: the FastText model and the third model may be DSSM models, and the five sets of vector sets may be obtained by using other models (e.g., token level embedded vectors extracted by a Bert model) or other technical means besides the three mentioned models, which is not specifically limited in this embodiment.
In the following, the extraction of five sets of vectors by the first model, the second model, and the third model will be described as an example.
S31, determining a first vector set and a second vector set corresponding to the participle set by using a first model, and taking the first vector set and the second vector set as the first input set of the first submodel.
And S32, taking the word segmentation set and the similar meaning word set as a second input set of the first sub-model.
And S33, determining a third vector set and a fourth vector set corresponding to the participle set by using a second model, determining a fifth vector set corresponding to the participle set by using a third model, and taking part or all of the first vector set, the second vector set, the third vector set, the fourth vector set and the fifth vector set as a third input set of the second sub-model.
The first model in this embodiment may be: the method comprises the steps of an ngram2vec model, inputting each token in the participle set to the ngram2vec model, extracting features of each token by the ngram2vec model from the dimension of a word to obtain token level embedding vectors corresponding to each token, extracting features of the tokens by the ngram2vec model from the dimension of a character to obtain char level embedding vectors corresponding to each token, taking all the token level embedding vectors as a first vector set, taking all the char level embedding vectors as a second vector set, taking the first vector set and the second vector set as first input sets, and taking all the token and the reservent tokens as second input sets.
Further, the second model may be: a FastText model, the third model may be a DSSM model; inputting all tokens in the word segmentation set into a FastText model, performing feature extraction on all tokens corresponding to a search text by the FastText model from the dimension of the text to obtain query-level embedding vectors corresponding to the search text, performing feature extraction on all tokens corresponding to the search text by the FastText model from the dimension of the text distribution to obtain query-channel distribution vectors corresponding to the search text, taking the query-level embedding vectors corresponding to the search text as a third vector set, and taking the query-channel distribution vectors corresponding to the search text as a fourth vector set; all tokens in the token set are input into a DSSM (direct sequence spread spectrum) model, the DSSM model extracts features of all tokens from the dimension of the token semantics to obtain DSSM token embedding vectors corresponding to all tokens, all DSSM token embedding vectors are used as a fifth vector set, part or all of the first vector set, the second vector set, the third vector set, the fourth vector set and the fifth vector set are used as a third input set of a second submodel, and the vectors of multiple dimensions are used as the input of the second submodel, so that the second submodel can respectively identify the search texts from part or all of the DSSR, token, query-level, query-channel and M token, the semantics of the search texts are enhanced in the identification process, and the accuracy of the search texts is improved.
For example, the third set of inputs may include: the first vector set, the second vector set, the third vector set, the fourth vector set, and the fifth vector set, the third input set may further include: the vector set specifically included in the third input set may be set according to actual requirements, which is not specifically limited in this embodiment.
Fig. 4 is a schematic flow chart of obtaining a third output result according to an embodiment of the present invention, and as shown in fig. 4, the method specifically includes:
and S41, combining at least two vector sets in the third input set into a multi-dimensional feature vector.
Further, the dimension of the feature vector may be determined according to the number of vector sets in the third input set, for example, the multidimensional feature vector may be a two-dimensional feature vector or a five-bit feature vector, and the dimension of the feature vector may be set according to actual requirements, which is not limited in this embodiment.
And S42, inputting the multi-dimensional feature vector into the second sub-model, so that the second sub-model outputs the third output result, wherein the third output result comprises the prediction probability value corresponding to the search text and the leaf node number features corresponding to the word segmentation set and the similar meaning word set.
In this embodiment, the second sub-model adopts an Xgboost model, and the number of trees in the Xgboost model may be set as: n _ estimators ═ 2000, the depth of the tree is set to: and (7) taking a token level embedding vector, a channel level embedding vector, a query-channel vector and a DSSM token embedding vector corresponding to each token in each search text as input of an Xgboost model, and outputting the prediction probability value of the search text, the participle set and the leaf node number characteristic corresponding to the similar meaning word set by the Xgboost model.
Fig. 5 is a schematic flowchart of training a first sub-model according to an embodiment of the present invention, and as shown in fig. 5, the method specifically includes:
s51, inputting the first input set to the neural network portion, so that the neural network portion extracts local features of the participle set from the first input set, fuses the local features to obtain global features of the participle set, and obtains the first output result of the search text by combining the global features with the probability value in the third output result.
In this embodiment, a first input set (token level embedding vector) is input to a neural network portion in a first sub-model, a local feature in a token sequence is extracted through a CNN (convolutional neural network) in the neural network portion, and then an Attention in the neural network portion is fused with the local feature of the token sequence to obtain a global feature.
Further, the predicted probability value output by the Xgboost model spliced by the CNN is used, and finally, a first output result of the search text is obtained through three full-connected layers, wherein the first output result is as follows: the predicted result of the neural network part of the text in the first submodel is searched.
And S52, inputting the second input set to the linear part, so that the linear part obtains the second output result of the search text according to the participle set and the synonym set and by combining the leaf node number characteristics in the third output result.
The linear part in the first submodel splices a second input set (token and recent token) with the Xgboost model to output the word segmentation set and the leaf node number characteristics corresponding to the word set to obtain a second output result of the search text, wherein the second output result is as follows: the prediction results of the linear part of the text in the first submodel are searched.
And S53, training the first sub-model according to each search text in the training sample.
And S54, determining a fourth output result of the search text recognition model according to the first output result and the second output result.
The fourth output result corresponding to the first submodel is: and the sum of the first output result and the second output result, and the fourth output result of the first submodel is also the fourth output result of the search text recognition model.
logit＝logitwide+logitdeep
Wherein logic is the fourth output result, logicdeepIs the first output result, logicwideIs the second output result.
S55, constructing the loss function according to the first output result, the second output result and the fourth output result.
S56, training the first sub-model based on the loss function in an auxiliary mode until the loss function meets a preset convergence condition, and determining that the training of the first sub-model is completed.
The loss function may be:
Loss＝CE(logitwide)+CE(logitdeep)+CE(logit)
where CE represents cross entropy.
The method for constructing the search text recognition model provided by the embodiment of the invention has the advantages that the semantics of the search text is not deficient any more through word segmentation and expansion processing of the search text, the semantic range of the search text is expanded, the generalization characteristic of the search text and the memory capacity of the search text are enhanced by combining the neural network part in the first sub-model, the recognition capacity of the linear part of the first sub-model is enhanced by the second sub-model, and the short and semantically deficient search text is accurately recognized on the whole.
Fig. 7 is a schematic flowchart of a method for identifying a search text according to an embodiment of the present invention, and as shown in fig. 7, the method specifically includes:
and S71, determining a participle set and a similar meaning word set corresponding to the search text to be recognized.
S72, carrying out multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set.
In this embodiment, the first input set includes two preset feature vector sets, and the feature vector set may be: the vector representation of the word dimension corresponding to each participle in the participle set and the vector representation of the character dimension corresponding to each participle in the participle set.
The third set of inputs may include, in addition to the first set of inputs: the vector representation of the text dimension corresponding to the search text, the vector representation of the text distribution dimension corresponding to the search text, and the vector representation of the participle semantics corresponding to each participle in the participle set are all or partially (where, some parts may be zero).
Further, for the first input set and the third input set similar to the first input set and the third input set in fig. 3, the related description of fig. 3 may be specifically referred to, and is not repeated herein.
And S73, taking the participle set and the similar meaning word set as a second input set.
And S74, inputting the third input set into a second submodel to obtain a third output result.
And S75, inputting the first input set, the second input set and the third output result into the first submodel, so that the first submodel outputs the identification result of the search text.
In this embodiment, the search text recognition model constructed in fig. 1 to 5 is used to perform recognition of the search text, and the process of the search text recognition is partially similar to the construction process of the search text recognition model.
In an alternative of the embodiment of the present invention, S71 may include the following sub-steps:
s711 (not shown in the figure), performing word segmentation processing on the search text to obtain a word segmentation set including a plurality of words and/or phrases.
S712 (not shown in the figure), performing semantic matching on each word and/or phrase in the segmented word set, and using one or more similar words with a semantic similarity greater than a set threshold and/or one or more similar phrases with a semantic similarity greater than a set threshold as a similar word set corresponding to the segmented word set.
In an alternative of the embodiment of the present invention, S75 may include the following sub-steps:
s751 (not shown in the figure), inputting the first input set and the third output result to the neural network portion, so that the neural network portion outputs the first output result corresponding to the search text.
S752 (not shown in the figure), and inputting the second input set and the third output result into the linear part, so that the linear part outputs the second output result corresponding to the search text.
S753 (not shown), and setting the first output result and the second output result as the recognition result of the search text.
S71(S711-S712) is similar to S21-S22 in fig. 2, S72-S73 is similar to S12 in fig. 1, S31-S33 in fig. 3, S74 is similar to S41-S42 in fig. 4, and S75(S751-S755) is similar to S51-S54 in fig. 5, which can specifically refer to the above description of fig. 2-5 and is not repeated herein.
According to the identification scheme of the search text provided by the embodiment of the invention, the segmentation and word expansion processing are carried out on the search text, so that the semantic meaning of the search text is not deficient, the semantic scope of the search text is expanded, multi-dimensional feature extraction is carried out on a segmentation set and a similar word set, the model carries out classification and identification on the extracted feature vectors, the identification result of the search text is obtained, and the short and semantic deficient search text is accurately identified on the whole.
Fig. 8 is a schematic structural diagram of a device for recognizing a search text according to an embodiment of the present invention, and as shown in fig. 8, the device is applied to training a search text recognition model, where the search text recognition model includes a first sub-model and a second sub-model, and specifically includes:
a set determining module 81, configured to determine a participle set and a synonym set corresponding to a search text to be identified;
an extraction module 82, configured to perform multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, where the first input set includes two preset feature vector sets, and the third input set at least includes the first input set;
the set determining module 81 is further configured to use the set of segmented words and the set of similar words as a second input set;
the output determining module 83 is configured to input the third input set into a second submodel to obtain a third output result;
a result determination module 84, configured to input the first input set, the second input set, and the third output result into the first submodel, so that the first submodel outputs the recognition result of the search text.
The device for identifying a search text provided in this embodiment may be the device for identifying a search text shown in fig. 8, and may perform all the steps of the method for identifying a search text shown in fig. 7, so as to achieve the technical effect of the method for identifying a search text shown in fig. 7.
Fig. 9 is a schematic structural diagram of a computer device according to an embodiment of the present invention, where the computer device 900 shown in fig. 9 includes: at least one processor 901, memory 902, at least one network interface 904, and other user interfaces 903. The various components in computer device 900 are coupled together by a bus system 905. It is understood that the bus system 905 is used to enable communications among the components. The bus system 905 includes a power bus, a control bus, and a status signal bus, in addition to a data bus. For clarity of illustration, however, the various buses are labeled in fig. 9 as bus system 905.
The user interface 903 may include, among other things, a display, a keyboard, or a pointing device (e.g., a mouse, trackball, touch pad, or touch screen, among others.
It is to be understood that the memory 902 in embodiments of the present invention may be either volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory. The non-volatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable PROM (EEPROM), or a flash Memory. Volatile Memory can be Random Access Memory (RAM), which acts as external cache Memory. By way of illustration and not limitation, many forms of RAM are available, such as Static random access memory (Static RAM, SRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic random access memory (Synchronous DRAM, SDRAM), Double Data Rate Synchronous Dynamic random access memory (ddr Data Rate SDRAM, ddr SDRAM), Enhanced Synchronous SDRAM (ESDRAM), synchlronous SDRAM (SLDRAM), and Direct Rambus RAM (DRRAM). The memory 902 described herein is intended to comprise, without being limited to, these and any other suitable types of memory.
In some embodiments, memory 902 stores the following elements, executable units or data structures, or a subset thereof, or an expanded set thereof: an operating system 9021 and application programs 9022.
The operating system 9021 includes various system programs, such as a framework layer, a core library layer, a driver layer, and the like, and is configured to implement various basic services and process hardware-based tasks. The application 9022 includes various applications, such as a Media Player (Media Player), a Browser (Browser), and the like, for implementing various application services. A program implementing the method of an embodiment of the present invention may be included in application 9022.
In the embodiment of the present invention, by calling a program or an instruction stored in the memory 902, specifically, a program or an instruction stored in the application 9022, the processor 901 is configured to execute the method steps provided by the method embodiments, for example, including:
determining a participle set and a similar meaning word set corresponding to a search text to be identified; performing multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set; taking the word segmentation set and the similar meaning word set as a second input set; inputting the third input set into a second submodel to obtain a third output result; inputting the first input set, the second input set and the third output result into the first submodel to cause the first submodel to output the recognition result of the search text.
In a possible implementation manner, performing word segmentation processing on the search text to obtain a word segmentation set containing a plurality of words and/or phrases; and performing semantic matching on each word and/or phrase in the word segmentation set, and taking one or more similar words with the similarity greater than a set threshold value with the word meaning and/or one or more similar phrases with the similarity greater than the set threshold value with the phrase meaning as a similar word set corresponding to the word segmentation set.
In one possible embodiment, the first set of inputs includes: a first vector set and a second vector set; the third set of inputs includes at least two of: a first set of vectors, a second set of vectors, a third set of vectors, a fourth set of vectors, or a fifth set of vectors; the first set of vectors comprises: vector representation of a word dimension corresponding to each participle in the participle set; the second set of vectors includes: vector representation of character dimension corresponding to each participle in the participle set; the third set of vectors comprises: vector representation of a text dimension corresponding to the search text; the fourth set of vectors comprises: vector representation of a text distribution dimension corresponding to the search text; the fifth set of vectors comprises: each participle in the participle set corresponds to vector representation of participle semantics.
In one possible embodiment, the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
inputting the first input set and the third output result into the neural network portion to cause the neural network portion to output the first output result corresponding to the search text; inputting the second input set and the third output result into the linear part to enable the linear part to output the second output result corresponding to the search text; and taking the first output result and the second output result as the recognition result of the search text.
In one possible embodiment, the second submodel is: a third output result output by the Xgboost model comprises a probability value corresponding to the search text and leaf node number characteristics corresponding to the word segmentation set and the similar word set;
inputting the first input set into the neural network part, so that the neural network part extracts local features of the word segmentation set from the first input set, fuses the local features to obtain global features of the word segmentation set, and obtains the first output result of the search text by combining the global features with the probability value in the third output result.
In one possible implementation, the second input set is input to the linear part, so that the linear part obtains the second output result of the search text according to the participle set and the synonym set and by combining the leaf node number characteristics in the third output result.
In one possible implementation manner, for any search text in a training sample, performing word segmentation on the search text to obtain a word segmentation set corresponding to the search text, and performing word expansion processing on each word segmentation in the word segmentation set to obtain a near-meaning word set corresponding to the word segmentation set; determining a first input set of the first submodel according to the word segmentation set, determining a second input set of the first submodel according to the word segmentation set and the word similarity set, and determining a third input set of the second submodel according to the word segmentation set; inputting the third input set into the second submodel to enable the second submodel to output a corresponding third output result; and training the first sub-model by using the first input set, the second input set and the third output result, and determining that the training of the search text recognition model is finished when the training of the first sub-model is finished.
In one possible embodiment, the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
inputting the first input set and the third output result into the neural network part, and training the neural network part; and inputting the second input set and the third output result into the linear part, and training the linear part.
In one possible embodiment, a loss function is constructed from a first output of the neural network portion and a second output of the linear portion; using the loss function to assist the first sub-model to train;
wherein, the condition that the training of the first sub-model is completed comprises: the loss function satisfies a preset convergence condition.
The method disclosed in the above embodiments of the present invention may be applied to the processor 901, or implemented by the processor 901. The processor 901 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be implemented by integrated logic circuits of hardware or instructions in the form of software in the processor 901. The Processor 901 may be a general-purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component. The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software elements in the decoding processor. The software elements may be located in ram, flash, rom, prom, or eprom, registers, among other storage media that are well known in the art. 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 the hardware thereof.
It is to be understood that the embodiments described herein may be implemented in hardware, software, firmware, middleware, microcode, or any combination thereof. For a hardware implementation, the Processing units may be implemented within one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), general purpose processors, controllers, micro-controllers, microprocessors, other electronic units configured to perform the functions described herein, or a combination thereof.
For a software implementation, the techniques described herein may be implemented by means of units performing the functions described herein. The software codes may be stored in a memory and executed by a processor. The memory may be implemented within the processor or external to the processor.
The computer device provided in this embodiment may be the computer device shown in fig. 9, and may perform all the steps of the identification method for searching for a text shown in fig. 7, so as to achieve the technical effect of the identification method for searching for a text shown in fig. 7.
The embodiment of the invention also provides a storage medium (computer readable storage medium). The storage medium herein stores one or more programs. Among others, the storage medium may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, a hard disk, or a solid state disk; the memory may also comprise a combination of memories of the kind described above.
When one or more programs in the storage medium are executable by one or more processors to implement the above-described method of recognizing a search text performed on the side of a device for recognizing a search text.
The processor is used for executing the identification program of the search text stored in the memory so as to realize the following steps of the identification method of the search text executed on the identification device side of the search text:
determining a participle set and a similar meaning word set corresponding to a search text to be identified; performing multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set; taking the word segmentation set and the similar meaning word set as a second input set; inputting the third input set into a second submodel to obtain a third output result; inputting the first input set, the second input set and the third output result into the first submodel to cause the first submodel to output the recognition result of the search text.
In a possible implementation manner, performing word segmentation processing on the search text to obtain a word segmentation set containing a plurality of words and/or phrases; and performing semantic matching on each word and/or phrase in the word segmentation set, and taking one or more similar words with the similarity greater than a set threshold value with the word meaning and/or one or more similar phrases with the similarity greater than the set threshold value with the phrase meaning as a similar word set corresponding to the word segmentation set.
In one possible embodiment, the first set of inputs includes: a first vector set and a second vector set; the third set of inputs includes at least two of: a first set of vectors, a second set of vectors, a third set of vectors, a fourth set of vectors, or a fifth set of vectors; the first set of vectors comprises: vector representation of a word dimension corresponding to each participle in the participle set; the second set of vectors includes: vector representation of character dimension corresponding to each participle in the participle set; the third set of vectors comprises: vector representation of a text dimension corresponding to the search text; the fourth set of vectors comprises: vector representation of a text distribution dimension corresponding to the search text; the fifth set of vectors comprises: each participle in the participle set corresponds to vector representation of participle semantics.
In one possible embodiment, the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
inputting the first input set and the third output result into the neural network portion to cause the neural network portion to output the first output result corresponding to the search text; inputting the second input set and the third output result into the linear part to enable the linear part to output the second output result corresponding to the search text; and taking the first output result and the second output result as the recognition result of the search text.
In one possible embodiment, the second submodel is: a third output result output by the Xgboost model comprises a probability value corresponding to the search text and leaf node number characteristics corresponding to the word segmentation set and the similar word set;
inputting the first input set into the neural network part, so that the neural network part extracts local features of the word segmentation set from the first input set, fuses the local features to obtain global features of the word segmentation set, and obtains the first output result of the search text by combining the global features with the probability value in the third output result.
In one possible implementation, the second input set is input to the linear part, so that the linear part obtains the second output result of the search text according to the participle set and the synonym set and by combining the leaf node number characteristics in the third output result.
In one possible implementation manner, for any search text in a training sample, performing word segmentation on the search text to obtain a word segmentation set corresponding to the search text, and performing word expansion processing on each word segmentation in the word segmentation set to obtain a near-meaning word set corresponding to the word segmentation set; determining a first input set of the first submodel according to the word segmentation set, determining a second input set of the first submodel according to the word segmentation set and the word similarity set, and determining a third input set of the second submodel according to the word segmentation set; inputting the third input set into the second submodel to enable the second submodel to output a corresponding third output result; and training the first sub-model by using the first input set, the second input set and the third output result, and determining that the training of the search text recognition model is finished when the training of the first sub-model is finished.
In one possible embodiment, the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
inputting the first input set and the third output result into the neural network part, and training the neural network part; and inputting the second input set and the third output result into the linear part, and training the linear part.
In one possible embodiment, a loss function is constructed from a first output of the neural network portion and a second output of the linear portion; using the loss function to assist the first sub-model to train;
wherein, the condition that the training of the first sub-model is completed comprises: the loss function satisfies a preset convergence condition.
Those of skill would further appreciate that the various illustrative components and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied in hardware, a software module executed by a processor, or a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (12)

1. A method of recognition of a search text, the search text recognition model comprising a first sub-model and a second sub-model, the method comprising:
determining a participle set and a similar meaning word set corresponding to a search text to be identified;
performing multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set;
taking the word segmentation set and the similar meaning word set as a second input set;
inputting the third input set into a second submodel to obtain a third output result;
inputting the first input set, the second input set and the third output result into the first submodel to cause the first submodel to output the recognition result of the search text.
2. The method of claim 1, wherein the determining a set of participles and a set of synonyms corresponding to the search text to be recognized comprises:
performing word segmentation processing on the search text to obtain a word segmentation set containing a plurality of words and/or phrases;
and performing semantic matching on each word and/or phrase in the word segmentation set, and taking one or more similar words with the similarity greater than a set threshold value with the word meaning and/or one or more similar phrases with the similarity greater than the set threshold value with the phrase meaning as a similar word set corresponding to the word segmentation set.
3. The method of claim 1, wherein the first set of inputs comprises: a first vector set and a second vector set;
the third set of inputs includes at least two of: a first set of vectors, a second set of vectors, a third set of vectors, a fourth set of vectors, or a fifth set of vectors;
the first set of vectors comprises: vector representation of a word dimension corresponding to each participle in the participle set;
the second set of vectors includes: vector representation of character dimension corresponding to each participle in the participle set;
the third set of vectors comprises: vector representation of a text dimension corresponding to the search text;
the fourth set of vectors comprises: vector representation of a text distribution dimension corresponding to the search text;
the fifth set of vectors comprises: each participle in the participle set corresponds to vector representation of participle semantics.
4. The method of claim 1, wherein the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
the inputting the first input set, the second input set, and the third output result into the first submodel to cause the first submodel to output the recognition result of the search text includes:
inputting the first input set and the third output result into the neural network portion to cause the neural network portion to output the first output result corresponding to the search text;
inputting the second input set and the third output result into the linear part to enable the linear part to output the second output result corresponding to the search text;
and taking the first output result and the second output result as the recognition result of the search text.
5. The method of claim 4, wherein the second submodel is: a third output result output by the Xgboost model comprises a probability value corresponding to the search text and leaf node number characteristics corresponding to the word segmentation set and the similar word set;
the inputting the first input set and the third output result to the neural network portion to cause the neural network portion to output the first output result corresponding to the search text includes:
inputting the first input set into the neural network part, so that the neural network part extracts local features of the word segmentation set from the first input set, fuses the local features to obtain global features of the word segmentation set, and obtains the first output result of the search text by combining the global features with the probability value in the third output result.
6. The method of claim 5, wherein the inputting the second input set and the third output result into the linear portion to cause the linear portion to output the second output result corresponding to the search text comprises:
and inputting the second input set into the linear part, so that the linear part obtains the second output result of the search text according to the participle set and the similar meaning word set and by combining the leaf node number characteristics in the third output result.
7. The method of claim 1, wherein the search text recognition model is constructed by:
performing word segmentation on the search text aiming at any search text in a training sample to obtain a word segmentation set corresponding to the search text, and performing word expansion processing on each word segmentation in the word segmentation set to obtain a near meaning word set corresponding to the word segmentation set;
determining a first input set of the first submodel according to the word segmentation set, determining a second input set of the first submodel according to the word segmentation set and the word similarity set, and determining a third input set of the second submodel according to the word segmentation set;
inputting the third input set into the second submodel to enable the second submodel to output a corresponding third output result;
and training the first sub-model by using the first input set, the second input set and the third output result, and determining that the training of the search text recognition model is finished when the training of the first sub-model is finished.
8. The method of claim 7, wherein the first submodel is: a Wide & Deep model comprising a neural network portion and a linear portion;
the training the first submodel using the first set of inputs, the second set of inputs, and the third output results includes:
inputting the first input set and the third output result into the neural network part, and training the neural network part;
and inputting the second input set and the third output result into the linear part, and training the linear part.
9. The method of claim 8, further comprising:
constructing a loss function from a first output result of the neural network portion and a second output result of the linear portion;
using the loss function to assist the first sub-model to train;
wherein, the condition that the training of the first sub-model is completed comprises: the loss function satisfies a preset convergence condition.
10. An apparatus for recognizing search text, wherein the search text recognition model includes a first submodel and a second submodel, the apparatus comprising:
the set determining module is used for determining a participle set and a similar meaning word set corresponding to the search text to be identified;
the extraction module is used for carrying out multi-dimensional feature extraction on the word segmentation set to obtain a first input set and a third input set, wherein the first input set comprises two preset feature vector sets, and the third input set at least comprises the first input set;
the set determination module is further configured to use the participle set and the synonym set as a second input set;
the output determining module is used for inputting the third input set into a second submodel to obtain a third output result;
a result determination module, configured to input the first input set, the second input set, and the third output result into the first submodel, so that the first submodel outputs the recognition result of the search text.
11. A computer device, comprising: a processor and a memory, wherein the processor is used for executing a construction program of a search text recognition model stored in the memory so as to realize the recognition method of the search text of any one of claims 1-9.
12. A storage medium storing one or more programs executable by one or more processors to implement the method of identifying search text according to any one of claims 1 to 9.
CN202110605909.XA 2021-05-31 2021-05-31 Search text recognition method and device, computer equipment and storage medium Active CN113254587B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110605909.XA CN113254587B (en) 2021-05-31 2021-05-31 Search text recognition method and device, computer equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110605909.XA CN113254587B (en) 2021-05-31 2021-05-31 Search text recognition method and device, computer equipment and storage medium

Publications (2)

Publication Number Publication Date
CN113254587A true CN113254587A (en) 2021-08-13
CN113254587B CN113254587B (en) 2023-10-13

Family

ID=77185570

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110605909.XA Active CN113254587B (en) 2021-05-31 2021-05-31 Search text recognition method and device, computer equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113254587B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080005091A1 (en) * 2006-06-28 2008-01-03 Microsoft Corporation Visual and multi-dimensional search
US20190065507A1 (en) * 2017-08-22 2019-02-28 Beijing Baidu Netcom Science And Technology Co., Ltd. Method and apparatus for information processing
CN110969023A (en) * 2018-09-29 2020-04-07 北京国双科技有限公司 Text similarity determination method and device
WO2020238053A1 (en) * 2019-05-31 2020-12-03 平安科技（深圳）有限公司 Neural grid model-based text data category recognition method and apparatus, nonvolatile readable storage medium, and computer device
CN112035599A (en) * 2020-11-06 2020-12-04 苏宁金融科技(南京)有限公司 Query method and device based on vertical search, computer equipment and storage medium
CN112287072A (en) * 2020-11-20 2021-01-29 公安部第一研究所 Multi-dimensional Internet text risk data identification method
CN112347791A (en) * 2020-11-06 2021-02-09 北京奇艺世纪科技有限公司 Method and system for constructing text matching model, computer equipment and storage medium
US20210081729A1 (en) * 2019-09-16 2021-03-18 Beijing Baidu Netcom Science Technology Co., Ltd. Method for image text recognition, apparatus, device and storage medium
CN112579767A (en) * 2019-09-29 2021-03-30 北京搜狗科技发展有限公司 Search processing method and device for search processing
CN112800758A (en) * 2021-04-08 2021-05-14 明品云(北京)数据科技有限公司 Method, system, equipment and medium for distinguishing similar meaning words in text

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080005091A1 (en) * 2006-06-28 2008-01-03 Microsoft Corporation Visual and multi-dimensional search
US20190065507A1 (en) * 2017-08-22 2019-02-28 Beijing Baidu Netcom Science And Technology Co., Ltd. Method and apparatus for information processing
CN110969023A (en) * 2018-09-29 2020-04-07 北京国双科技有限公司 Text similarity determination method and device
WO2020238053A1 (en) * 2019-05-31 2020-12-03 平安科技（深圳）有限公司 Neural grid model-based text data category recognition method and apparatus, nonvolatile readable storage medium, and computer device
US20210081729A1 (en) * 2019-09-16 2021-03-18 Beijing Baidu Netcom Science Technology Co., Ltd. Method for image text recognition, apparatus, device and storage medium
CN112579767A (en) * 2019-09-29 2021-03-30 北京搜狗科技发展有限公司 Search processing method and device for search processing
CN112035599A (en) * 2020-11-06 2020-12-04 苏宁金融科技(南京)有限公司 Query method and device based on vertical search, computer equipment and storage medium
CN112347791A (en) * 2020-11-06 2021-02-09 北京奇艺世纪科技有限公司 Method and system for constructing text matching model, computer equipment and storage medium
CN112287072A (en) * 2020-11-20 2021-01-29 公安部第一研究所 Multi-dimensional Internet text risk data identification method
CN112800758A (en) * 2021-04-08 2021-05-14 明品云(北京)数据科技有限公司 Method, system, equipment and medium for distinguishing similar meaning words in text

Also Published As

Publication number Publication date
CN113254587B (en) 2023-10-13

Similar Documents

Publication Publication Date Title
CN108711422B (en) Speech recognition method, speech recognition device, computer-readable storage medium and computer equipment
CN107085581B (en) Short text classification method and device
CN104615589A (en) Named-entity recognition model training method and named-entity recognition method and device
CN110717332B (en) News and case similarity calculation method based on asymmetric twin network
CN112528637B (en) Text processing model training method, device, computer equipment and storage medium
CN106528694B (en) semantic judgment processing method and device based on artificial intelligence
WO2021134524A1 (en) Data processing method, apparatus, electronic device, and storage medium
CN111291177A (en) Information processing method and device and computer storage medium
CN113948066B (en) Error correction method, system, storage medium and device for real-time translation text
CN110309504B (en) Text processing method, device, equipment and storage medium based on word segmentation
JP2020004382A (en) Method and device for voice interaction
CN110781687B (en) Same intention statement acquisition method and device
CN111737997A (en) Text similarity determination method, text similarity determination equipment and storage medium
JP2023022845A (en) Method of processing video, method of querying video, method of training model, device, electronic apparatus, storage medium and computer program
CN111985228A (en) Text keyword extraction method and device, computer equipment and storage medium
CN113536795B (en) Method, system, electronic device and storage medium for entity relation extraction
CN111814477B (en) Dispute focus discovery method and device based on dispute focus entity and terminal
CN112766319A (en) Dialogue intention recognition model training method and device, computer equipment and medium
CN116304748A (en) Text similarity calculation method, system, equipment and medium
TWI752406B (en) Speech recognition method, speech recognition device, electronic equipment, computer-readable storage medium and computer program product
CN115544303A (en) Method, apparatus, device and medium for determining label of video
CN113553848A (en) Long text classification method, system, electronic equipment and computer readable storage medium
CN114298055B (en) Retrieval method and device based on multilevel semantic matching, computer equipment and storage medium
CN115374325A (en) Website classification method and device, classification equipment and storage medium
CN113672731B (en) Emotion analysis method, device, equipment and storage medium based on field information

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant