A brain speech assessment method and system based on DSM-V cognitive domain

Abstract

The present invention relates to the field of computer software technology, and specifically discloses a brain speech assessment method and system based on the DSM-V cognitive domain, including: obtaining the user's speech information; extracting features of the speech information based on the DSM-V cognitive domain to obtain spoken feature information, wherein the spoken feature information includes lexical response feature information, lexical semantic feature information, and lexical association feature information; obtaining multiple stages of lexical feature information according to the lexical response feature information, and obtaining lexical response feature evaluation values ​​according to all the lexical feature information of the stages; and obtaining multiple lexical node feature information according to the lexical semantic feature information. The present invention lays the foundation for comprehensive assessment by obtaining speech information, extracts multi-dimensional spoken feature information based on the DSM-V cognitive domain, can deeply analyze the brain's speech function, and solves the problem of single dimension of traditional assessment, such as accurately locating the root cause of the problem in the diagnosis of language disorders.

Description

Technical Field

[0001] The present invention relates to the field of computer software technology, and in particular to a brain speech assessment method and system based on the DSM-V cognitive domain. Background Art

[0002] Brain speech assessment is of great significance in many fields. In the field of language education, accurate assessment of students' brain speech function helps teachers understand students' language learning status, develop personalized teaching plans, improve teaching effectiveness, and promote the comprehensive development of students' language ability.

[0003] Existing brain speech assessment methods only focus on one aspect, such as focusing only on pronunciation accuracy and evaluating by calculating pronunciation error rates. However, brain speech function is a complex system involving multiple cognitive domains such as vocabulary comprehension, semantic application, and vocabulary association. Evaluating from only a single dimension cannot fully reflect the user's true brain speech ability, and may miss important information, resulting in inaccurate evaluation results. Summary of the invention

[0004] The purpose of the present invention is to provide a brain speech assessment method and system based on the DSM-V cognitive domain to solve the technical problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A brain speech assessment method based on the DSM-V cognitive domain, including:

[0007] Get the user's voice information;

[0008] Extracting features of speech information based on the DSM-V cognitive domain to obtain spoken language feature information, wherein the spoken language feature information includes lexical response feature information, lexical semantic feature information and lexical association feature information;

[0009] Acquire multiple stages of vocabulary feature information according to the vocabulary response feature information, and acquire vocabulary response feature evaluation values ​​according to all the stages of vocabulary feature information;

[0010] Acquire multiple vocabulary node feature information according to the vocabulary semantic feature information, and acquire a semantic feature evaluation value according to each vocabulary node feature information;

[0011] Acquire a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and acquire a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information;

[0012] A speech comprehensive evaluation value is obtained according to the vocabulary reaction feature evaluation value, the semantic feature evaluation value and the vocabulary association feature evaluation value, and a brain speech evaluation is performed on the user according to the speech comprehensive evaluation value.

[0013] Preferably, the step of extracting features from speech information based on the DSM-V cognitive domain to obtain spoken language feature information includes:

[0014] Acquiring voice sample information according to the voice information;

[0015] Acquire speech vocabulary reaction time according to the speech sample information;

[0016] Extracting features from the speech sample information based on the speech-vocabulary reaction time to obtain vocabulary reaction feature information;

[0017] Get the preset corpus;

[0018] Divide the speech sample information based on the preset corpus to obtain multiple node words;

[0019] Extracting semantic features of each node word to obtain semantic feature information of the node word;

[0020] Collecting the lexical semantic feature information of all the nodes to obtain the lexical semantic feature information;

[0021] Acquire multiple associated words according to the node words;

[0022] Performing semantic division on all the associative words to obtain multiple semantic categories;

[0023] Get the similar vocabulary ratio of each node vocabulary in the semantic classification;

[0024] Acquire the similarity of node vocabulary based on the similar vocabulary ratio;

[0025] Determine whether the similarity of the node vocabulary is less than a preset value;

[0026] If it is not less than, the node vocabulary is removed;

[0027] If it is less than, the node vocabulary is retained, and the associated features are extracted based on the retained node vocabulary to obtain the vocabulary associated feature information.

[0028] Preferably, the step of acquiring multiple stages of vocabulary feature information according to the vocabulary reaction feature information, and acquiring vocabulary reaction feature evaluation values ​​according to all the stage vocabulary feature information includes:

[0029] Acquire multiple recognition words according to the voice sample information;

[0030] Acquiring phoneme feature information according to the recognition vocabulary, wherein the phoneme feature information includes pronunciation duration, pitch value, and sound intensity value;

[0031] Acquiring vocabulary accuracy according to the pronunciation duration, pitch value, and sound intensity value;

[0032] Get the recognition time interval between each recognition word;

[0033] Acquire a vocabulary extraction delay time of each recognized vocabulary word according to the recognition time interval;

[0034] Obtaining a semantic response decision for each of the recognized words, wherein the semantic response decision includes a semantic extraction time, a semantic response duration, and a semantic response operation time;

[0035] Acquire a vocabulary response value according to the semantic extraction time, the semantic response duration, and the semantic response operation time;

[0036] A vocabulary response feature evaluation value is obtained based on the vocabulary accuracy and the vocabulary response value.

[0037] Preferably, the step of obtaining a semantic feature evaluation value according to each of the lexical node feature information comprises:

[0038] Acquire the context of each node vocabulary according to the vocabulary node feature information;

[0039] Acquire the semantic coherence of each node word according to the context, and acquire the semantic coherence value according to the semantic coherence of all node words;

[0040] Obtaining position information of each node word in the semantic hierarchy;

[0041] Obtaining the semantic depth of the node vocabulary according to the position information, and obtaining the semantic depth value of the vocabulary node according to the semantic depths corresponding to all the semantic depths;

[0042] The semantic feature evaluation value is obtained according to the semantic coherence value and the semantic depth value of the vocabulary node.

[0043] Preferably, the step of obtaining a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and obtaining a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information comprises:

[0044] Acquire vocabulary association features according to the vocabulary association feature information;

[0045] Acquire multiple associative words based on the word association features;

[0046] Performing semantic classification on all the associative words to obtain a plurality of associative semantic categories;

[0047] Obtaining the associative lexical strength of the associative lexical items in each associative semantic category based on the vector space;

[0048] Acquire all the associative word strengths to obtain word associative strength feature information;

[0049] Obtaining the overlap rate of each of the associative words in the associative semantic category;

[0050] The word association feature evaluation value is obtained according to the associative word strength and the overlap rate.

[0051] Preferably, the step of obtaining the comprehensive speech evaluation value according to the vocabulary reaction feature evaluation value, the semantic feature evaluation value and the vocabulary association feature evaluation value comprises:

[0052] Obtaining the vocabulary response feature evaluation value;

[0053] Acquire a first weight coefficient according to the vocabulary reaction feature evaluation value;

[0054] Obtaining the semantic feature evaluation value;

[0055] Acquire a second weight coefficient according to the semantic feature evaluation value;

[0056] Obtaining word association feature evaluation values;

[0057] Acquire a third weight coefficient according to the word association feature evaluation value;

[0058] The speech comprehensive evaluation value is obtained according to the vocabulary reaction feature evaluation value, the first weight coefficient, the semantic feature evaluation value, the second weight coefficient, the vocabulary association feature evaluation value and the third weight coefficient.

[0059] The present invention also provides a brain speech assessment system based on the DSM-V cognitive domain, comprising:

[0060] The first acquisition module is used to acquire the user's voice information;

[0061] A second acquisition module is used to extract features from speech information based on the DSM-V cognitive domain to obtain spoken language feature information, wherein the spoken language feature information includes lexical response feature information, lexical semantic feature information and lexical association feature information;

[0062] A third acquisition module is used to acquire multiple stages of vocabulary feature information according to the vocabulary reaction feature information, and acquire vocabulary reaction feature evaluation values ​​according to all the stage vocabulary feature information;

[0063] A fourth acquisition module, configured to acquire a plurality of vocabulary node feature information according to the vocabulary semantic feature information, and acquire a semantic feature evaluation value according to each of the vocabulary node feature information;

[0064] A fifth acquisition module, configured to acquire a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and acquire a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information;

[0065] The evaluation module is used to obtain a comprehensive speech evaluation value according to the vocabulary reaction feature evaluation value, the semantic feature evaluation value and the vocabulary association feature evaluation value, and to perform a brain speech evaluation on the user according to the comprehensive speech evaluation value.

[0066] Preferably, the second acquisition module includes:

[0067] A first acquisition unit, configured to acquire voice sample information according to the voice information;

[0068] A second acquisition unit, used for acquiring a speech vocabulary reaction time according to the speech sample information;

[0069] A third acquisition unit, configured to extract features from the speech sample information based on the speech vocabulary reaction time to obtain vocabulary reaction feature information;

[0070] A fourth acquisition unit, used to acquire a preset corpus;

[0071] A fifth acquisition unit, configured to divide the speech sample information based on a preset corpus to obtain a plurality of node words;

[0072] A sixth acquisition unit, configured to extract semantic features of each node word to obtain semantic feature information of the node word;

[0073] A seventh acquisition unit, used for collecting the lexical semantic feature information of all the nodes to obtain the lexical semantic feature information;

[0074] An eighth acquisition unit, configured to acquire a plurality of associated words according to the node words;

[0075] A ninth acquisition unit, configured to semantically divide all the associative words to obtain a plurality of semantic categories;

[0076] a tenth acquisition unit, used for acquiring a similar vocabulary ratio of each node vocabulary in the semantic classification;

[0077] an eleventh obtaining unit, configured to obtain the similarity of node words based on the similar word ratio;

[0078] A judging unit, used to judge whether the similarity of the node words is less than a preset value;

[0079] If it is not less than, the node vocabulary is removed;

[0080] If it is less than, the node vocabulary is retained, and the associated features are extracted based on the retained node vocabulary to obtain the vocabulary associated feature information.

[0081] Preferably, the third acquisition module includes:

[0082] a twelfth acquisition unit, configured to acquire a plurality of recognition words according to the voice sample information;

[0083] a thirteenth acquiring unit, configured to acquire phoneme feature information according to the recognition vocabulary, wherein the phoneme feature information includes pronunciation duration, pitch value, and sound intensity value;

[0084] A fourteenth acquisition unit, configured to acquire vocabulary accuracy according to the pronunciation duration, pitch value, and sound intensity value;

[0085] A fifteenth acquiring unit, used for acquiring a recognition time interval between each recognition vocabulary;

[0086] a sixteenth acquisition unit, configured to acquire a vocabulary extraction delay time of each recognized vocabulary word according to the recognition time interval;

[0087] A seventeenth acquisition unit, configured to acquire a semantic response decision of each of the recognized words, wherein the semantic response decision includes a semantic extraction time, a semantic response duration, and a semantic response operation time;

[0088] An eighteenth acquisition unit, configured to acquire a vocabulary response value according to the semantic extraction time, the semantic response duration, and the semantic response operation time;

[0089] A vocabulary response feature evaluation value is obtained based on the vocabulary accuracy and the vocabulary response value.

[0090] Preferably, the fourth acquisition module includes:

[0091] A nineteenth acquisition unit, configured to acquire the context of each of the node words according to the vocabulary node feature information;

[0092] a twentieth acquisition unit, configured to acquire the semantic coherence of each node word according to the context, and acquire a semantic coherence value according to the semantic coherence of all node words;

[0093] A twenty-first acquisition unit, configured to acquire position information of each of the node words in the semantic hierarchy;

[0094] A twenty-second acquisition unit, configured to acquire the semantic depth of the node vocabulary according to the position information, and acquire the semantic depth value of the vocabulary node according to the semantic depths corresponding to all the semantic depths;

[0095] The twenty-third acquisition unit is used to acquire the semantic feature evaluation value according to the semantic coherence value and the semantic depth value of the vocabulary node.

[0096] The beneficial effects of the present application are as follows: the present invention lays the foundation for comprehensive evaluation by acquiring voice information, extracts multi-dimensional oral feature information based on the DSM-V cognitive domain, can deeply analyze the brain's voice function, and solve the problem of single traditional evaluation dimension, such as accurately locating the root cause of the problem in language disorder diagnosis. The lexical reaction, semantics, and association features are carefully analyzed and evaluated respectively, which can improve students' abilities in language learning in a targeted manner, for example, helping learners optimize vocabulary application strategies. The comprehensive evaluation value of each evaluation value is obtained and the brain voice evaluation is performed. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 The figure is a schematic diagram of a method flow of an embodiment of the present application.

[0098] Figure 2 A schematic diagram of the system structure of an embodiment of the present application.

[0099] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0100] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0101] like Figure 1 As shown, the present application provides a brain speech assessment method based on the DSM-V cognitive domain, comprising:

[0102] S1. Obtain the user's voice information;

[0103] S2, extracting features of speech information based on DSM-V cognitive domain to obtain spoken language feature information, wherein the spoken language feature information includes lexical response feature information, lexical semantic feature information and lexical association feature information;

[0104] S3, acquiring multiple stages of vocabulary feature information according to the vocabulary response feature information, and acquiring vocabulary response feature evaluation values ​​according to all the stages of vocabulary feature information;

[0105] S4, obtaining a plurality of vocabulary node feature information according to the vocabulary semantic feature information, and obtaining a semantic feature evaluation value according to each of the vocabulary node feature information;

[0106] S5, obtaining a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and obtaining a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information;

[0107] S6. Obtain a comprehensive speech evaluation value according to the vocabulary reaction feature evaluation value, the semantic feature evaluation value, and the vocabulary association feature evaluation value, and perform a brain speech evaluation on the user according to the comprehensive speech evaluation value.

[0108] There are many problems with existing brain speech assessment technologies. Traditional assessment methods are often single-dimensional, focusing only on local aspects such as pronunciation or vocabulary, and are difficult to fully reflect the brain's speech function. Many assessments lack the support of scientific theoretical models, resulting in insufficient accuracy and reliability of the assessment results.

[0109] As described in the above steps S1-S6, the present invention extracts features from speech information based on the DSM-V cognitive domain to obtain spoken language feature information, including vocabulary response feature information, vocabulary semantic feature information and vocabulary association feature information. Speech is deeply analyzed from multiple key dimensions. Taking vocabulary semantic feature information as an example, by analyzing the node features in the vocabulary semantic network, the coherence and depth of vocabulary semantics can be accurately evaluated. Then, vocabulary feature information of multiple stages is obtained according to the vocabulary response feature information and an evaluation value is obtained. By accurately evaluating the performance of the brain in stages such as vocabulary recognition, semantic extraction and reaction decision-making. In the vocabulary recognition stage, the analysis of features such as pronunciation duration, pitch value, and sound intensity value can determine the user's perception ability of vocabulary. For example, the user can accurately recognize vocabulary by adjusting the pronunciation duration and sound intensity, indicating that his vocabulary perception ability is strong. This solves the problem of rough evaluation of the vocabulary response process in the past, and accurate training plans can be formulated for problems at different stages. For users with slow vocabulary response problems, their performance at each stage can be analyzed to determine whether it is difficult to recognize vocabulary or slow semantic extraction, and then targeted training can be carried out to improve language ability and efficiency. Then, the vocabulary node feature information is obtained based on the vocabulary semantic feature information and the semantic feature evaluation value is calculated; and the good grasp of semantic depth solves the problem of insufficiently detailed evaluation of vocabulary semantic understanding, which helps to improve the diagnostic accuracy of semantic-related language disorders. By evaluating their vocabulary semantic features, the degree of impairment of their semantic understanding and expression can be more accurately understood. Then, the vocabulary association strength feature information is obtained based on the vocabulary association feature information and the evaluation value is obtained. This can quantify the vocabulary association features and effectively evaluate the association strength and thinking flexibility between words. After that, the comprehensive speech evaluation value is obtained based on the above-mentioned feature evaluation values ​​and the brain speech evaluation is performed. Its beneficial effect is to integrate the evaluation values ​​of multiple dimensions to achieve a comprehensive and objective evaluation of the brain speech function. The setting of different weight coefficients can be adjusted according to the evaluation purpose and the characteristics of the user group to make the evaluation results more targeted. For people engaged in creative work, the weight of the vocabulary association feature can be increased. This solves the problem that a single assessment value cannot fully reflect the brain's speech function, and provides a reliable basis for accurately judging the state of the brain's speech function and formulating personalized intervention plans. Whether it is rehabilitation treatment for patients with speech disorders or language ability improvement training for normal people, accurate and effective guidance can be obtained based on this.

[0110] In one embodiment, the step of extracting features from speech information based on the DSM-V cognitive domain to obtain spoken language feature information includes:

[0111] S201, obtaining voice sample information according to the voice information;

[0112] S202, obtaining voice vocabulary reaction time according to the voice sample information;

[0113] S203, extracting features from the speech sample information based on the speech-vocabulary reaction time to obtain vocabulary reaction feature information;

[0114] S204, obtaining a preset corpus;

[0115] S205, dividing the speech sample information based on a preset corpus to obtain a plurality of node words;

[0116] S206, extracting semantic features of each node vocabulary to obtain semantic feature information of the node vocabulary;

[0117] S207, collecting lexical semantic feature information of all the nodes to obtain lexical semantic feature information;

[0118] S208, obtaining a plurality of associative words according to the node words;

[0119] S209, performing semantic division on all the associative words to obtain a plurality of semantic categories;

[0120] S210, obtaining the similar vocabulary ratio of each node vocabulary in the semantic classification;

[0121] S211, obtaining the similarity of node words based on the similar word ratio;

[0122] S212, determining whether the similarity of the node vocabulary is less than a preset value;

[0123] If it is not less than, the node vocabulary is removed;

[0124] If it is less than, the node vocabulary is retained, and the associated features are extracted based on the retained node vocabulary to obtain the vocabulary associated feature information.

[0125] As described in the above steps S201-S212, the present invention constructs a vocabulary reaction evaluation basis through steps S201-S203, and obtains voice sample information from voice information. For example, in language ability assessment, accurate voice sample information is the cornerstone of subsequent analysis, avoiding evaluation bias caused by data source errors. Acquiring voice vocabulary reaction time accurately quantifies the speed at which the brain processes vocabulary. By extracting features from voice sample information through voice vocabulary reaction time to obtain vocabulary reaction feature information, it is possible to comprehensively analyze the performance of the brain in the whole process from vocabulary perception, recognition to reaction output, and then mining the overall semantic features of vocabulary through steps S204-S207 to obtain a preset corpus, providing a standardized reference for subsequent operations. In the language research comparison experiment, a unified corpus ensures the consistency of vocabulary division and analysis between different samples. Based on the preset corpus, voice sample information is divided to obtain multiple node vocabularies, which helps to decompose complex voice information into analyzable vocabulary units. Semantic feature extraction of node vocabulary is performed to obtain node vocabulary semantic feature information, and the semantic connotation of a single vocabulary is deeply explored. The semantic feature information of node vocabulary is collected to obtain the semantic feature information of vocabulary, realizing the semantic integration from individual to whole. Then, the vocabulary association is analyzed through the steps S208-S211, and multiple associative vocabulary is obtained according to the node vocabulary, which opens the exploration of vocabulary association ability. The associative vocabulary is semantically divided to obtain multiple semantic classifications, and the logical structure of vocabulary association is sorted out. For example, when constructing a vocabulary knowledge graph, a clear semantic classification helps to establish an accurate vocabulary association relationship. The proportion of similar vocabulary in the semantic classification of node vocabulary is obtained, and the semantic similarity between vocabulary is quantitatively measured. The similarity of node vocabulary is obtained based on the proportion of similar vocabulary, and the evaluation of vocabulary association degree is further refined. Finally, the vocabulary association feature extraction step is optimized through the step S212, the similarity of node vocabulary is judged and screened, and the valid node vocabulary is retained for association feature extraction to obtain vocabulary association feature information. When constructing a language model, this step can exclude noise vocabulary, solve the problem of irrelevant vocabulary interfering with vocabulary association feature extraction, and ensure that the acquired vocabulary association features are more representative and effective, thereby improving the analysis ability of the entire brain speech evaluation system at the vocabulary association level.

[0126] In one embodiment, the step of acquiring multiple stages of vocabulary feature information according to the vocabulary reaction feature information, and acquiring vocabulary reaction feature evaluation values ​​according to all the stage vocabulary feature information, includes:

[0127] S301, acquiring a plurality of recognition words according to the voice sample information;

[0128] S302, acquiring phoneme feature information according to the recognition vocabulary, wherein the phoneme feature information includes pronunciation duration, pitch value, and sound intensity value;

[0129] S303, obtaining vocabulary accuracy according to the pronunciation duration, pitch value, and sound intensity value;

[0130] S304, obtaining the recognition time interval between each recognition word;

[0131] S305, acquiring a vocabulary extraction delay time of each recognized vocabulary according to the recognition time interval;

[0132] S306, obtaining a semantic response decision of each of the recognized words, wherein the semantic response decision includes a semantic extraction time, a semantic response duration, and a semantic response operation time;

[0133] S307, obtaining a vocabulary response value according to the semantic extraction time, the semantic response duration, and the semantic response operation time;

[0134] S308: Obtain a vocabulary response feature evaluation value according to the vocabulary accuracy and the vocabulary response value.

[0135] As described in the above steps S301-S308, the present invention can accurately locate the vocabulary recognition and pronunciation ability through S301-S303, accurately obtain multiple recognition words from the voice sample information, and further mine the phoneme feature information of the recognition words. This operation is like installing a microscope on the pronunciation of words. Taking speech therapy as an example, for patients with pronunciation disorders, the precise analysis of pronunciation duration, pitch value and sound intensity value can accurately locate the pronunciation problem, such as determining whether the pronunciation duration of a certain phoneme is too long or too short, or the pitch is unstable, etc., which overcomes the dilemma of insufficient grasp of pronunciation details in traditional evaluation. Based on these phoneme features, the vocabulary accuracy is then calculated, and then the quantified vocabulary recognition speed and extraction efficiency are realized through S304-S305. By obtaining the recognition time interval between the recognition words, in the reading ability assessment, this data is like an accurate ruler. For example, in rapid reading training, by measuring the recognition time interval of the reader to the continuous words, the reading speed and information processing fluency can be clearly judged, which solves the problem of not being able to effectively measure the brain's continuous processing speed of words. By calculating the vocabulary retrieval delay time, in the study of vocabulary memory, if the subject's retrieval delay time is too long when recalling vocabulary, it may imply that there are obstacles in their memory storage structure or retrieval path. This provides key clues for in-depth exploration of the vocabulary memory mechanism and solves the difficulty of accurately quantifying the speed of vocabulary retrieval from memory. Then, through S306-S307, a comprehensive evaluation of the semantic response decision process is achieved. By obtaining the relevant time information of the semantic response decision, in daily communication scenarios, this helps us understand how the brain quickly understands semantics and responds. For example, in a group discussion, a person's answer speed and accuracy to questions reflect the comprehensive performance of their semantic retrieval time, reaction duration and operation time. If the semantic retrieval time is too long, it may be difficult to understand complex topics; unstable reaction duration may mean that there are problems with the coherence of thinking; too long operation time may affect communication efficiency. This information overcomes the limitations of the previous fragmented evaluation of the semantic response decision process. After that, the vocabulary response value is calculated by combining these time information, and finally the comprehensive evaluation of the vocabulary response characteristics is achieved through S308, and then the vocabulary response feature evaluation value is obtained by combining the vocabulary accuracy and vocabulary response value.

[0136] In one embodiment, the step of obtaining a semantic feature evaluation value according to each of the lexical node feature information comprises:

[0137] S401, acquiring the context of each node vocabulary according to the vocabulary node feature information;

[0138] S402, obtaining the semantic coherence of each node vocabulary according to the context, and obtaining the semantic coherence value according to the semantic coherence of all node vocabulary;

[0139] S403, obtaining the position information of each node word in the semantic hierarchy;

[0140] S404, obtaining the semantic depth of the node vocabulary according to the position information, and obtaining the semantic depth value of the vocabulary node according to the semantic depths corresponding to all the semantic depths;

[0141] S405: Obtain a semantic feature evaluation value according to the semantic coherence value and the semantic depth value of the vocabulary node.

[0142] As described in the above steps S401-S405, the present invention realizes context and semantic coherence evaluation through S401-S402, and avoids the ambiguity generated when analyzing vocabulary in isolation by obtaining the contextual context of node vocabulary, and solves the problem of inaccurate semantic understanding caused by being out of context in traditional evaluation. Evaluating semantic coherence based on contextual context and obtaining semantic coherence value are of great significance in text quality evaluation. Reports with high semantic coherence values ​​can allow readers to smoothly understand the whole picture of the event. If a reply text has poor vocabulary semantic coherence, it may be difficult to understand the logic of the event. This overcomes the limitation of being difficult to quantify semantic coherence in the past, and provides an objective basis for evaluating text quality and diagnosing logical confusion caused by improper use of vocabulary in writing. Then, through S403-S404, a deep analysis of the semantic level is realized, and by obtaining the position information of the node vocabulary in the semantic hierarchy, a navigation role is played in vocabulary learning and knowledge construction. The problem of learners' lack of clear cognition of vocabulary semantic categories and hierarchical relationships is solved, and a framework is provided for building a knowledge network. Then, the semantic depth is determined according to the position information and the semantic depth value of the lexical node is obtained. Finally, a comprehensive evaluation of the semantic features is achieved through step S405. The semantic feature evaluation value is obtained by combining the semantic coherence value and the semantic depth value of the lexical node. This solves the problem that a single indicator cannot comprehensively evaluate the semantic features of the lexical word, integrates key semantic dimension information, improves the effectiveness and practicality of language ability assessment, and makes the semantic ability assessment in language-related fields more accurate.

[0143] In one embodiment, the step of obtaining a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and obtaining a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information comprises:

[0144] S501, obtaining vocabulary association features according to the vocabulary association feature information;

[0145] S502, acquiring a plurality of associative words based on the word association features;

[0146] S503, semantically classifying all the associative words to obtain a plurality of associative semantic categories;

[0147] S504, acquiring the associative vocabulary strength of the associative vocabulary in each associative semantic category based on the vector space;

[0148] S505, acquiring all the associative word strengths to obtain word associative strength feature information;

[0149] S506, obtaining the coincidence rate of each of the associative words in the associative semantic category;

[0150] S507: Obtain a word association feature evaluation value according to the associative word strength and the overlap rate.

[0151] As described in the above steps S501-S507, the present invention can mine vocabulary associations and construct semantic logic, and accurately obtain vocabulary association features from vocabulary association feature information. This process is like exploring a hidden treasure map in the brain's vocabulary treasure house. In the rehabilitation treatment of language disorders, for example, for patients with aphasia, therapists can find the patient's impaired vocabulary association pattern by analyzing vocabulary association features, such as patients may have difficulty establishing connections between certain semantically related words, thereby designing rehabilitation training programs in a targeted manner to help patients re-establish a normal vocabulary association network, solving the dilemma of the previous difficulty in accurately locating the vocabulary association problem of patients with language disorders. Then, multiple associative words are obtained based on vocabulary association features, which is crucial to improving the richness of language expression. Taking the language fluency training in the DSMV-V cognitive domain as an example, when a patient describes a scene, through the expansion of associative words, such as from "park" to "bench", "green tree", "tourist", etc., the description can be made more vivid and vivid, overcoming the problem of monotonous expression due to lack of vocabulary association. Semantic classification of associative words to obtain multiple associative semantic categories plays a key role in constructing the brain semantic knowledge map. In the process of studying children's language development, researchers can understand the formation process of their semantic cognitive structure by analyzing the semantic categories of children's associative vocabulary. For example, observing whether children can correctly classify natural and social vocabulary can help discover the characteristics of children's development stage in semantic understanding and classification ability, and improve the lack of systematic analysis of children's vocabulary semantic cognition. Quantifying the strength of vocabulary association and overall evaluation, obtaining the strength of associative vocabulary based on vector space, which provides a powerful tool for accurately measuring the degree of semantic association between words. In the design of intelligent auxiliary communication system, for users with language barriers, the system can recommend semantically closely associated vocabulary to users based on the strength of associative vocabulary to help users express their ideas more accurately. For example, when the user enters "doctor", the system recommends words such as "hospital", "treatment" and "patient" based on the strength of associative vocabulary, which improves the efficiency of communication and solves the problem that traditional systems are difficult to provide accurate vocabulary recommendations based on semantic association. Taking all associative vocabulary strengths to obtain the characteristic information of vocabulary association strength realizes the overall quantitative description of vocabulary association relationship. In the study of brain cognitive function, researchers can compare the characteristic information of vocabulary association strength of people of different age groups (such as teenagers and the elderly) and analyze the changes in brain cognitive ability in vocabulary association. For example, it was found that teenagers have a higher intensity of association when associating some emerging scientific and technological words, while the elderly have a relatively stable intensity of association with traditional life words. This helps to deeply understand the laws of brain cognitive development and overcome the problem of lack of overall quantitative indicators to evaluate vocabulary association relationships. Afterwards, by obtaining the overlap rate of associative words in associative semantic categories, it is of great value in monitoring the stability of brain language function.Taking Alzheimer's patients as an example, as the disease progresses, the overlap rate of associative words in the word association task may gradually decrease. Regular testing of this indicator can assist in early diagnosis and disease monitoring, solving the problem of the difficulty in effectively evaluating changes in the stability of brain language function. The word association feature evaluation value is obtained based on the associative word strength and overlap rate, realizing a comprehensive evaluation of word association features.

[0152] In one embodiment, the step of obtaining the speech comprehensive evaluation value according to the vocabulary reaction feature evaluation value, the semantic feature evaluation value and the vocabulary association feature evaluation value comprises:

[0153] S601, obtaining the vocabulary reaction feature evaluation value;

[0154] S602, obtaining a first weight coefficient according to the vocabulary reaction feature evaluation value;

[0155] S603, obtaining the semantic feature evaluation value;

[0156] S604, obtaining a second weight coefficient according to the semantic feature evaluation value;

[0157] S605, obtaining a vocabulary association feature evaluation value;

[0158] S606, obtaining a third weight coefficient according to the word association feature evaluation value;

[0159] S607: Obtain a comprehensive speech evaluation value according to the vocabulary reaction feature evaluation value, the first weight coefficient, the semantic feature evaluation value, the second weight coefficient, the vocabulary association feature evaluation value and the third weight coefficient.

[0160] As described in the above steps S601-S607, the present invention can clearly reflect the ability level of an individual in terms of vocabulary processing speed, accuracy, and reaction decision-making by obtaining a vocabulary reaction feature evaluation value. When aphasia patients are evaluated for rehabilitation, the evaluation value can accurately record the patient's reaction changes to vocabulary at different rehabilitation stages. For example, the patient may have a long recognition reaction time for simple vocabulary in the early stage. As the rehabilitation training progresses, the reaction time gradually shortens, and the vocabulary reaction feature evaluation value increases accordingly, which provides an intuitive and quantitative basis for rehabilitation therapists to adjust the rehabilitation training plan and solves the problem of being difficult to objectively measure the dynamic changes of vocabulary reaction ability. Obtaining the first weight coefficient according to the vocabulary reaction feature evaluation value realizes the personalized customization of the evaluation system. Taking the children's language development evaluation as an example, for children in the early stage of language learning, the vocabulary reaction speed and accuracy have a greater impact on their overall language ability development, so a higher weight can be given to the vocabulary reaction feature evaluation value. This makes the evaluation results more able to highlight the key abilities of children at this stage, provide targeted guidance for parents and educators, and overcome the drawbacks of fixed weights and lack of flexibility in traditional evaluations. Then, the semantic feature evaluation value is obtained. This process deeply explores the depth of individual understanding of lexical semantics, semantic coherence, and the ability to use vocabulary in semantic networks. In the diagnosis of patients with dyslexia, the semantic feature evaluation value can accurately reflect the patient's grasp of lexical semantics during reading comprehension. For example, patients may have difficulties in understanding the metaphorical meaning and semantic logical relationship of vocabulary, resulting in a low semantic feature evaluation value, which helps professionals accurately locate the patient's semantic understanding obstacles. The second weight coefficient is determined based on the semantic feature evaluation value, which provides precise adaptation for evaluations in different application scenarios. When evaluating the ability to write academic papers, semantic accuracy, richness, and logical coherence are crucial, so a higher weight is given to the semantic feature evaluation value. Such a weight configuration makes the evaluation results more able to reflect the author's ability to control lexical semantics in academic writing, provides a scientific basis for academic review, and improves the situation in which previous evaluations could not reasonably highlight the importance of semantic features according to task requirements. Then, the evaluation value of lexical association features is obtained, which effectively measures the individual's ability in the richness, intensity and flexibility of lexical association. In the evaluation of creative writing ability, the evaluation value of lexical association features can reflect the creator's ability to build a unique semantic network and create novel expressions through lexical association. For example, excellent poets or writers are often able to establish unique and creative associations between words, and their lexical association feature evaluation values ​​are relatively high, which provides a quantitative indicator for the training and selection of creative writing talents. The third weight coefficient is obtained according to the evaluation value of lexical association features, which further optimizes the pertinence of the evaluation system.When evaluating the creativity of advertising copy, the uniqueness, diversity of word associations, and the degree of resonance with the target audience are key factors. Therefore, the reasonable adjustment of the weight of the word association feature evaluation value can make the evaluation results more focused on the creative highlights of the copy, screen out more attractive and influential advertising copy, and solve the problem of lack of dynamic adaptability in the evaluation of the importance of word association features. Finally, the comprehensive speech evaluation value is calculated by combining the various evaluation values ​​and weight coefficients, where the calculation formula is:.

[0161] ;

[0162] Among them, F represents the comprehensive speech evaluation value, R represents the vocabulary response feature evaluation value, M represents the semantic feature evaluation value, N represents the vocabulary association feature evaluation value, a represents the first weight coefficient, b represents the second weight coefficient, and c represents the third weight coefficient, which realizes a comprehensive and objective evaluation of the brain's speech ability. In clinical diagnosis, for patients with language disorders, the comprehensive speech evaluation value can integrate information such as vocabulary response, semantic understanding, and vocabulary association, providing doctors with a comprehensive basis for diagnosis of the disease. For example, in the language assessment of autistic patients, the comprehensive speech evaluation value can comprehensively reflect the patient's overall condition in terms of language expression, understanding, and the use of vocabulary association for communication, helping doctors accurately judge the type and severity of the disorder and formulate personalized treatment plans.

[0163] like Figure 2 As shown, the present invention also provides a brain speech assessment system based on the DSM-V cognitive domain, comprising:

[0164] The first acquisition module 1 is used to acquire the user's voice information;

[0165] A second acquisition module 2 is used to extract features from speech information based on the DSM-V cognitive domain to obtain spoken language feature information, wherein the spoken language feature information includes lexical response feature information, lexical semantic feature information and lexical association feature information;

[0166] A third acquisition module 3 is used to acquire multiple stages of vocabulary feature information according to the vocabulary reaction feature information, and acquire a vocabulary reaction feature evaluation value according to all the stage vocabulary feature information;

[0167] A fourth acquisition module 4 is used to acquire multiple vocabulary node feature information according to the vocabulary semantic feature information, and acquire a semantic feature evaluation value according to each vocabulary node feature information;

[0168] A fifth acquisition module 5, configured to acquire a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and acquire a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information;

[0169] Evaluation module 6 is used to obtain a comprehensive speech evaluation value according to the vocabulary reaction feature evaluation value, the semantic feature evaluation value and the vocabulary association feature evaluation value, and to perform a brain speech evaluation on the user according to the comprehensive speech evaluation value.

[0170] In one embodiment, the second acquisition module 2 includes:

[0171] A first acquisition unit, configured to acquire voice sample information according to the voice information;

[0172] A second acquisition unit, used for acquiring a speech vocabulary reaction time according to the speech sample information;

[0173] A third acquisition unit, configured to extract features from the speech sample information based on the speech vocabulary reaction time to obtain vocabulary reaction feature information;

[0174] A fourth acquisition unit, used to acquire a preset corpus;

[0175] A fifth acquisition unit, configured to divide the speech sample information based on a preset corpus to obtain a plurality of node words;

[0176] A sixth acquisition unit, configured to extract semantic features of each node word to obtain semantic feature information of the node word;

[0177] A seventh acquisition unit, used for collecting the lexical semantic feature information of all the nodes to obtain the lexical semantic feature information;

[0178] An eighth acquisition unit, configured to acquire a plurality of associated words according to the node words;

[0179] A ninth acquisition unit, configured to semantically divide all the associative words to obtain a plurality of semantic categories;

[0180] a tenth acquisition unit, used for acquiring a similar vocabulary ratio of each node vocabulary in the semantic classification;

[0181] an eleventh obtaining unit, configured to obtain the similarity of node words based on the similar word ratio;

[0182] A judging unit, used to judge whether the similarity of the node words is less than a preset value;

[0183] If it is not less than, the node vocabulary is removed;

[0184] If it is less than, the node vocabulary is retained, and the associated features are extracted based on the retained node vocabulary to obtain the vocabulary associated feature information.

[0185] In one embodiment, the third acquisition module 3 includes:

[0186] a twelfth acquisition unit, configured to acquire a plurality of recognition words according to the voice sample information;

[0187] a thirteenth acquiring unit, configured to acquire phoneme feature information according to the recognition vocabulary, wherein the phoneme feature information includes pronunciation duration, pitch value, and sound intensity value;

[0188] A fourteenth acquisition unit, configured to acquire vocabulary accuracy according to the pronunciation duration, pitch value, and sound intensity value;

[0189] A fifteenth acquiring unit, used for acquiring a recognition time interval between each recognition vocabulary;

[0190] a sixteenth acquisition unit, configured to acquire a vocabulary extraction delay time of each recognized vocabulary word according to the recognition time interval;

[0191] A seventeenth acquisition unit, configured to acquire a semantic response decision of each of the recognized words, wherein the semantic response decision includes a semantic extraction time, a semantic response duration, and a semantic response operation time;

[0192] An eighteenth acquisition unit, configured to acquire a vocabulary response value according to the semantic extraction time, the semantic response duration, and the semantic response operation time;

[0193] A vocabulary response feature evaluation value is obtained based on the vocabulary accuracy and the vocabulary response value.

[0194] In one embodiment, the fourth acquisition module 4 includes:

[0195] A nineteenth acquisition unit, configured to acquire the context of each of the node words according to the vocabulary node feature information;

[0196] a twentieth acquisition unit, configured to acquire the semantic coherence of each node word according to the context, and acquire a semantic coherence value according to the semantic coherence of all node words;

[0197] A twenty-first acquisition unit, configured to acquire position information of each of the node words in the semantic hierarchy;

[0198] A twenty-second acquisition unit, configured to acquire the semantic depth of the node vocabulary according to the position information, and acquire the semantic depth value of the vocabulary node according to the semantic depths corresponding to all the semantic depths;

[0199] The twenty-third acquisition unit is used to acquire the semantic feature evaluation value according to the semantic coherence value and the semantic depth value of the vocabulary node.

[0200] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0201] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent results or equivalent process changes made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A brain speech assessment method based on the DSM-V cognitive domain, characterized in that: include: Get the user's voice information; Acquiring voice sample information according to the voice information; Acquire speech vocabulary reaction time according to the speech sample information; Extracting features from the speech sample information based on the speech-vocabulary reaction time to obtain vocabulary reaction feature information; Get the preset corpus; Divide the speech sample information based on the preset corpus to obtain multiple node words; Extracting semantic features of each node word to obtain semantic feature information of the node word; Collecting the lexical semantic feature information of all the nodes to obtain the lexical semantic feature information; Acquire multiple associated words according to the node words; Performing semantic division on all the associative words to obtain multiple semantic categories; Get the similar vocabulary ratio of each node vocabulary in the semantic classification; Acquire the similarity of node vocabulary based on the similar vocabulary ratio; Determine whether the similarity of the node vocabulary is less than a preset value; If it is not less than, the node vocabulary is removed; If it is less than, the node vocabulary is retained, and the associated features are extracted based on the retained node vocabulary to obtain the vocabulary associated feature information; Acquire multiple stages of vocabulary feature information according to the vocabulary response feature information, and acquire vocabulary response feature evaluation values ​​according to all the stages of vocabulary feature information; Acquire multiple vocabulary node feature information according to the vocabulary semantic feature information; Acquire the context of each node vocabulary according to the vocabulary node feature information; Acquire the semantic coherence of each node word according to the context, and acquire the semantic coherence value according to the semantic coherence of all node words; Obtaining position information of each node word in the semantic hierarchy; Obtaining the semantic depth of the node vocabulary according to the position information, and obtaining the semantic depth value of the vocabulary node according to the semantic depths corresponding to all the semantic depths; Obtaining a semantic feature evaluation value according to a semantic coherence value and a semantic depth value of a lexical node; Acquire a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and acquire a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information; A speech comprehensive evaluation value is obtained according to the vocabulary reaction feature evaluation value, the semantic feature evaluation value and the vocabulary association feature evaluation value, and a brain speech evaluation is performed on the user according to the speech comprehensive evaluation value.

2. A method for brain speech assessment based on the DSM-V cognitive domain according to claim 1, characterized in that: The step of acquiring multiple stages of vocabulary feature information according to the vocabulary reaction feature information, and acquiring vocabulary reaction feature evaluation values ​​according to all the stage vocabulary feature information, comprises: Acquire multiple recognition words according to the voice sample information; Acquiring phoneme feature information according to the recognition vocabulary, wherein the phoneme feature information includes pronunciation duration, pitch value, and sound intensity value; Acquiring vocabulary accuracy according to the pronunciation duration, pitch value, and sound intensity value; Get the recognition time interval between each recognition word; Acquire a vocabulary extraction delay time of each recognized vocabulary word according to the recognition time interval; Obtaining a semantic response decision for each of the recognized words, wherein the semantic response decision includes a semantic extraction time, a semantic response duration, and a semantic response operation time; Acquire a vocabulary response value according to the semantic extraction time, the semantic response duration, and the semantic response operation time; A vocabulary response feature evaluation value is obtained based on the vocabulary accuracy and the vocabulary response value.

3. A method for brain speech assessment based on the DSM-V cognitive domain according to claim 1, characterized in that: The step of obtaining a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and obtaining a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information comprises: Acquire vocabulary association features according to the vocabulary association feature information; Acquire multiple associative words based on the word association features; Performing semantic classification on all the associative words to obtain a plurality of associative semantic categories; Obtaining the associative lexical strength of the associative lexical items in each associative semantic category based on the vector space; Acquire all the associative word strengths to obtain word associative strength feature information; Obtaining the overlap rate of each of the associative words in the associative semantic category; The word association feature evaluation value is obtained according to the associative word strength and the overlap rate.

4. The method for brain speech assessment based on the DSM-V cognitive domain according to claim 1, characterized in that: The step of obtaining the comprehensive speech evaluation value according to the lexical reaction feature evaluation value, the semantic feature evaluation value and the lexical association feature evaluation value comprises: Obtaining the vocabulary response feature evaluation value; Acquire a first weight coefficient according to the vocabulary reaction feature evaluation value; Obtaining the semantic feature evaluation value; Acquire a second weight coefficient according to the semantic feature evaluation value; Obtaining word association feature evaluation values; Acquire a third weight coefficient according to the word association feature evaluation value; The speech comprehensive evaluation value is obtained according to the vocabulary reaction feature evaluation value, the first weight coefficient, the semantic feature evaluation value, the second weight coefficient, the vocabulary association feature evaluation value and the third weight coefficient.

5. A brain speech assessment system based on the DSM-V cognitive domain, characterized in that: include: The first acquisition module is used to acquire the user's voice information; A second acquisition module is used to extract features from speech information based on the DSM-V cognitive domain to obtain spoken language feature information, wherein the spoken language feature information includes lexical response feature information, lexical semantic feature information and lexical association feature information; A third acquisition module is used to acquire multiple stages of vocabulary feature information according to the vocabulary reaction feature information, and acquire vocabulary reaction feature evaluation values ​​according to all the stage vocabulary feature information; A fourth acquisition module, configured to acquire a plurality of vocabulary node feature information according to the vocabulary semantic feature information, and acquire a semantic feature evaluation value according to each of the vocabulary node feature information; A fifth acquisition module, configured to acquire a plurality of vocabulary association strength feature information according to the vocabulary association feature information, and acquire a vocabulary association feature evaluation value according to the plurality of vocabulary association strength feature information; An evaluation module, used to obtain a comprehensive speech evaluation value according to the lexical response feature evaluation value, the semantic feature evaluation value and the lexical association feature evaluation value, and to perform a brain speech evaluation on the user according to the comprehensive speech evaluation value; Wherein, the second acquisition module includes: A first acquisition unit, configured to acquire voice sample information according to the voice information; A second acquisition unit, used for acquiring a speech vocabulary reaction time according to the speech sample information; A third acquisition unit, configured to extract features from the speech sample information based on the speech vocabulary reaction time to obtain vocabulary reaction feature information; A fourth acquisition unit, used to acquire a preset corpus; A fifth acquisition unit, configured to divide the speech sample information based on a preset corpus to obtain a plurality of node words; A sixth acquisition unit, configured to extract semantic features of each node word to obtain semantic feature information of the node word; A seventh acquisition unit, used for collecting the lexical semantic feature information of all the nodes to obtain the lexical semantic feature information; An eighth acquisition unit, configured to acquire a plurality of associated words according to the node words; A ninth acquisition unit, configured to semantically divide all the associative words to obtain a plurality of semantic categories; a tenth acquisition unit, used for acquiring a similar vocabulary ratio of each node vocabulary in the semantic classification; an eleventh obtaining unit, configured to obtain the similarity of node words based on the similar word ratio; A judging unit, used to judge whether the similarity of the node words is less than a preset value; If it is not less than, the node vocabulary is removed; If it is less than, the node vocabulary is retained, and the associated features are extracted based on the retained node vocabulary to obtain the vocabulary associated feature information; Wherein, the fourth acquisition module includes: A nineteenth acquisition unit, configured to acquire the context of each of the node words according to the vocabulary node feature information; a twentieth acquisition unit, configured to acquire the semantic coherence of each node word according to the context, and acquire a semantic coherence value according to the semantic coherence of all node words; A twenty-first acquisition unit, configured to acquire position information of each of the node words in the semantic hierarchy; A twenty-second acquisition unit, configured to acquire the semantic depth of the node vocabulary according to the position information, and acquire the semantic depth value of the vocabulary node according to the semantic depths corresponding to all the semantic depths; The twenty-third acquisition unit is used to acquire the semantic feature evaluation value according to the semantic coherence value and the semantic depth value of the vocabulary node.

6. A brain speech assessment system based on the DSM-V cognitive domain according to claim 5, characterized in that: The third acquisition module includes: a twelfth acquisition unit, configured to acquire a plurality of recognition words according to the voice sample information; a thirteenth acquiring unit, configured to acquire phoneme feature information according to the recognition vocabulary, wherein the phoneme feature information includes pronunciation duration, pitch value, and sound intensity value; A fourteenth acquisition unit, configured to acquire vocabulary accuracy according to the pronunciation duration, pitch value, and sound intensity value; A fifteenth acquiring unit, used for acquiring a recognition time interval between each recognition vocabulary; a sixteenth acquisition unit, configured to acquire a vocabulary extraction delay time of each recognized vocabulary word according to the recognition time interval; A seventeenth acquisition unit, configured to acquire a semantic response decision of each of the recognized words, wherein the semantic response decision includes a semantic extraction time, a semantic response duration, and a semantic response operation time; An eighteenth acquisition unit, configured to acquire a vocabulary response value according to the semantic extraction time, the semantic response duration, and the semantic response operation time; A vocabulary response feature evaluation value is obtained based on the vocabulary accuracy and the vocabulary response value.

Images