Scoring method, scoring device, computer equipment and storage medium

文档序号：1965045 发布日期：2021-12-14 浏览：15次中文

阅读说明：本技术 评分方法、装置、计算机设备和存储介质 (Scoring method, scoring device, computer equipment and storage medium ) 是由区正辉张九龙贾亚龙李锋于 2021-08-27 设计创作，主要内容包括：本申请涉及一种评分方法、装置、计算机设备和存储介质。所述方法包括：通过响应于目标问题对应的回答指令,获取目标用户针对目标问题的回答信息、以及获取目标问题对应的多个评价维度的评价参数,可以预先配置多个评价维度的评价参数,降低进行评分处理的时延。通过根据多个评价维度的评价参数分别对回答信息进行评分,得到评分反馈信息,评分反馈信息包括评分分数信息以及评分分数对应的评分含义信息；输出评分分数信息以及评分分数对应的评分含义信息,可以在短时间内为用户提供高准确度的评分分数以及评分含义,供用户学习以及改进。(The application relates to a scoring method, a scoring device, computer equipment and a storage medium. The method comprises the following steps: by responding to the answer instruction corresponding to the target question, acquiring the answer information of the target user for the target question and acquiring the evaluation parameters of the multiple evaluation dimensions corresponding to the target question, the evaluation parameters of the multiple evaluation dimensions can be configured in advance, and the time delay of scoring processing is reduced. Scoring the answer information according to the evaluation parameters of the evaluation dimensions to obtain scoring feedback information, wherein the scoring feedback information comprises scoring score information and scoring meaning information corresponding to the scoring scores; the scoring score information and the scoring meaning information corresponding to the scoring score are output, and the scoring score and the scoring meaning with high accuracy can be provided for the user in a short time for the user to learn and improve.)

1. A scoring method, the method comprising:

responding to an answer instruction corresponding to a target question, acquiring answer information of the target user for the target question, and acquiring evaluation parameters of a plurality of evaluation dimensions corresponding to the target question;

scoring the answer information according to the evaluation parameters of the evaluation dimensions to obtain scoring feedback information, wherein the scoring feedback information comprises scoring score information and scoring meaning information corresponding to the scoring scores;

and outputting the scoring score information and scoring meaning information corresponding to the scoring score.

2. The method of claim 1, wherein the plurality of evaluation dimensions comprise at least one or more of content standardization, completeness, fluency, and speech rate;

the scoring the answer information according to the evaluation parameters of the evaluation dimensions to obtain scoring feedback information includes:

performing standard word matching in the answer information through a preset content standard degree matching algorithm, and determining a content standard degree score of the answer information according to a matching result;

counting the text length of the answer information through a preset integrity counting algorithm, and determining the integrity score of the answer information according to the counting result;

counting the number of preset tone words in the answer information through a preset fluency statistical algorithm, and determining the fluency score of the answer information according to the statistical result;

calculating the speed of speech of the answer information through a preset speech speed determining algorithm, and obtaining the speed of speech score of the answer information according to the speed of speech;

and analyzing the content standard degree score, the integrity degree score, the fluency score and the speech speed score to obtain score feedback information.

3. The method according to claim 2, wherein the evaluation parameters of the content standardization level comprise a preset standard quantity threshold value and standard words corresponding to the target question;

the matching of the standard words in the answer information by a preset content standard degree matching algorithm and the determination of the content standard degree score of the answer information according to the matching result comprise:

matching the standard words in the answer information through a preset content standard degree matching algorithm, and counting the number of matched standard words of the answer information;

and determining the content standard degree score of the answer information according to the standard words and the preset standard quantity threshold.

4. The method according to claim 3, wherein the matching the standard words in the answer information by a preset content standard degree matching algorithm and counting the number of matched standard words of the answer information comprises:

converting the answer information into a pinyin sequence, and performing voice recognition correction on the pinyin sequence of the answer information through a preset proper noun dictionary to obtain corrected answer information;

generating a regular expression of the standard words through a preset regular expression generation algorithm;

and matching the standard words in the corrected answer information according to the regular expression, and counting the matching number of the first standard words.

5. The method according to claim 3, wherein the evaluation parameter of the content standardization further comprises a preset standard character length threshold;

the matching of the standard words in the answer information through a preset content standard degree matching algorithm and the statistics of the number of matched standard words of the answer information comprise:

if the character length of the standard word is smaller than the preset standard character length threshold, generating a regular expression of the standard word through a preset regular expression generation algorithm;

and matching the standard words in the answer information according to the regular expression, and counting the matching number of the second standard words.

6. The method according to claim 3, wherein the evaluation parameters of the content standardization further comprise a preset standard character length threshold value and a standard editing distance threshold value;

if the character length of the standard word is larger than or equal to the preset standard character length threshold value, performing word segmentation processing on the standard word to obtain a plurality of standard word segments;

respectively extracting character strings containing the standard word segmentation from the answer information through the preset matching window to form a matching candidate set, wherein the length of the preset matching window is the character length of the standard word;

and calculating the editing distance between the standard words and each character string in the matching candidate set, determining that the standard words are successfully matched when the editing distance is smaller than or equal to the standard editing distance threshold, and counting the matching number of third standard words.

7. The method according to claim 3, wherein the evaluation parameter of the content standardization further comprises a preset standard character length threshold;

if the character length of the standard word is smaller than the preset standard character length threshold, the answer information and the standard word are respectively converted into an answer information pinyin sequence and a standard word pinyin sequence;

generating a regular expression of the standard word pinyin sequence by a preset regular expression generation algorithm;

and matching in the answer information pinyin sequence according to the regular expression of the standard word pinyin sequence, and counting the matching number of the fourth standard word.

8. The method according to claim 3, wherein the evaluation parameters of the content standardization further comprise a preset standard character length threshold value and a standard editing distance threshold value;

if the number of the characters of the standard word is larger than or equal to the preset standard character length threshold, performing word segmentation processing on the standard word to obtain a plurality of standard word segments;

converting the answer information and the standard word segmentation into an answer information pinyin sequence and a standard word segmentation pinyin sequence respectively;

extracting character strings containing the pinyin sequence of the participle of the standard word from the pinyin sequence of the answer information through the preset matching window to form a matching candidate set, wherein the length of the preset matching window is the number of characters contained in the pinyin sequence of the standard word;

and calculating the edit distance between the pinyin sequence of the standard word and each character string in the matching candidate set, determining that the standard word is successfully matched when the edit distance is less than or equal to the standard edit distance threshold, and counting the matching quantity of a fifth standard word.

9. The method of any one of claims 2 to 8, wherein the content criteria includes at least one of keyword hit level, compliance level, emotion level, and usage level of conversational words.

10. The method according to claim 3, wherein before the step of matching the standard words in the answer information by a preset content standard degree matching algorithm and counting the number of matched standard words of the answer information, the method further comprises:

recognizing and eliminating punctuations in the answer information through a preset punctuation dictionary to obtain the answer information after the punctuations are eliminated; and/or converting the numerical values in the answer information into Chinese character expressions to obtain the converted answer information.

11. The method of claim 2, wherein the evaluation parameters of the completeness comprise a total question character length threshold value and a sub-question character length threshold value, and the answer information comprises a plurality of sub-question answer texts;

the step of counting the text length of the answer information by a preset integrity counting algorithm to obtain the integrity score of the answer information comprises the following steps:

if the character length of the answer information is larger than or equal to the total question character length threshold value, determining a first score according to a numerical value of a preset total length score;

if the character length of the sub-question answer text is larger than or equal to the sub-question character length threshold value, determining a second score according to a numerical value of a preset sub-length score;

determining a third score for the plurality of sub-question answer texts according to the content standardization score;

and taking the sum of the first score, the second score and the third score as a completeness score of the answer information.

12. A scoring device, the device comprising:

the acquisition module is used for responding to an answer instruction corresponding to a target question, acquiring answer information of the target user aiming at the target question and acquiring evaluation parameters of the target question for a plurality of evaluation dimensions;

the scoring module is used for scoring the answer information according to the evaluation parameters of the evaluation dimensions to obtain scoring feedback information, and the scoring feedback information comprises scoring score information and scoring meaning information corresponding to the scoring scores;

and the output module is used for outputting the scoring score information and scoring meaning information corresponding to the scoring score.

13. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor realizes the steps of the method of any one of claims 1 to 11 when executing the computer program.

14. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 11.

Technical Field

The present application relates to the field of natural language processing technologies, and in particular, to a scoring method, an apparatus, a computer device, and a storage medium.

Background

With the development of technology, more and more work can be done by machines. When a text input by a user is scored, a random forest model is used as a scorer in the related technology, the target feature of the text is extracted according to a preset feature extraction mode, and the target feature value is input into a preset random forest model which is trained in advance to be scored so as to obtain the score corresponding to the text. However, this method can only obtain the overall score of the text, and no scoring reason is given, so that the score lacks resolvable performance, and the user is difficult to improve according to the score.

Disclosure of Invention

In view of the above, there is a need to provide a scoring method, apparatus, computer device and storage medium that can save a lot of labor cost and provide scoring details in multiple evaluation dimensions to help users to improve.

A scoring method, the method comprising:

and outputting the scoring score information and scoring meaning information corresponding to the scoring score.

In one embodiment, the plurality of evaluation dimensions at least comprise one or more of content standard degree, completeness, fluency, and speech speed;

the scoring the answer information from a plurality of evaluation dimensions through the evaluation parameters to obtain scoring feedback information includes:

counting the text length of the answer information through a preset integrity counting algorithm, and determining the integrity score of the answer information according to the counting result;

and analyzing the content standard degree score, the integrity degree score, the fluency score and the speech speed score to obtain scoring feedback information.

In one embodiment, the evaluation parameters of the content standardization level include a preset standard quantity threshold and standard words corresponding to the target question;

matching the standard words in the answer information through a preset content standard degree matching algorithm, and counting the number of matched standard words of the answer information;

and determining the content standard degree score of the answer information according to the standard words and the preset standard quantity threshold.

In one embodiment, the matching the standard words in the answer information by a preset content standard degree matching algorithm, and counting the number of matched standard words of the answer information includes:

generating a regular expression of the standard words through a preset regular expression generation algorithm;

and matching the standard words in the corrected answer information according to the regular expression, and counting the matching number of the first standard words.

In one embodiment, the evaluation parameter of the content standardization further comprises a preset standard character length threshold;

and matching the standard words in the answer information according to the regular expression, and counting the matching number of the second standard words.

In one embodiment, the evaluation parameters of the content standardization further include a preset standard character length threshold and a standard editing distance threshold;

In one embodiment, the evaluation parameter of the content standardization further comprises a preset standard character length threshold;

generating a regular expression of the standard word pinyin sequence by a preset regular expression generation algorithm;

and matching in the answer information pinyin sequence according to the regular expression of the standard word pinyin sequence, and counting the matching number of the fourth standard word.

In one embodiment, the evaluation parameters of the content standardization further include a preset standard character length threshold and a standard editing distance threshold;

converting the answer information and the standard word segmentation into an answer information pinyin sequence and a standard word segmentation pinyin sequence respectively;

In one embodiment, the content standard degree includes at least one of keyword hit degree, compliance degree, emotion degree and usage degree of words.

In one embodiment, before the step of matching the standard words in the answer information by a preset content standard degree matching algorithm and counting the number of matched standard words of the answer information, the method further includes:

In one embodiment, the evaluation parameters of the integrity comprise a total question character length threshold value and a sub-question character length threshold value, and the answer information comprises a plurality of sub-question answer texts;

the step of counting the text length of the answer information by a preset integrity counting algorithm to obtain the integrity score of the answer information comprises the following steps:

determining a third score for the plurality of sub-question answer texts according to the content standardization score;

and taking the sum of the first score, the second score and the third score as a completeness score of the answer information.

A scoring device, the device comprising:

and the output module is used for outputting the scoring score information and scoring meaning information corresponding to the scoring score.

A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:

outputting the scoring score information and scoring meaning information corresponding to the scoring score

A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:

outputting the scoring score information and scoring meaning information corresponding to the scoring score

According to the scoring method, the scoring device, the computer equipment and the storage medium, the answer information of the target user aiming at the target question is obtained by responding to the answer instruction corresponding to the target question, the evaluation parameters of a plurality of evaluation dimensions corresponding to the target question are obtained, the evaluation parameters of the plurality of evaluation dimensions can be configured in advance, and the time delay of scoring processing is reduced. Scoring the answer information according to the evaluation parameters of the evaluation dimensions to obtain scoring feedback information, wherein the scoring feedback information comprises scoring score information and scoring meaning information corresponding to the scoring scores; the scoring score information and the scoring meaning information corresponding to the scoring score are output, and the scoring score and the scoring meaning with high accuracy can be provided for the user in a short time for the user to learn and improve.

Drawings

FIG. 1 is a schematic flow chart of a scoring method in one embodiment;

FIG. 2 is a schematic flow chart illustrating the steps of determining scoring feedback information in one embodiment;

FIG. 3 is a flowchart illustrating the step of determining a content normalization score in one embodiment;

FIG. 4 is a flowchart illustrating the step of counting the number of matches of the first criterion word in one embodiment;

FIG. 5 is a flowchart illustrating the step of counting the number of second criterion word matches in one embodiment;

FIG. 6 is a flowchart illustrating the step of counting the number of matches of the third criterion word in one embodiment;

FIG. 7 is a flowchart illustrating the step of counting the number of matches of the fourth criterion word in one embodiment;

FIG. 8 is a flowchart illustrating the step of counting the number of matches of the fifth criterion word in one embodiment;

FIG. 9 is a flowchart illustrating the step of determining the integrity score in one embodiment;

FIG. 10 is a schematic diagram of an online dialogue training service system;

FIG. 11 is a block diagram showing the structure of a scoring device according to an embodiment;

FIG. 12 is a diagram illustrating an internal structure of a computer device according to an embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

In one embodiment, as shown in fig. 1, a scoring method is provided, which is exemplified by applying the method to a scoring device, it is to be understood that the method may also be applied to a server, and may also be applied to a system including a scoring device and a server, and is implemented by interaction between the scoring device and the server, where the scoring device may be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices, and the server may be implemented by an independent server or a server cluster formed by a plurality of servers. In this embodiment, the scoring method includes the following steps:

step 101, in response to an answer instruction corresponding to the target question, acquiring answer information of the target user for the target question and acquiring evaluation parameters of a plurality of evaluation dimensions corresponding to the target question.

The target question and the evaluation parameters of the plurality of evaluation dimensions corresponding to the target question may be pre-stored in a database in the scoring device, and the evaluation parameters of different evaluation dimensions are different from each other. The scoring device can determine the selection number of the multiple evaluation dimensions and which evaluation dimensions are specifically selected according to the actual requirements of the target user.

Specifically, the scoring device outputs a target question to a target user in response to a target question acquisition operation of the target user. In this way, the target user can acquire the target question and then perform an answering operation. And the scoring equipment responds to the answer instruction corresponding to the target question and acquires the answer information generated by the target user aiming at the target question. Meanwhile, the scoring device can also extract evaluation parameters of a plurality of evaluation dimensions corresponding to the target problem from a preset database.

Optionally, the answer information, which is acquired by the scoring device and generated by the target user for the target question, may be in a text information form or a voice information form. When the answer information is in a voice form, the scoring device needs to convert the answer information in the voice form into answer information in a text form according to a preset voice conversion algorithm.

And 102, scoring the answer information according to the evaluation parameters of the plurality of evaluation dimensions to obtain scoring feedback information, wherein the scoring feedback information comprises scoring score information and scoring meaning information corresponding to the scoring scores.

Specifically, the scoring device may score the answer information in the multiple evaluation dimensions respectively through a preset scoring algorithm corresponding to the multiple evaluation dimensions respectively and evaluation parameters of the multiple evaluation dimensions, and obtain corresponding scoring scores and scoring meaning information represented by the scoring scores.

And 103, outputting scoring score information and scoring meaning information corresponding to the scoring scores.

Specifically, the scoring device provides examples of at least two ways of generating scoring feedback information.

One possible example may be that the scoring device outputs the scoring score information of each evaluation dimension and the scoring meaning corresponding to the scoring score respectively for reference by the target user. For example, the plurality of evaluation dimensions may include completeness, fluency, and speech rate. In this case, the scoring device generates evaluation feedback information according to the completeness score, the fluency score, and the speech rate score. For example, the evaluation score information may be: integrity score 90, fluency score 100, pace score 10. Correspondingly, the scoring meaning information corresponding to the integrity score can be enough sections of answer information of the user and is information related to the answer target question; the scoring meaning information corresponding to the fluency score can be that the user is very smooth in answering operation and has no stumbling; the score meaning information corresponding to the speech rate score 10 may be that the speech rate is slow when the user performs the response operation. The user may make improvements based on the scoring feedback information.

Another possible example may be that the scoring device performs weighted calculation on the scoring score information of the multiple evaluation dimensions according to a preset weighting algorithm to obtain a total score of the answer information of the target user for the target question and scoring meaning information corresponding to the total score. Specifically, the scoring meaning information corresponding to the total score can comprehensively represent the scoring condition of the target user in each evaluation dimension.

For example, the plurality of evaluation dimensions may include completeness, fluency, and speech rate. The scoring device may pre-configure the weight of each evaluation dimension in the preset weighting algorithm, for example, the weight of the integrity is 0.8, the weight of the fluency is 0.1, and the weight of the speech rate is 0.1. The scoring device may calculate a total score (10) according to scores (a completeness score of 10, a fluency score of 10, a pace score of 10) of each evaluation dimension and a preset weight. In this case, the score meaning information corresponding to the total score may be that the answer information of the user is insufficient, most of the contents are information irrelevant to the answer target question, and the user is unsmooth and the speech speed is not moderate when performing the answer operation.

According to the scoring method, the answer information of the target user aiming at the target question is obtained by responding to the answer instruction corresponding to the target question, and the evaluation parameters of the plurality of evaluation dimensions corresponding to the target question are obtained, so that the evaluation parameters of the plurality of evaluation dimensions can be configured in the scoring equipment in advance, and the time delay of scoring processing is reduced. Scoring the answer information according to the evaluation parameters of the evaluation dimensions to obtain scoring feedback information, wherein the scoring feedback information comprises scoring score information and scoring meaning information corresponding to the scoring scores; the scoring score information and the scoring meaning information corresponding to the scoring score are output, and the scoring score and the scoring meaning with high accuracy can be provided for the user in a short time for the user to learn and improve.

In one embodiment, the evaluation dimension may include one or more evaluation dimensions, in which case the evaluation dimension may include at least one or more of content standardization, completeness, fluency, and speech rate. Accordingly, as shown in fig. 2, a specific process of step 102 "scoring the answer information according to the evaluation parameters of the multiple evaluation dimensions to obtain the scoring feedback information" includes:

step 201, performing standard word matching in the answer information through a preset content standard degree matching algorithm, and determining the content standard degree score of the answer information according to the matching result.

Specifically, in the case where the evaluation dimension is the content standardization level, the scoring device may score the answer information in the evaluation dimension of the content standardization level by a preset content standardization level matching algorithm. And the scoring equipment matches the standard words in the answer information through a preset content standard degree matching algorithm, and determines a content standard degree score according to a matching result output by the preset content standard degree matching algorithm.

Alternatively, the standard words may be answer keywords pre-configured according to actual needs of the target user and the target question. The standard words may also be violation words that are pre-configured according to the actual needs of the target user, such as words with insulting meanings. The standard words may also be emotion words that are pre-configured according to the actual needs of the target user, such as words with negative emotions. The standard words may also be words pre-configured according to the actual needs of the target users, for example, words that can achieve sales promotion effects.

Step 202, counting the text length of the answer information through a preset integrity counting algorithm, and determining the integrity score of the answer information according to the counting result.

Specifically, the scoring algorithm used by the scoring device in calculating the integrity score of the answer information may be a preset integrity statistic algorithm. The scoring device can count the text length of the answer information through a preset integrity statistical algorithm, and determine the integrity score of the answer information according to a statistical result output by the preset integrity statistical algorithm.

And 203, counting the number of preset tone words in the answer information through a preset fluency statistical algorithm, and determining the fluency score of the answer information according to the statistical result.

Specifically, in the case where the evaluation dimension is fluency, the scoring device may determine that the scoring algorithm that scores the answer information is a preset fluency statistical algorithm. And the scoring device is used for counting the occurrence times of voice activity words (such as forehead, kayak and hiccup) in the answer information through a preset fluency statistical algorithm, and determining fluency score according to the statistical result output by the preset fluency statistical algorithm.

And 204, calculating the speed of speech of the answer information through a preset speech speed determining algorithm, and obtaining the speed of speech score of the answer information according to the speed of speech.

Specifically, in the case where the evaluation dimension is the speech rate, the scoring device may determine that the scoring algorithm that scores the answer information is a preset speech rate calculation algorithm. And the scoring equipment calculates the speech rate of the answer information according to the text length and the answer time of the answer information by a preset speech rate calculation algorithm, and determines the speech rate score according to the calculated speech rate.

Step 205, analyzing the content standard degree score, integrity degree score, fluency score and speed score to obtain the scoring feedback information.

Specifically, the scoring device may determine meanings of the content standardization score, the completeness score, the fluency score, and the speech rate score according to the obtained content standardization score, the completeness score, the fluency score, and the speech rate score, and then generate scoring feedback information.

It should be further noted that, in the embodiment of the present invention, none of the step 201, the step 202, the step 203, the step 204, and the step 205 needs to distinguish the execution sequence. The execution sequence of the above steps is not specifically limited in the embodiment of the present invention, and those skilled in the art can determine the execution sequence according to the actual application scenario.

In this embodiment, the content standardization score, the completeness score, the fluency score, and the speech rate score are obtained by scoring the evaluation dimensions of the content standardization, the completeness, the fluency, and the speech rate, so that the obtained scores have resolvable performance, a quantifiable scoring index is given, and reference and learning of a target user are facilitated.

In one embodiment, the evaluation parameters of the content standardization level include a preset standard quantity threshold and standard words corresponding to the target questions; in this case, as shown in fig. 3, the specific processing procedure of "performing standard word matching in the answer information by using the preset standard degree matching algorithm, and determining the content standard degree score of the answer information according to the matching result" in step 201 includes:

step 301, performing standard word matching in the answer information through a preset content standard degree matching algorithm, and counting the number of matched standard words in the answer information.

Specifically, the scoring device performs matching of the standard words in the answer information through a preset content standard degree matching algorithm, that is, searches and screens the standard words in the answer information, and counts the number of the successfully matched standard words in the answer information.

Step 302, obtaining a content standard degree score of the answer information according to the matching number of the standard words and a preset standard number threshold.

Specifically, the preset standard quantity threshold is the number of times that a standard word should appear in the preset answer information. If the number of matched standard words obtained by performing step 301 is greater than or equal to the preset standard number threshold, the scoring device may determine that the content standard degree score is a preset full score (e.g., 10 points, 100 points). If the matching number of the standard words is smaller than the threshold value of the preset standard number, in this case, the scoring device determines the corresponding content standardization score according to the ratio of the matching number of the standard words to the threshold value of the preset standard number.

Alternatively, the number of criterion words in the target answer information corresponding to the target question may be the target number, and the evaluation parameter of the content criterion degree may further include a preset threshold. In this case, the scoring device may obtain the preset standard number threshold according to a ratio of the first number to a preset threshold.

Alternatively, the content standardization score may be calculated by the following formula:

A＝min(B×(C÷D÷E)，C),

wherein, A is the content standard degree score, B is the standard word matching number, C is the preset full score, D is the target number, E is the preset threshold value, and D/E is the preset standard number threshold value.

In one embodiment, if the scoring device determines that the number of matching standard words in the answer information has reached the preset threshold, that is, the number of matching standard words in the answer information has become greater than or equal to the preset threshold, in this case, the scoring device may assign a preset full score to the answer information in the evaluation dimension of the content standardization. For example, the preset threshold may be 0.8, in which case, when the scoring device confirms that the matching number of the standard words in the answer information has reached eighty percent of the target number, a preset full score may be assigned to the answer information in the evaluation dimension of the content standardization.

In this embodiment, by comparing the matching number of the standard words with the preset threshold of the number of the standards, the scoring device may assign a high-precision score to the answer information in the evaluation dimension of the content standardization.

In one embodiment, as shown in fig. 4, the specific processing procedure of "performing standard word matching in the answer information by using a preset standard degree matching algorithm, and determining a content standard degree score of the answer information according to the matching result" in step 301 includes:

step 401, converting the answer information into a pinyin sequence, and performing voice recognition correction on the pinyin sequence of the answer information through a preset proper noun dictionary to obtain corrected answer information.

The answer information may be a character text sequence, and the preset proper noun dictionary may be a proper noun dictionary collected in advance in a situation adapted to an actual application scenario. A proper noun may be a word having a particular standard meaning, such as a word containing a particular name (e.g., XX Bank, XX fund, etc.).

Specifically, the scoring device converts the acquired answer information in the form of a character text sequence into a sequence in the form of pinyin. In this way, the scoring device can filter proper nouns in the answer information according to a preset proper noun dictionary and correct the answer information according to the filtering result so as to obtain corrected answer information. The screening result comprises pinyin sequences of a plurality of words, and the edit distance between the pinyin sequence of each word and a proper noun in a preset proper noun dictionary conforms to the preset correction edit distance.

For example, the target answer information corresponding to the target question may be "XX bank and XX museum deep collaboration", where the proper terms are XX bank and XX museum. The answer information of the target user input acquired by the scoring device may be "XX bank collaborating with XX joint depth". The scoring equipment screens proper nouns in the answer information according to a preset proper noun dictionary, namely, the pinyin sequence of the answer information is aligned with the pinyin sequence of the proper nouns, namely, the pinyin sequence of 'national museum' is aligned with the pinyin sequence of 'national museum', and the editing distance between the pinyin sequence of the answer information and the pinyin sequence of the proper nouns is calculated. If the calculated editing distance is smaller than or equal to the preset correction editing distance, the scoring equipment rewrites proper nouns in the answer information according to proper nouns in a preset proper noun dictionary to obtain corrected answer information, namely 'deep cooperation between bank A and national museum'.

Alternatively, the edit distance may be calculated by the following formula:

where i, j represent the subscripts of string a and string b, respectively, beginning with 1. A [ i ], B [ j ] represent the characters corresponding to the subscript positions of the character strings. d (i, j) represents the edit distance between the substring of the character string a from the beginning to A [ i ] and the substring of the character string B from the beginning to B [ j ].

Step 402, generating a regular expression of the standard word by a preset regular expression generation algorithm.

Specifically, the scoring device may generate the regular expressions for the matched standard words through a preset regular expression generation algorithm.

And 403, performing standard word matching in the corrected answer information according to the regular expression, and counting the matching number of the first standard words.

Specifically, the scoring device matches each standard word in the answer information respectively through the generated regular expressions of the plurality of standard words, counts the occurrence times of the standard words in the answer information, and determines the matching number of the first standard words according to the statistical result.

In the embodiment, the answer information is preprocessed through the preset proper noun dictionary, wrongly written information of the answer information can be filtered in time, and the accuracy of the scoring equipment in scoring is improved.

In one embodiment, as shown in fig. 5, the specific processing procedure of "performing standard word matching in the answer information by using a preset standard degree matching algorithm, and determining a content standard degree score of the answer information according to the matching result" in step 301 includes:

step 501, if the character length of the standard word is smaller than a preset standard character length threshold, generating a regular expression of the standard word by a preset regular expression generation algorithm.

Specifically, the preset standard character length threshold is a standard word length threshold configured in advance by the scoring device. If the scoring device judges that the character length of the standard word needing to be subjected to standard word matching at present is smaller than the preset standard character length threshold value, the scoring device can generate a regular expression for matching according to a preset regular expression generation algorithm.

Step 502, according to the regular expression, performing standard word matching in the answer information, and counting the matching number of the second standard words.

Specifically, the scoring device matches the standard words in the answer information through the generated regular expressions of the standard words, counts the occurrence times of the standard words in the answer information, and determines the matching number of the second standard words according to the statistical result.

In the embodiment, the standard words are classified in advance through the character lengths of the standard words, so that the energy consumption of the scoring equipment can be reduced, and the computing capability of the scoring equipment is improved.

In one embodiment, the character length of the standard word has multiple conditions, in which case, the evaluation parameter of the content standardization further includes a preset standard character length threshold and a standard edit distance threshold. As shown in fig. 6, the specific processing procedure of "performing standard word matching in the answer information by using a preset standard degree matching algorithm, and determining a content standard degree score of the answer information according to a matching result" in step 301 includes:

step 601, if the character length of the standard word is greater than or equal to a preset standard character length threshold, performing word segmentation processing on the standard word to obtain a plurality of standard word segments.

Specifically, when the character length of the standard word is greater than or equal to the preset standard character length threshold, the scoring device may perform word segmentation processing on the standard word according to a preset word segmentation algorithm to obtain a plurality of standard word segments. The number of the standard word segments can be determined according to the character length of the standard word.

For example, when the standard word is "Bo-Shi fund", the scoring device performs word segmentation on the standard word according to a preset word segmentation algorithm, and the obtained multiple standard word segments may be "Bo, Shi, Bas, and jin".

Step 602, extracting character strings containing the standard word segmentation from the answer information respectively through a preset matching window to form a matching candidate set, wherein the length of the preset matching window is the character length of the standard word.

Specifically, the scoring device takes the character length of the standard word as the length of a preset matching window, and in this case, the scoring device extracts a plurality of character strings containing the standard word segments in the answer information through the preset matching window. In this way, the length of the plurality of character strings extracted by the scoring device is the length of the preset matching window, and the plurality of extracted character strings are taken as a matching candidate set.

For example, the standard word may be "Boji fund" and the answer message may be "product to select doctor fund". In this case, the scoring device may segment the word "doctor" by the first standard word, extract a plurality of character strings including "doctor" in the answer information, and compose a matching candidate set, i.e., "select doctor, select doctor base, and doctor fund".

Step 603, calculating the editing distance between the standard words and each character string in the matching candidate set, and when the editing distance is smaller than or equal to the standard editing distance threshold, determining that the standard words are successfully matched to obtain the matching number of the third standard words.

Specifically, the scoring device may calculate the edit distance of each character string in the matching candidate set from the standard word, respectively. If the edit distance corresponding to the character string in the matching candidate set is smaller than or equal to the standard edit distance threshold, the scoring device may determine that the standard word is successfully matched, and obtain the third standard word matching number. If the edit distances corresponding to the character strings in the matching candidate set are all greater than the standard edit distance threshold, the scoring device may determine that the standard word segmentation matching fails, in this case, the scoring device may match the answer information again through the second standard word segmentation, and execute the processes of step 602 to step 603 until the standard word matching succeeds or the standard word matching fails.

For example, the scoring device calculates edit distances of the respective character strings "select doctor", "select doctor base", "doctor fund" in the matching candidate set from the standard word "bock fund", and K2< K1< K3, respectively. And the scoring device judges that the K2 is smaller than a preset standard editing distance threshold, and in the case, the matching of the standard word 'Boshi fund' is determined to be successful.

In this embodiment, word segmentation processing is performed on each standard word, and the answer information is matched through the standard word segmentation to obtain a plurality of matching candidate sets, so that the probability of successful matching can be improved, and the matching accuracy is also improved.

In one embodiment, the evaluation parameter of the content standardization further comprises a preset standard character length threshold. As shown in fig. 7, the specific processing procedure of "performing standard word matching in the answer information by using a preset standard degree matching algorithm, and determining a content standard degree score of the answer information according to a matching result" in step 301 includes:

step 701, if the character length of the standard word is smaller than a preset standard character length threshold, the answer information and the standard word are respectively converted into an answer information pinyin sequence and a standard word pinyin sequence.

Specifically, the preset standard character length threshold is a preset standard word length threshold. If the scoring device judges that the character length of the standard word needing standard word matching at present is smaller than a preset standard character length threshold value, the scoring device converts the acquired answer information into an answer information pinyin sequence and converts the standard word into a standard word pinyin sequence.

Step 702, generating a regular expression of the standard word pinyin sequence by a preset regular expression generation algorithm.

Specifically, the scoring device may generate the regular expression of the pinyin sequence of each standard word for matching through a preset regular expression generation algorithm.

And 703, matching in the answer information pinyin sequence according to the regular expression of the standard word pinyin sequence, and counting the matching number of the fourth standard word.

Specifically, the scoring device matches each standard word in the answer information respectively through the generated regular expressions of the pinyin sequences of the plurality of standard words, counts the occurrence times of the standard words in the answer information, and determines the matching number of the fourth standard words according to the statistical result.

In one embodiment, the character length of the standard word has multiple conditions, in which case, the evaluation parameter of the content standardization further includes a preset standard character length threshold and a standard editing distance threshold. As shown in fig. 8, the specific processing procedure of "performing standard word matching in the answer information by using a preset standard degree matching algorithm, and determining a content standard degree score of the answer information according to a matching result" in step 301 includes:

step 801, if the number of characters of the standard word is greater than or equal to a preset standard character length threshold, performing word segmentation processing on the standard word to obtain a plurality of standard word segments.

Step 802, the answer information and the standard word segmentation are converted into an answer information pinyin sequence and a standard word segmentation pinyin sequence respectively.

Step 803, extracting character strings containing the pinyin sequence of the participles of the standard words from the pinyin sequence of the answer information through a preset matching window to form a matching candidate set, wherein the length of the preset matching window is the number of characters contained in the pinyin sequence of the standard words.

Specifically, the scoring device takes the character length of the standard word pinyin sequence as the length of a preset matching window, and in this case, the scoring device extracts a plurality of character strings containing the standard word segmentation pinyin sequence from the answer information pinyin sequence through the preset matching window. In this way, the length of the plurality of character strings extracted by the scoring device is the length of the preset matching window, and the plurality of extracted character strings are taken as a matching candidate set.

And step 804, calculating the editing distance between the pinyin sequence of the standard word and each character string in the matching candidate set, and when the editing distance is smaller than or equal to the threshold value of the standard editing distance, determining that the standard word is successfully matched to obtain the matching quantity of the fifth standard word.

Specifically, the scoring device may calculate the edit distance between each character string in the matching candidate set and the pinyin sequence of the standard word, respectively. If the edit distance corresponding to the character string in the matching candidate set is smaller than or equal to the standard edit distance threshold, the scoring device may determine that the standard word is successfully matched, and obtain the fifth standard word matching number. If the edit distances corresponding to the character strings in the matching candidate set are all larger than the standard edit distance threshold, the scoring device can determine that the word segmentation matching of the standard word fails, in this case, the scoring device can match the answer information again through the pinyin sequence of the second standard word segmentation, and execute the processes from the step 803 to the step 804 until the matching of the standard word succeeds or the matching of the standard word fails.

In this embodiment, the word segmentation processing is performed on each standard word, and the answer information is matched through the standard word segmentation to obtain a plurality of matching candidate sets, so that the probability of successful matching can be improved, the matching accuracy is also improved, and the recall probability of the standard words can be improved on the premise of correct recognition.

In one embodiment, there may be a plurality of standard words corresponding to the target question pre-configured in the scoring device, and the matching of the standard words may be performed by the method of any of the above embodiments, so that the matching number of the standard words may be determined according to the sum of one or more of the first standard word matching number, the second standard word matching number, the third standard word matching number, the fourth standard word matching number, and the fifth standard word matching number.

In one example, for the target question Q1, the standard words preset by the scoring device may be a1, a2, A3, a4, a5, a 6. Matching the standard words A1 and A3 in the answer information successfully by the method from step 401 to step 403, and counting the matching number of the first standard words as 2; the standard words A2 and A4 are successfully matched in the answer information through the method from the step 501 to the step 503, and the matching number of the second standard words is counted to be 2; the standard words A5 and A6 are successfully matched in the answer information through the methods from step 801 to step 803, and the matching number of the fifth standard words is counted to be 2; accordingly, the scoring device may determine that the standard word matching number may be the sum of the first standard word matching number, the second standard word matching number, and the fifth standard word matching number (6).

In one embodiment, the content criteria includes at least one of keyword hit level, compliance level, emotion level, and usage level of spoken words. The keyword hit degree score, the compliance degree score, the emotion degree score and the word use degree score can be calculated through the preset content standard degree matching algorithm provided by the embodiment, and different content standard degrees are different only in evaluation parameters of each evaluation dimension when the preset content standard degree matching algorithm is used.

In one example, when the content criterion includes a keyword hit degree, the criterion word may be an answer keyword configured in advance according to an actual demand of the target user and the target question. When the content criterion includes a compliance, the criterion words may be offending words that are pre-configured according to the actual needs of the target user, such as words with an insulting meaning. When the content standardization level includes emotion level, the standardization word may also be an emotion word configured in advance according to actual needs of the target user, for example, a word with negative emotion. When the content standard degree comprises the usage degree of the jargon, the standard word can also be the jargon which is configured in advance according to the actual requirement of the target user, for example, the jargon which can achieve the effect of promoting sales.

Optionally, when the scoring device matches the standard word in the answer information, the scoring device may match generalized words of the standard word together. The generalization words of the standard words can be pre-configured according to the actual application scene.

In one embodiment, before the step of performing standard word matching in the answer information by using a preset matching algorithm to obtain the number of standard word matches of the answer information, the scoring method further includes:

and recognizing and eliminating punctuations in the answer information through a preset punctuation dictionary to obtain the answer information with the punctuations eliminated. And/or converting the numerical values in the answer information into Chinese character expressions to obtain the converted answer information.

In a possible example, if the answer information input by the target user is in a voice form, in this case, the answer information obtained by performing voice conversion may have a sentence-break error. And the scoring equipment identifies punctuation marks in the answer information through a preset punctuation mark dictionary and removes the punctuation marks. In addition, if the answer information input by the target user is in a voice form, and the voice information contains numerical information. Therefore, the answer information obtained by the scoring equipment through voice conversion has the condition of random numerical value conversion, and under the condition, the scoring equipment can uniformly convert the standard words and the numerical value expressions in the answer information into the Chinese character expressions.

In the embodiment, by removing the punctuation marks, introduction of external errors can be reduced, and the scoring accuracy is improved. The Chinese character expression is uniformly converted through numerical values, and the detection capability of the marking equipment for the standard words is improved.

In one embodiment, the response information generated by the target user may be obtained for multiple sub-questions in a lump. When the scoring device scores on the evaluation dimension of the integrity degree, the length of the text and whether the text is related to the standard word are concerned. Therefore, the evaluation parameters of the integrity include a total question character length threshold value and a subproblem character length threshold value, and the answer information includes a plurality of subproblem answer texts. Accordingly, as shown in fig. 9, the specific implementation process of "counting the text length of the answer information by a preset integrity statistic algorithm, and determining the integrity score of the answer information according to the statistical result" in step 302 includes:

step 901, if the character length of the answer information is greater than or equal to the total question character length threshold, determining a first score according to a numerical value of a preset total length score.

Specifically, when the scoring device determines that the total character length of the answer information is greater than or equal to the total question character length threshold, in this case, the scoring device may assign a preset total length score to the answer information in the evaluation dimension of the integrity degree, that is, determine the first score of the integrity degree score.

And step 902, if the character length of the sub-question answer text is greater than or equal to the sub-question character length threshold, determining a second score according to the numerical value of the preset sub-length score.

Specifically, when the scoring device determines that the character length of the answer text of the plurality of sub-questions of the answer information is greater than or equal to the sub-question character length threshold, in this case, the scoring device may assign a preset sub-length score to the answer information in the evaluation dimension of the completeness, that is, determine the second score of the completeness score, according to the number of sub-question answer texts exceeding the sub-question order length threshold.

And step 903, determining a third score of the plurality of sub-question answer texts according to the content standard degree score.

Specifically, according to the method described in step 201, the standard words corresponding to the sub-questions are respectively matched in the sub-question answer texts, and the third score of the integrity score is determined according to the number of the sub-question answer texts with successfully matched standard words.

And step 904, taking the sum of the first score, the second score and the third score as the integrity score of the answer information.

Specifically, the completeness score of the answer information may be calculated by the following formula:

z＝z1+∑(z2+z3)

where z is the completeness score, z1 is the first score, Σ (z2+ z3) is the sum of the second and third scores, z2 is the sub-length, and z3 is the score for the sub-question hit keyword.

In an embodiment, the specific implementation process of step 203 "counting the number of preset linguistic words in the answer information by using a preset fluency statistical algorithm, and determining the fluency score of the answer information according to the statistical result" includes:

when the scoring device scores the answer information in the evaluation dimension of the fluency, the scoring basis is the number of preset tone words appearing in the answer information. The preset tone word may be an unsmooth tone word (e.g., hiccup). The scoring device can preset the fluency full-score and the configuration tolerance number, namely, in an actual application scene, the scoring device can tolerate the occurrence of a certain number of unsmooth tone words when a target user answers a question. When the number of the appearance times of the unsmooth tone words in the answer information exceeds the tolerance number, the scoring device determines a punishment deduction score according to the number of the appearance times of the unsmooth tone words exceeding the tolerance data until the preset fluency full-score is deducted to 0.

Alternatively, the fluency score may be calculated by the following formula:

F＝max(M-P*max(R-U,0)，0)

wherein F is fluency score, M is preset fluency full score, P is punishment mark, R is the number of appearance of the non-fluent mood words, and U is tolerance number.

In an embodiment, the specific implementation process of "calculating the speech rate of the answer information by using a preset speech rate determining algorithm, and obtaining the speech rate score of the answer information according to the speech rate" in step 204 includes:

and the scoring equipment counts the number of text words in the answer information and the duration of the answer information through a preset speech rate determining algorithm, and then calculates the speech rate of the answer information. In this way, the scoring device may determine the speech rate score of the answer information in the speech rate evaluation dimension according to the interval where the speech rate of the answer information is located.

Alternatively, the interval in which the speech rate of the answer information is located may include too slow, somewhat slow, normal, somewhat fast, and too fast, and the corresponding speech rate scores are 10, 50, 100, 75, and 25, respectively.

Optionally, the scoring device may determine, according to a preset speech rate interval determination standard, a speech rate interval corresponding to the calculated speech rate, where the preset speech rate interval determination standard may be specifically determined according to an actual application scenario.

In an embodiment, the compliance score may also be determined by the following process: and the scoring equipment counts the occurrence times of the illegal words in the answer information through a preset standard word matching algorithm, and determines the compliance score according to the occurrence times of the illegal words.

Alternatively, the compliance score may be calculated by the following formula:

H＝max(0,H1-H2*H3)

wherein, H is the compliance score, H1 is the preset compliance full score, H2 is the number of occurrences of the violation word, and H3 is the preset score for deducting the number of occurrences of the violation word.

In one embodiment, the sentiment score may also be determined by the following process: and the scoring equipment counts the occurrence times of the negative emotion words in the answer information through a preset standard word matching algorithm, and determines the emotion degree score according to the occurrence times of the negative emotion words.

Alternatively, the sentiment score may be calculated by the following formula:

G＝max(0,G1-G2*G3)

wherein G is the emotion degree score, G1 is the preset emotion degree full score, G2 is the number of occurrences of the negative emotion word, and G3 is the preset score for each occurrence of the negative emotion word.

In one embodiment, the conversational word usage score may also be determined by the following process: and the scoring equipment counts the occurrence times of the preset dialect words in the answer information through a preset standard word matching algorithm, and determines the usage degree score of the dialect words according to the occurrence times of the preset dialect words.

Alternatively, the pronoun word usage score may be calculated by the following formula:

S＝max(0,S1-S2*S3)

wherein, S is the score of the usage degree of the speech term, S1 is the score of the full score of the usage degree of the preset speech term, S2 is the number of occurrences of the preset speech term, and S3 is the score of the preset speech term per occurrence.

It should be understood that although the steps in the flowcharts of fig. 1 to 9 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1 to 9 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of performing the steps or stages is not necessarily sequential, but may be performed alternately or alternately with other steps or at least some of the other steps or stages.

The embodiment of the invention also provides a scoring method based on the online conversation training service system, and as shown in fig. 10, the online conversation training service system comprises scoring equipment and a management terminal. The scoring device comprises a scoring configuration service system, a database, a voice recognition system, a scoring system and a result feedback system. The management terminal may configure evaluation parameters corresponding to the plurality of target questions in the scoring configuration service system in advance, and store the evaluation parameters in the database. When the on-line dialogue training service system responds to the target question display instruction and displays the target question, the user equipment generates answer information corresponding to the target question. The online dialogue training service system responds to the answer instruction corresponding to the target question, obtains answer information of the target user aiming at the target question through the voice recognition system, and obtains evaluation parameters of multiple evaluation dimensions corresponding to the target question in the database. Therefore, the online dialogue training service system can grade the answer information on a plurality of evaluation dimensions through the grading system to generate grading feedback information, and the grading feedback information is fed back to the target user through the result feedback system, so that the target user can correspondingly improve the next answer to the question.

In one embodiment, as shown in fig. 11, there is provided a scoring apparatus including: an obtaining module 1001, a scoring module 1002, and an outputting module 1003, wherein:

an obtaining module 1001, configured to, in response to an answer instruction corresponding to a target question, obtain answer information of a target user for the target question, and obtain evaluation parameters of the target question for multiple evaluation dimensions;

the scoring module 1002 is configured to score the answer information according to the evaluation parameters of the multiple evaluation dimensions to obtain scoring feedback information, where the scoring feedback information includes scoring score information and scoring meaning information corresponding to the scoring scores;

the output module 1003 is configured to output the scoring score information and scoring meaning information corresponding to the scoring score.

In one embodiment, the plurality of evaluation dimensions at least comprise one or more of content standard degree, completeness, fluency, and speech speed; the scoring module 1002 includes:

the content standard degree scoring unit is used for performing standard word matching in the answer information through a preset content standard degree matching algorithm and determining the content standard degree score of the answer information according to a matching result;

the integrity scoring unit is used for counting the text length of the answer information through a preset integrity counting algorithm and determining the integrity score of the answer information according to a counting result;

the fluency scoring unit is used for counting the number of preset tone words in the answer information through a preset fluency statistical algorithm and determining the fluency score of the answer information according to a statistical result;

the speech rate scoring unit is used for calculating the speech rate of the answer information through a preset speech rate determining algorithm and obtaining the speech rate score of the answer information according to the speech rate;

and the scoring unit is used for analyzing the content standard degree score, the integrity degree score, the fluency score and the speech speed score to obtain scoring feedback information.

the quantity determining subunit is used for matching the standard words in the answer information through a preset content standard degree matching algorithm and counting the standard word matching quantity of the answer information;

and the score determining subunit is used for determining the content standard degree score of the answer information according to the standard words and the preset standard quantity threshold.

In one embodiment, the number determining subunit is specifically configured to:

generating a regular expression of the standard words through a preset regular expression generation algorithm;

and matching the standard words in the corrected answer information according to the regular expression, and counting the matching number of the first standard words.

In one embodiment, the number determining subunit is specifically configured to:

and matching the standard words in the answer information according to the regular expression, and counting the matching number of the second standard words.

In one embodiment, the evaluation parameters of the content standardization further include a preset standard character length threshold and a standard editing distance threshold; the number determination subunit is specifically configured to:

In one embodiment, the evaluation parameter of the content standardization further comprises a preset standard character length threshold; the number determination subunit is specifically configured to:

generating a regular expression of the standard word pinyin sequence by a preset regular expression generation algorithm;

and matching in the answer information pinyin sequence according to the regular expression of the standard word pinyin sequence, and counting the matching number of the fourth standard word.

converting the answer information and the standard word segmentation into an answer information pinyin sequence and a standard word segmentation pinyin sequence respectively;

In one embodiment, the apparatus further comprises:

the preprocessing subunit is used for identifying and eliminating punctuations in the answer information through a preset punctuation dictionary to obtain the answer information with the punctuations eliminated; and/or converting the numerical values in the answer information into Chinese character expressions to obtain the converted answer information.

the integrity scoring unit is specifically configured to determine a first score according to a numerical value of a preset total length score if the character length of the answer information is greater than or equal to the total question character length threshold; if the character length of the sub-question answer text is larger than or equal to the sub-question character length threshold value, determining a second score according to a numerical value of a preset sub-length score; determining a third score for the plurality of sub-question answer texts according to the content standardization score; and taking the sum of the first score, the second score and the third score as a completeness score of the answer information.

For the specific limitations of the scoring device, reference may be made to the limitations of the scoring method above, which are not described herein again. The respective modules in the above scoring apparatus may be wholly or partially implemented by software, hardware, and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.

In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 12. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing evaluation parameter data and scoring feedback information. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a scoring method.

Those skilled in the art will appreciate that the architecture shown in fig. 12 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is further provided, which includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above method embodiments when executing the computer program.

In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when being executed by a processor, carries out the steps of the above-mentioned method embodiments.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

29页详细技术资料下载

上一篇：一种医用注射器针头装配设备

下一篇：一种多特征融合的说话人识别方法

Scoring method, scoring device, computer equipment and storage medium

相关技术

网友询问留言