Confidence coefficient prediction method and device based on emotional behaviors, equipment and medium
By extracting and analyzing deterministic and integral texts in emotional behavior texts, combining behavior weights and emotional impact values, we automatically predict the confidence of emotional risk among emotional objects, and solving the stability and efficiency of emotional risk assessment in the prior art. The problem is to achieve efficient evaluation of multiple emotional objects.
Patent Information
- Application Number
- CN202411073332.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-11-22
AI Technical Summary
In the prior art, emotional risk assessment relies on manual consultation, making it difficult to evaluate multiple emotional objects at the same time, and the stability and effectiveness of the evaluation results are poor.
By extracting deterministic and integral texts in emotional behavior texts, combining behavior weights and emotional impact values, we automatically predict emotional risk confidence among emotional objects.
It realizes automated emotional risk prediction, reduces human resource consumption, improves the stability and effectiveness of assessment, and supports simultaneous evaluation of multiple emotional objects.
Smart Images

Figure CN119005201A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of computer technology, and in particular to a confidence prediction method, device, equipment and medium based on emotional behavior. Background Art
[0002] In order to reduce the possibility of emotional risk between emotional objects, emotional risk assessment can be carried out and timely intervention can be carried out. In the related art, emotional risk assessment is carried out by professionals in the form of manual consultation. However, this method consumes a lot of manpower, it is difficult to conduct risk assessment on multiple pairs of emotional objects at the same time, and the final assessment result depends on the knowledge level of professionals, and the stability and effectiveness of the assessment fluctuate greatly. Summary of the invention
[0003] In order to solve the above technical problems or at least partially solve the above technical problems, the present disclosure provides a confidence prediction method, device, equipment and medium based on emotional behavior.
[0004] The present disclosure provides a confidence prediction method based on emotional behavior, including:
[0005] Extracting a first text from the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior;
[0006] If the first text is not extracted, extracting a second text from the emotional behavior text, and determining an emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral emotional behavior;
[0007] Emotional trend prediction is performed based on the behavior weight corresponding to the integral-type emotional behavior and the emotional impact value to obtain the emotional risk confidence between the corresponding emotional objects.
[0008] The present disclosure also provides a confidence prediction device based on emotional behavior, including:
[0009] A first extraction module is used to extract a first text from the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior;
[0010] A second extraction module is used to extract a second text from the emotional behavior text if the first text is not extracted, and determine the emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral emotional behavior;
[0011] The first prediction module is used to predict the emotional trend according to the behavior weight corresponding to the integral emotional behavior and the emotional influence value, and obtain the emotional risk confidence between the corresponding emotional objects.
[0012] An embodiment of the present disclosure also provides an electronic device, which includes: a processor; a memory for storing executable instructions of the processor; the processor is used to read the executable instructions from the memory and execute the instructions to implement the confidence prediction method based on emotional behavior as provided in the embodiment of the present disclosure.
[0013] The embodiments of the present disclosure also provide a computer-readable storage medium, which stores a computer program, and the computer program is used to execute the confidence prediction method based on emotional behavior as provided in the embodiments of the present disclosure.
[0014] The technical solution provided by the embodiment of the present disclosure has the following advantages over the prior art: the confidence prediction scheme based on emotional behavior provided in the embodiment of the present disclosure comprises: extracting the first text in the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior; if the first text is not extracted, extracting the second text in the emotional behavior text, and determining the emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral-type emotional behavior; predicting the emotional trend according to the behavior weight and emotional impact value corresponding to the integral-type emotional behavior, and obtaining the emotional risk confidence between the corresponding emotional objects. By adopting the above technical solution, when the text corresponding to the judgment-type emotional behavior is not extracted from the emotional behavior text, the text corresponding to the integral-type emotional behavior is extracted, and the emotional risk confidence is predicted in combination with the behavior weight and emotional impact value of the integral-type emotional behavior, so as to realize the prediction of the result of emotional behavior according to the integral-type emotional behavior when there is no judgment-type emotional behavior between the emotional objects, realize the automatic prediction of the confidence corresponding to the emotional risk according to the emotional behavior, reduce the human resources consumed by the emotional risk prediction, and create a basis for realizing the risk assessment of multiple pairs of emotional objects at the same time, and the confidence result generated is stable and effective. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.
[0016] In order to more clearly illustrate the embodiments of the present invention 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.
[0017] Figure 1 A flowchart of a confidence prediction method based on emotional behavior provided in an embodiment of the present disclosure;
[0018] Figure 2 A flowchart of another confidence prediction method based on emotional behavior provided in an embodiment of the present disclosure;
[0019] Figure 3 A schematic diagram of the structure of a confidence prediction device based on emotional behavior provided by an embodiment of the present disclosure;
[0020] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0021] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.
[0022] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.
[0023] In order to solve the above problems, the embodiments of the present disclosure provide a confidence prediction method based on emotional behavior, which is introduced below in conjunction with specific embodiments.
[0024] Figure 1 The present invention provides a flow chart of a confidence prediction method based on emotional behavior. The confidence prediction method based on emotional behavior can be applied to a confidence prediction device based on emotional behavior. The confidence prediction device based on emotional behavior can be implemented by software and / or hardware. The confidence prediction device based on emotional behavior can generally be integrated into an electronic device. Figure 1 As shown, the confidence prediction method based on emotional behavior includes:
[0025] Step 101, extracting the first text in the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior.
[0026] Among them, the emotional behavior text can be a text that records the emotional object and emotion-related behavior, the emotional object can be an object with an emotional relationship, and the emotional objects can be paired. The emotional behavior text can include structured and / or unstructured text. Structured emotional behavior texts can include questionnaires, tests, etc. of emotional objects, and unstructured emotional behavior texts can include texts that record the communication between emotional objects and / or texts that record the behavior of emotional objects. The first text can be a text that records specific judgment-type emotional behaviors that occur between emotional objects. Judgment-type emotional behaviors can be emotional behavior indicators that have a decisive influence on emotional risk confidence. The judgment-type emotional behavior can be understood as the type of emotional behavior, and the first text can be understood as a specific label under the judgment-type emotional behavior. This embodiment does not limit the judgment-type emotional behavior.
[0027] In some embodiments of the present disclosure, extracting a first text from an emotional behavior text includes: extracting text of sensitive behaviors from the emotional behavior text through a text extraction model to obtain the first text.
[0028] The text extraction model may be a trained language model with text extraction function. This embodiment does not limit the type of the text extraction model. Sensitive behavior can be understood as an emotional behavior that has a veto effect on the emotional results between emotional objects. The sensitive behavior can be restricted according to user needs, etc., and this embodiment does not limit it.
[0029] In this embodiment, the confidence prediction device based on emotional behavior can input the emotional behavior text into the text extraction model, perform word segmentation processing on the emotional behavior text through the text extraction model to obtain a word text, and perform semantic similarity matching between the word text and the preset sample sensitive words, determine the word text that successfully matches as the first text, and the text extraction model outputs the first text. Thus, the automatic extraction of sensitive behavior related text is realized based on the text extraction model.
[0030] Step 102: If the first text is not extracted, extract the second text in the emotional behavior text, and determine the emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral emotional behavior.
[0031] The second text may be a text recording the specific integral emotional behavior between emotional objects, and the second text may be understood as a specific label under the integral emotional behavior. This embodiment does not limit the data format of the second text. For example, the format of the second text may include providing at least one of a string format, an enumeration format, a value format, and a date format.
[0032] Integral-type emotional behavior may be an emotional behavior that has a non-determinative impact on the emotional risk confidence. The integral-type emotional behavior has a smaller impact on the emotional risk confidence than the judgment-type emotional behavior. The integral-type emotional behavior can be understood as a type of emotional behavior. This embodiment does not limit the integral-type emotional behavior. The integral-type emotional behavior may include emotional object attributes and non-sensitive behaviors. Emotional object attributes may be properties, characteristics, etc. possessed by emotional objects. Emotional object attributes may include one or more of the health status, education status, and work status of the emotional object. It can be understood that the emotional object attributes characterize the subject that performs the specific emotional behavior, and the specific attributes of the subject will also affect the final emotional risk confidence. Non-sensitive behavior may be an emotional behavior that has a cumulative comprehensive influence on the emotional results between emotional objects. The non-sensitive behavior may be limited according to user needs, etc., and this embodiment does not limit it.
[0033] In this embodiment, if the first text extracted from the emotional behavior text is empty, that is, the first text is not extracted from the emotional behavior text, the second text representing the integral-type emotional behavior is extracted from the emotional behavior text.
[0034] In some embodiments of the present disclosure, extracting the second text from the emotional behavior text includes: extracting the text of the emotional object attributes and the non-sensitive behavior from the emotional behavior text through a text extraction model to obtain the second text. The second text may include text representing the emotional object attributes and text representing the non-sensitive behavior.
[0035] In this embodiment, after the emotional behavior text is input into the text extraction model, the confidence prediction device based on emotional behavior can input the prompt words representing the emotional object attributes and the extraction of non-sensitive behavior text into the text extraction model, and the text extraction model outputs the second text extracted from the emotional behavior text. Thus, the extraction of emotional object attributes and non-sensitive behavior related text is realized automatically based on the text extraction model.
[0036] In some embodiments of the present disclosure, determining the sentiment impact value corresponding to the second text includes: performing query processing on a preset text impact value relationship according to the second text to obtain the sentiment impact value.
[0037] The text influence value relationship may be a corresponding relationship between the text and the emotion influence value. The emotion influence value may represent the influence of a specific behavior under the corresponding integral-type emotion behavior on the emotion risk confidence.
[0038] In this embodiment, the confidence prediction device based on emotional behavior can query the text influence value relationship according to the second text, and determine the emotional influence value corresponding to the second text in the text influence value relationship. For example, in the text influence value relationship, the same integral emotional behavior can correspond to multiple second texts, and different second texts record different degrees of the integral emotional behavior, and each second text has a different emotional influence value. After determining the current second text, determine the second text that is the same as the current second text in the text influence value relationship, and use the emotional influence value of the second text as the emotional influence value corresponding to the current second text.
[0039] In some embodiments of the present disclosure, the confidence prediction method based on emotional behavior also includes: if the first text is extracted, setting the emotional risk confidence to the highest risk value.
[0040] The emotional risk confidence level may be a credibility level that characterizes the emotional risk between emotional objects. The highest risk value may be a value that characterizes the greatest possibility of emotional risk between emotional objects. This embodiment does not limit the highest risk value.
[0041] In this embodiment, if a specific first text is extracted from the emotional behavior text, indicating that sensitive behavior has occurred between emotional objects, the emotional risk confidence is set to the highest risk value. Thus, in the case of sensitive behavior between emotional objects, the maximum intensity warning of emotional risk is achieved.
[0042] Optionally, the confidence prediction method based on emotional behavior may further include: extracting emotional object identity information and contact information from the emotional behavior text.
[0043] Step 103, predicting the emotional trend according to the behavior weight and emotional influence value corresponding to the integral emotional behavior, and obtaining the emotional risk confidence between the corresponding emotional objects.
[0044] Among them, the behavior weight can represent the importance of the integral emotional behavior.
[0045] In the disclosed embodiment, for each integral-type emotional behavior, the confidence prediction device based on the emotional behavior obtains a predetermined behavior weight for the integral-type emotional behavior, and obtains an emotional impact value corresponding to the integral-type emotional behavior determined based on the second text, and multiplies the behavior weight and the emotional impact value to obtain a weighted impact value corresponding to the integral-type emotional behavior. The weighted impact values corresponding to each integral-type emotional behavior are added together to obtain an emotional risk prediction value between emotional objects. Among the pre-set multiple prediction value intervals, the prediction value interval where the emotional risk prediction value is located is determined as a target prediction value interval, and the confidence corresponding to the target prediction value interval is determined as the emotional risk confidence between emotional objects.
[0046] The confidence prediction method based on emotional behavior provided by the embodiment of the present disclosure includes: extracting the first text in the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior; if the first text is not extracted, extracting the second text in the emotional behavior text, and determining the emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral-type emotional behavior; predicting the emotional trend according to the behavior weight and emotional impact value corresponding to the integral-type emotional behavior, and obtaining the emotional risk confidence between the corresponding emotional objects. By adopting the above technical scheme, when the text corresponding to the judgment-type emotional behavior is not extracted in the emotional behavior text, the text corresponding to the integral-type emotional behavior is extracted, and the emotional risk confidence is predicted in combination with the behavior weight and emotional impact value of the integral-type emotional behavior, so as to realize the prediction of the result of emotional behavior according to the integral-type emotional behavior when there is no judgment-type emotional behavior between the emotional objects, realize the automatic prediction of the confidence corresponding to the emotional risk according to the emotional behavior, reduce the human resources consumed by the emotional risk prediction, and create a basis for realizing the risk assessment of multiple pairs of emotional objects at the same time, and the confidence result generated is stable and effective.
[0047] In some embodiments of the present disclosure, the confidence prediction method based on emotional behavior also includes: obtaining sample emotional text, extracting keywords from the sample emotional text through a text extraction model to obtain emotional keywords; and determining judgment-type emotional behavior and integral-type emotional behavior according to the behavior type to which the emotional keywords belong.
[0048] The sample emotional text may be an emotional text used as sample data. This embodiment does not limit the sample emotional text. For example, the sample emotional text may include one or more of a text recording emotional disputes, a text recording emotional events, a text recording emotional rights protection, a text recording emotional mediation, and a text recording emotional public opinion. Thus, the sample emotional text is comprehensively acquired, so that the data for determining the integral emotional behavior is extensive and diverse, and the final determined emotional risk confidence is more comprehensive and accurate.
[0049] In this embodiment, before extracting the first text in the emotional behavior text, the confidence prediction device based on emotional behavior can obtain a sample emotional text, and extract keywords of the emotional dimension from the sample emotional text through the trained text extraction model to obtain the emotional keywords in the sample emotional text. Furthermore, the confidence prediction device based on emotional behavior can perform semantic analysis on the specific emotional keyword, determine the behavior type to which the emotional keyword belongs, and display the behavior type to the user. In response to the user's classification operation, it is determined that the behavior type belongs to a judgment-type emotional behavior or an integral-type emotional behavior.
[0050] In the above scheme, the emotional behavior involved in the emotional risk confidence prediction is determined based on the behavior type to which the keywords extracted from the sample emotional text belong, and the classification of the behavior type is realized, which creates a basis for the subsequent determination of the emotional risk confidence.
[0051] In some embodiments of the present disclosure, the process of obtaining the sample emotional text includes:
[0052] Obtain multiple original emotional texts, and use pre-set data governance tools to pre-process and cleanse each of them, thereby improving the integrity and accuracy of the text. Through data fusion technology, different original emotional texts can be converged to obtain a unified sample emotional text.
[0053] Through data governance tools, the quality of the original sentiment text is analyzed in multiple dimensions such as null value rate, efficiency, accuracy, and completeness based on the logical rule matching model and the regular expression matching model. The data governance dimension is determined based on the analysis results, and data governance is performed based on the data instruction dimension. In addition, the basic information of the sentiment object in the original sentiment text is extracted through the semantic analysis algorithm, and the extracted address is used as the data base for subsequent text processing. In addition, invalid data in the original sentiment text is eliminated to improve the efficiency of subsequent text processing.
[0054] Figure 2 A flow chart of another confidence prediction method based on emotional behavior provided in an embodiment of the present disclosure. In some embodiments of the present disclosure, the confidence prediction method based on emotional behavior further includes:
[0055] Step 201, determining the comparative value of the impact of pairwise integral emotional behaviors on the emotional risk confidence.
[0056] The impact comparison value may be a comparison result of the degree of influence of one of the two integral-type emotional behaviors on the emotional risk confidence of the other integral-type emotional behavior. The impact comparison value may include equally important, slightly important, more important, strongly important, extremely important, etc. For example, for the first integral-type emotional behavior and the second integral-type emotional behavior, the impact comparison value may characterize the importance of the first integral-type emotional behavior on the emotional risk confidence relative to the second integral-type emotional behavior.
[0057] In this embodiment, the influence comparison value of the pairwise integral emotional behavior on the emotional risk confidence can be manually determined in advance, and the influence comparison value is input into the confidence prediction device based on emotional behavior. The confidence prediction device based on emotional behavior obtains the influence comparison value.
[0058] Step 202, establishing a comparison value matrix according to the influence comparison values; wherein the integral-type emotional behaviors used as comparison references in the influence comparison values in the same column of the comparison value matrix are the same, and the integral-type emotional behaviors used as comparison subjects in the influence comparison values in the same row of the comparison value matrix are the same.
[0059] Among them, the comparison value matrix can be a matrix whose elements are influence comparison values, and the comparison value matrix is also called a judgment matrix. The comparison reference can be the integral emotional behavior compared with other integral emotional behaviors in the influence comparison value. The comparison subject can be the integral emotional behavior compared with other integral emotional behaviors in the influence comparison value. For example, if the influence comparison value represents the importance of the first integral emotional behavior relative to the second integral emotional behavior on the emotional risk confidence, then the first integral emotional behavior is the comparison subject and the second integral emotional behavior is the comparison reference.
[0060] In this embodiment, the confidence prediction device based on emotional behavior can arrange the same impact comparison values of comparison references in the same column of the matrix, and arrange the same impact comparison values of comparison subjects in the same row of the matrix to obtain a comparison value matrix.
[0061] For example, if the integral emotional behavior includes y1, y2, ...y 16 There are 16 in total. Take two integral emotional behaviors each time. For example, the two integral emotional behaviors include y i and j , use b ij Represents y i and j For the impact comparison value of the emotional risk confidence, two integral emotional behaviors are selected from the 16 integral emotional behaviors in turn, and the impact comparison values of the two integral emotional behaviors are determined. The comparison value matrix B is determined according to the impact comparison value. The comparison value matrix can be:
[0062]
[0063] Step 203: If the comparison value matrix passes the consistency check, the average value of each column affecting the comparison value in the comparison value matrix is calculated to obtain the comparison value average value, and the sum of the comparison value average values is calculated to obtain the sum of the average values.
[0064] In this embodiment, if the comparison value matrix passes the consistency check, it means that the comparison value matrix meets the consistency requirement. Then, taking each column in the comparison value matrix as a basic unit, the average value of the comparison value influencing each column is calculated to obtain the average value of the comparison value corresponding to each column. The sum of the comparison value averages is calculated to obtain the sum of the average values.
[0065] Step 204, calculating the quotient between the average value of the comparison values and the sum of the average values, and obtaining the behavior weight of the corresponding integral-type emotional behavior.
[0066] In this embodiment, the quotient between the average value of each comparison value and the sum of the average values is calculated to obtain the behavior weight of the integral type emotional behavior corresponding to the column where the average value of the comparison value is located.
[0067] In this embodiment, if the integral emotional behavior includes y1, y2, ...y 16 There are 16 in total. The corresponding behavioral weights of integral emotional behaviors are w1, w2, ... w 16 . For integral emotional behavior i The corresponding behavior weight w i , w i The value of is the average value of the influence comparison value of the i-th column divided by the sum of the average values of the influence comparison values of each column. Correspondingly, the behavior weight matrix W determined based on the behavior weight is: [w1w2…w 16 ].
[0068] In the above scheme, based on the impact comparison value between pairwise integral emotional behaviors, the behavior weight corresponding to each integral emotional behavior is determined, so that the behavior weight can more accurately represent the importance of the corresponding integral emotional behavior to the emotional risk confidence.
[0069] In some embodiments of the present disclosure, the comparison value matrix passes a consistency check, including: normalizing the comparison value matrix with column vectors as basic units to obtain a column normalized matrix; summing the column normalized matrix by rows and normalizing it to obtain a normalized column vector; determining the maximum eigenvalue approximation based on the normalized column vector and the comparison value matrix; determining a consistency index based on the maximum eigenvalue approximation and the order of the comparison value matrix; if the consistency index is not greater than a preset index threshold, determining that the comparison value matrix passes the consistency check.
[0070] The maximum eigenvalue approximation may be an approximation of the maximum eigenvalue of the comparison value matrix. The consistency index may be an index value representing the consistency of the comparison value matrix. The larger the consistency index, the more serious the representation inconsistency. The preset index threshold may be a preset maximum value of the consistency index representing the comparison value matrix passing the consistency check.
[0071] In this embodiment, the confidence prediction device based on emotional behavior can perform normalization processing on each column vector in the comparison matrix to obtain a column normalized matrix. The column normalized matrix is summed row by row, and the summation result is normalized to obtain a normalized column vector. The maximum eigenvalue approximation of the comparison value matrix is calculated based on the normalized column vector and the comparison value matrix. Further, the order of the comparison value matrix is determined, and the consistency index is determined based on the maximum eigenvalue approximation and the order of the comparison value matrix. It is judged whether the consistency index is greater than the preset index threshold. If so, it is determined that the comparison value matrix has not passed the consistency check. If not, it is determined that the comparison value matrix has passed the consistency check.
[0072] For example, if the comparison value matrix is B, each column vector in B is normalized to obtain a column normalized matrix, and the elements in the column normalized matrix are Specifically, b ij Represents y i and j Comparison of the impact on the confidence level of emotional risk. Sum the column normalization matrix row by row to obtain the intermediate column vector, the elements of which are Specifically, Normalize the middle column vector to get the normalized column vector. The element in the normalized column vector is w i , specifically,
[0073] Furthermore, for the maximum eigenvalue approximation λ, For the consistency index CI, Where n is the order of the comparison value matrix. If CI = 0, it means that there is complete consistency; if CI is close to 0, it means that there is a relatively satisfactory consistency; the larger the CI, the more serious the inconsistency.
[0074] In the above scheme, the consistency check of the comparison value matrix is implemented, and it is determined whether the behavior weight can be calculated based on the comparison value matrix.
[0075] In some embodiments of the present disclosure, each prediction value interval may have a corresponding color, and the color may be set according to user needs, etc., and this embodiment does not impose any restrictions. Taking the three colors of green, yellow, and red as an example, the prediction value interval corresponding to green represents that the emotional risk confidence is in a low-risk interval, the prediction value interval corresponding to yellow represents that the emotional risk confidence is in a medium-risk interval, and the prediction value interval corresponding to red represents that the emotional risk confidence is in a high-risk interval. Based on this color, a reminder of the current emotional risk of the emotional object can be realized, and then the emotional objects with higher emotional risks can be displayed more clearly.
[0076] In the disclosed embodiment, quantitative prediction of emotional trends is achieved through veto-type judgment-type emotional behaviors and integral-type emotional behaviors, combined with calculation rules such as specific texts, emotional impact values, and indicator weights, and periodic quantitative calculations of emotional risks can be performed to achieve automatic early warning of emotional risks. Each pair of integral-type emotional behaviors is compared pairwise to construct a comparison value matrix. The elements in the comparison value matrix represent the relative importance of two integral-type emotional behaviors, and the impact comparison values in the comparison value matrix are numerical values obtained by quantifying the relative importance. The comparison value matrix is processed by methods such as normalization and eigenvector method to obtain the behavioral weight of each integral-type emotional behavior. The scientific measurement of integral weights is achieved, which improves the effectiveness and credibility of the final emotional risk confidence level.
[0077] In addition, it takes into account both criticality and comprehensiveness, taking judgment-type emotional behavior as the representative of criticality and integral-type emotional behavior as the representative of comprehensiveness, and conducts an emotional risk assessment at multiple levels and dimensions, making the assessment more accurate and comprehensive, and able to effectively identify and warn of potential emotional risks, thus avoiding further escalation of emotional risks.
[0078] Figure 3 This is a schematic diagram of the structure of a confidence prediction device based on emotional behavior provided by an embodiment of the present disclosure. The device can be implemented by software and / or hardware, and the device can also be integrated in an electronic device. Figure 3 As shown, the confidence prediction device based on emotional behavior includes:
[0079] The first extraction module 301 is used to extract the first text in the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior;
[0080] A second extraction module 302 is used to extract a second text from the emotional behavior text if the first text is not extracted, and determine the emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral emotional behavior;
[0081] The first prediction module 303 is used to predict the emotional trend according to the behavior weight corresponding to the integral emotional behavior and the emotional influence value, and obtain the emotional risk confidence between the corresponding emotional objects.
[0082] Optionally, the first extraction module 301 is used to:
[0083] Extracting sensitive behavior text from the emotional behavior text using a text extraction model to obtain the first text;
[0084] The extracting the second text from the emotional behavior text includes:
[0085] The text extraction model is used to extract the emotion object attributes and non-sensitive behavior text from the emotion behavior text to obtain the second text.
[0086] Optionally, determining the sentiment impact value corresponding to the second text includes:
[0087] A query process is performed on a preset text impact value relationship according to the second text to obtain the sentiment impact value.
[0088] Optionally, the confidence prediction device based on emotional behavior further includes:
[0089] The second prediction module is configured to set the sentiment risk confidence level to a highest risk value if the first text is extracted.
[0090] Optionally, the confidence prediction device based on emotional behavior further includes:
[0091] The third extraction module is used to obtain sample emotional text, and extract keywords from the sample emotional text through a text extraction model to obtain emotional keywords;
[0092] The first determination module is used to determine the judgment-type emotional behavior and the integral-type emotional behavior according to the behavior type to which the emotional keyword belongs.
[0093] Optionally, the confidence prediction device based on emotional behavior further includes:
[0094] A second determination module is used to determine the comparison value of the influence of each of the integral-type emotional behaviors on the emotional risk confidence level;
[0095] A matrix establishment module, used to establish a comparison value matrix according to the influence comparison value; wherein the integral-type emotional behaviors used as comparison references in the influence comparison values of the same column in the comparison value matrix are the same, and the integral-type emotional behaviors used as comparison subjects in the influence comparison values of the same row in the comparison value matrix are the same;
[0096] A first calculation module is used to calculate the average value of each column affecting the comparison value in the comparison value matrix to obtain the comparison value average value if the comparison value matrix passes the consistency check, and calculate the sum of the comparison value average values to obtain the sum of the average values;
[0097] The second calculation module is used to calculate the quotient between the average value of the comparison value and the sum of the average values to obtain the behavior weight of the corresponding integral-type emotional behavior.
[0098] Optionally, the comparison value matrix passes a consistency check, including:
[0099] The comparison value matrix is normalized with column vectors as basic units to obtain a column normalized matrix; the column normalized matrix is summed by rows and normalized to obtain a normalized column vector; the maximum eigenvalue approximation is determined according to the normalized column vector and the comparison value matrix; a consistency index is determined according to the maximum eigenvalue approximation and the order of the comparison value matrix; if the consistency index is not greater than a preset index threshold, it is determined that the comparison value matrix passes the consistency check.
[0100] The confidence prediction device based on emotional behavior provided by the embodiments of the present disclosure can execute the confidence prediction method based on emotional behavior provided by any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method.
[0101] Figure 4 The structure diagram of an electronic device provided by the embodiment of the present disclosure is shown in FIG. Figure 4 As shown, the electronic device 400 includes one or more processors 401 and a memory 402 .
[0102] The processor 401 may be a central processing unit (CPU) or other forms of processing units having confidence prediction capabilities based on emotional behavior and / or instruction execution capabilities, and may control other components in the electronic device 400 to perform desired functions.
[0103] The memory 402 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 401 may run the program instructions to implement the confidence prediction method based on emotional behavior of the embodiment of the present disclosure described above and / or other desired functions. Various contents such as input signals, signal components, noise components, etc. may also be stored in the computer-readable storage medium.
[0104] In one example, the electronic device 400 may further include: an input device 403 and an output device 404, and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).
[0105] In addition, the input device 403 may also include, for example, a keyboard, a mouse, and the like.
[0106] The output device 404 can output various information to the outside, including the determined distance information, direction information, etc. The output device 404 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.
[0107] Of course, to simplify, Figure 4 Only some of the components related to the present disclosure in the electronic device 400 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 400 may also include any other appropriate components.
[0108] In addition to the above-mentioned methods and devices, an embodiment of the present disclosure may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the confidence prediction method based on emotional behavior provided by the embodiment of the present disclosure.
[0109] The computer program product may be written in any combination of one or more programming languages to write program code for performing the operations of the disclosed embodiments, including object-oriented programming languages such as Java, C++, etc., and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.
[0110] In addition, the embodiments of the present disclosure may also be a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the confidence prediction method based on emotional behavior provided by the embodiments of the present disclosure.
[0111] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.
[0112] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0113] The above description is only a specific embodiment of the present disclosure, so that those skilled in the art can understand or implement the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A confidence prediction method based on emotional behavior, It is characterized in that include: Extracting a first text from the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior; If the first text is not extracted, extracting a second text from the emotional behavior text, and determining an emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral emotional behavior; Emotional trend prediction is performed based on the behavior weight corresponding to the integral-type emotional behavior and the emotional impact value to obtain the emotional risk confidence between the corresponding emotional objects.
2. The method according to claim 1, It is characterized in that The step of extracting the first text from the emotional behavior text includes: Extracting sensitive behavior text from the emotional behavior text using a text extraction model to obtain the first text; The extracting the second text from the emotional behavior text includes: The text extraction model is used to extract the emotion object attributes and non-sensitive behavior text from the emotion behavior text to obtain the second text.
3. The method according to claim 1, It is characterized in that The determining the emotional impact value corresponding to the second text includes: A query process is performed on a preset text impact value relationship according to the second text to obtain the sentiment impact value.
4. The method according to claim 1, It is characterized in that The method further comprises: If the first text is extracted, the emotional risk confidence is set to the highest risk value.
5. The method according to claim 1, It is characterized in that The method further comprises: Acquire sample sentiment text, and extract keywords from the sample sentiment text using a text extraction model to obtain sentiment keywords; The judgment-type emotional behavior and the integral-type emotional behavior are determined according to the behavior type to which the emotional keyword belongs.
6. The method according to claim 1, It is characterized in that The method further comprises: Determine the comparative values of the effects of the two integral emotional behaviors on the emotional risk confidence level; A comparison value matrix is established according to the influence comparison values; wherein the integral-type emotional behaviors used as comparison references in the influence comparison values in the same column of the comparison value matrix are the same, and the integral-type emotional behaviors used as comparison subjects in the influence comparison values in the same row of the comparison value matrix are the same; If the comparison value matrix passes the consistency check, the average value of each column affecting the comparison value in the comparison value matrix is calculated to obtain the comparison value average value, and the sum of the comparison value average values is calculated to obtain the sum of the average values; The quotient between the average value of the comparison value and the sum of the average values is calculated to obtain the behavior weight of the corresponding integral-type emotional behavior.
7. The method according to claim 6, It is characterized in that The comparison value matrix passes a consistency check, including: Normalizing the comparison value matrix using column vectors as basic units to obtain a column normalized matrix; Summing and normalizing the column normalization matrix row by row to obtain a normalized column vector; Determine a maximum eigenvalue approximation according to the normalized column vector and the comparison value matrix; Determining a consistency index according to the maximum eigenvalue approximation and the order of the comparison value matrix; If the consistency index is not greater than the preset index threshold, it is determined that the comparison value matrix passes the consistency check.
8. A confidence prediction device based on emotional behavior, It is characterized in that include: A first extraction module is used to extract a first text from the emotional behavior text; wherein the first text is used to characterize the judgment-type emotional behavior; A second extraction module is used to extract a second text from the emotional behavior text if the first text is not extracted, and determine the emotional impact value corresponding to the second text; wherein the second text is used to characterize the integral emotional behavior; The first prediction module is used to predict the emotional trend according to the behavior weight corresponding to the integral emotional behavior and the emotional influence value, and obtain the emotional risk confidence between the corresponding emotional objects.
9. An electronic device, It is characterized in that The electronic device comprises: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the confidence prediction method based on emotional behavior as described in any one of claims 1-7.
10. A computer-readable storage medium, It is characterized in that The storage medium stores a computer program, and the computer program is used to execute the confidence prediction method based on emotional behavior as described in any one of claims 1-7 above.
