Athlete psychological consulting service system and method based on natural language processing

Abstract

The present invention relates to an athlete psychological counseling service system and method based on natural language processing, and belongs to the field of sports psychology and artificial intelligence application technology. The present invention utilizes regular expression matching and replacement methods in natural language processing to replace text, thereby realizing real-time emotional dialogue feedback to athlete inquirers, guiding visitors to talk, and completing information collection in the initial stage of psychological counseling. Through a probability-based text classification method, combined with a corpus of psychological questions, the information collected in the initial stage of consultation is classified, and the psychological problems described by athlete inquirers are professionally and reliably classified. For different categories of psychological problems, psychological intervention is implemented through a standardized psychological chat process, and psychological assistance and psychological support are provided to athlete inquirers in a professional psychological counseling manner. The present invention is of great significance to maintaining the mental health of athletes during preparation and competition, and improving competitive sports performance.

Description

Technical Field

[0001] The present invention relates to an athlete-based psychological consulting service system and method, belonging to the technical field of sports psychology and artificial intelligence application. Background Art

[0002] Athletes face high pressure during competition preparation and have a high demand for psychological support. Their mental health is most directly impacted by their competitive performance. Even when competitors possess similar physical conditions, talent, and technical abilities, psychological factors often become the key factor in determining victory or defeat. Psychological counseling, utilizing psychological methods, provides psychological support to those experiencing challenges with psychological adaptation and seeking solutions.

[0003] However, the current human resources for athlete psychological intervention and support are severely insufficient. With a small number of practitioners and short periods of time on-site, traditional face-to-face counseling is the only form of support. Traditional psychological counseling relies on a manual consultation model, requiring clients to select a counselor and receive consultation via chat. This model no longer meets the flexible, around-the-clock, and timely response needs of modern athletes.

[0004] With the rapid development of artificial intelligence, big data, and cloud computing, intelligent systems for psychological counseling and mental health have emerged. Currently, research institutions both domestically and internationally have developed AI systems or mobile applications for psychological counseling or psychotherapy. Some, such as the Woebot mobile application for psychological counseling, have been deployed for the general public or in medical settings.

[0005] However, existing intelligent psychological counseling systems are mostly designed and developed for the general public or individuals with medical conditions, providing simple mental health counseling services. Research has found that athletes' physical and psychological development is characterized by imbalance, instability, and plasticity. In particular, their mental state is not a stable constant; it is easily influenced and becomes increasingly complex. Athletes' unique status increases the number of factors contributing to their psychological changes. If not properly addressed, this can easily lead to psychological imbalances, impacting their normal learning, training, and daily lives. Currently, there are no intelligent systems or methods for psychological counseling intervention and service support specifically targeting this special group of athletes. Summary of the Invention

[0006] The purpose of this invention is to address the problems and shortcomings of the existing technology and creatively propose an athlete psychological counseling service system and method based on natural language processing, which can provide athletes with identification, evaluation, early warning and counseling services for common sports psychology problems.

[0007] The innovation of this invention lies in its use of regular expression matching and replacement methods in natural language processing to replace text, providing real-time emotional dialogue feedback to athletes seeking counseling. This can guide clients to express themselves and complete information collection in the initial stages of psychological counseling. A probability-based text classification method, combined with a corpus of psychological questions, classifies the information collected in the initial stages of counseling, providing a professional and reliable classification of the psychological issues described by athletes seeking counseling. Finally, psychological intervention is implemented through a standardized psychological chat process, targeting different psychological issue categories, providing psychological assistance and support to athletes seeking counseling through professional psychological counseling.

[0008] Beneficial effects

[0009] 1. During the relationship-building and information-gathering phases, this invention uses a text-replacement-based approach to generate responses, avoiding the natural language understanding difficulties of existing chat systems. Furthermore, because responses include the caller's own narrative, athletes are more likely to perceive the system as empathetic, making it easier to establish relationships with the system.

[0010] 2. The present invention uses whether the information entropy is less than a certain threshold to determine whether the psychological problem recognition is successful, rather than judging by adding an "unrecognizable" class.

[0011] 3. The present invention classifies sports psychology problems and uses single-choice, multiple-choice, picture, text video, psychological assessment and other forms for each type of problem. It also formulates standardized processes based on the characteristics of sports psychology problems to achieve intelligent and efficient companionship and psychological counseling guarantees, which is of great significance to maintaining the mental health of athletes during preparation and competition, and improving their competitive sports performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic diagram of the composition structure of the system of the present invention.

[0013] Figure 2 It is a schematic flow chart of the method of the present invention.

[0014] Figure 3 This is a typical simplified version of the standardized flow chart of psychological services included in the present invention. DETAILED DESCRIPTION

[0015] The present invention will be further described in detail below with reference to the accompanying drawings.

[0016] like Figure 1 As shown, a psychological counseling service system for athletes based on natural language processing includes a human-computer interaction interface, a relationship building and information collection module, a service goal establishment module, a psychological service standardization process module, a key information extraction module, and a sentence conversion module. Among them:

[0017] Human-computer interaction interface: Provide psychological service interaction for athletes by providing input and output methods such as text and voice.

[0018] Relationship Building Module: This module uses regular expressions from natural language processing to replace athlete input text and generate responses, encouraging athletes to express themselves more openly. When performing chat text replacement, the use of "pattern matching" and substitution methods to simulate conversations demonstrates the superficial nature of communication between humans and machines. This method gives users the illusion of program understanding, but lacks a built-in framework for contextualizing events.

[0019] Service Target Establishment Module: Responsible for problem identification and confirmation. Problem identification uses a probability-based machine learning text classification algorithm based on the athlete's input to establish a service target. Problem confirmation calculates information entropy based on the probabilities output by the text classification algorithm. If the information entropy is less than a set threshold, the service target is considered successfully established.

[0020] The standardized psychological service module is responsible for the support and change phases, as well as the closing phase. During the support and change phase, case stories, psychological training, assessment tools, and psychological pre-planning strategies are used, depending on the service objectives, to help athletes address cognitive or behavioral issues. The closing phase is used to review key service points and consolidate the effectiveness of the service.

[0021] Key information extraction module: Use the sentence importance evaluation method to extract important sentences from the athlete's input information, input them into the sentence conversion module, and generate reply sentences.

[0022] Sentence conversion module: Uses machine learning to classify the collected athlete expressions. For each type of expression, a matching rule and multiple conversion rules are formulated. Each conversion rule converts the athlete's input sentence into an output, including empathy conversion rules, concretization conversion rules, open-ended question conversion rules, self-disclosure conversion rules, universal responses, etc. According to different psychological service stages, each type of rule is assigned a weight, and these weights are used to randomly select conversion rules. At the same time, the sentence conversion module also includes person and sentence conversion rules (such as switching first-person pronouns and second-person pronouns, or converting declarative sentences into interrogative sentences, rhetorical questions, exclamatory sentences, etc.).

[0023] The connection relationship of the above modules is as follows:

[0024] Athletes input information into the system through the human-computer interface. If the input is voice, it undergoes speech recognition and is converted into text, which is then fed into the Service Goal Establishment Module. The Service Goal Establishment Module, based on whether it has identified the athlete's desired problem, feeds the athlete's input into the Relationship Establishment Module or the Psychological Service Standardization Module. In the Relationship Establishment Module, the Key Information Extraction Module first extracts the athlete's key sentences, and then the Sentence Transformation Module generates a response.

[0025] Furthermore, based on the above system, the present invention proposes a psychological consulting service method for athletes based on natural language processing, such as Figure 2 As shown, the following steps are included:

[0026] Step 1: Output the greeting message through the human-computer interaction interface.

[0027] The greetings are formulated by psychology experts, who randomly select one each time and wait for the athlete's response.

[0028] Step 2: For all the information input by the athletes in this service, use a probability-based method to identify the athletes' psychological problems.

[0029] This method outputs the probability of which category of psychological problems the athlete belongs to, which is given by H = -Σp i log(p i ) Calculate the entropy of the recognition result, where H represents entropy, p i is the probability that the psychological problem belongs to the i-th category. If H≤γ, the classification is considered successful and jump to step 6. γ is a pre-set threshold. Otherwise, execute step 3.

[0030] Step 3: Use an extractive text summarization algorithm to extract the most important sentences from the athlete's response in this round.

[0031] Step 4: Use the matching rules to match the key sentences of the athlete's response in this round, and then randomly select the conversion rules based on the weights to generate a response to the athlete.

[0032] If no suitable rule is matched, a common statement is randomly selected for reply.

[0033] If multiple reply rules are matched, the longest match principle is used to select the rule that matches the most words.

[0034] If there is a match with the same number of words, the rule written in front is matched according to the order of the rules.

[0035] Specifically, the matching rules are formulated as follows:

[0036] First, we crawled athletes’ Weibo accounts and expert interviews to obtain athletes’ commonly used expressions and formed an athlete expression corpus.

[0037] Then, the text similarity algorithm and grammatical analysis results are used to cluster the athlete expression corpus;

[0038] Afterwards, for each type of statement, psychology experts will draft responses in categories such as empathy and specificity. When drafting responses, psychology experts should use phrases from the original statement as much as possible.

[0039] Finally, the response rules of the psychology experts are summarized to form matching rules such as "I am worried (.*)" and "Why are you worried (%1)?", where (.*) represents any text here, for example, "I am worried that I have not trained enough"; (%1) represents the athlete's statement matched by (.*) after person conversion. For example, after the above statement and rules are converted, the result is "Why are you worried that you have not trained enough?".

[0040] Step 5: Delay for a period of time (determined according to actual needs, such as 800 seconds) to simulate real-person input, feed back the generated response to the athlete through the human-computer interaction interface, and wait for the athlete's next round of input before jumping to step 2.

[0041] Step 6: Select the service target corresponding to the category with the highest probability, and select the standardized process for psychological problems based on the service target. Typical standardized processes are as follows: Figure 3 shown.

[0042] Step 7: Use a single-choice method to ask athletes to provide feedback on the service effect.

Claims

1. An athlete psychological consulting service system based on natural language processing, characterized in that: It includes human-computer interaction interface, relationship building and information collection module, service goal establishment module, psychological service standardization process module, key information extraction module, and sentence conversion module; Among them, the human-computer interaction interface provides psychological service interaction for athletes; The relationship building module uses regular expressions in natural language processing to replace athlete input text to generate responses, encouraging athletes to be more willing to talk. When replacing chat text, it uses "pattern matching" and substitution methods to simulate conversations. The service goal establishment module is responsible for problem identification and problem confirmation. Problem identification uses the athlete's input information and a probability-based machine learning text classification algorithm to establish the service goal. Problem confirmation calculates the information entropy based on the probability output by the text classification algorithm. If the information entropy is less than the set threshold, the service goal is considered to have been successfully established. The psychological service standardization module is responsible for providing assistance, change, and the ending stage. During the assistance and change stage, athletes are helped to change their cognitive or behavioral problems based on different service goals. The ending stage is used to review the key points of the service and consolidate the service results. The key information extraction module is responsible for using the sentence importance evaluation method to extract important sentences from the athlete's input information, inputting them into the sentence conversion module to generate reply sentences; The sentence conversion module uses machine learning to classify the collected athlete expressions. For each type of expression, a matching rule and multiple conversion rules are developed. Each conversion rule converts the athlete's input sentence into an output, including empathy conversion rules, concretization conversion rules, open-ended question conversion rules, self-disclosure conversion rules, and universal response conversion rules. Each type of rule is assigned a weight based on different psychological service stages, and these weights are used to randomly select conversion rules. The sentence conversion module also includes person and sentence structure conversion rules. The connection relationship of the above modules is as follows: Athletes input information into the system through the human-computer interaction interface. If the input is voice, it is converted into text after voice recognition and input into the service goal establishment module. The service goal establishment module sends the athlete's input to the relationship establishment module or the psychological service standardization module based on whether it has determined the problem the athlete wants to solve. In the relationship establishment module, the key information extraction module first extracts the athlete's key sentences, and then the sentence conversion module generates a response. The system working method includes the following steps: Step 1: Output greetings through the human-computer interaction interface; The greetings are formulated by psychology experts, who randomly select one each time and wait for the athlete's response. Step 2: Using a probability-based approach to identify athletes’ psychological problems based on all the information they input in this service; This method outputs the probability of which category of psychological problems the athlete belongs to, which is given by H = -∑p i log(p i ) Calculate the entropy of the recognition result, where H represents entropy, p i is the probability that the psychological problem belongs to the i-th category. If H≤γ, the classification is considered successful and the process goes to step 6. γ is a pre-set threshold. Otherwise, the process goes to step 3. Step 3: Use an extractive text summarization algorithm to extract the most important sentences from the athlete's response in this round; Step 4: Use the matching rules to match the key sentences of the athlete's response in this round, and then randomly select the conversion rules based on the weights to generate a response to the athlete; If no suitable rule is matched, a common statement is randomly selected for reply; If multiple reply rules are matched, the longest match principle is used to select the rule with the most matching words; If there is a match with the same number of words, the rule written in front is matched according to the order of the rules; Step 5: After a delay to simulate real-person input, the generated response is fed back to the athlete through the human-computer interaction interface. After waiting for the athlete's next round of input, jump to step 2; Step 6: Select the service target corresponding to the category with the highest probability, and select the standardized process for psychological problems based on the service target; Step 7: Use a single-choice method to ask athletes to provide feedback on the service effect; The matching rules are formulated as follows: First, we obtain the expressions commonly used by athletes and form an athlete expression corpus; Then, the text similarity algorithm and grammatical analysis results are used to cluster the athlete expression corpus; Afterwards, for each type of statement, the psychology expert will formulate a response in the empathy and specificity categories; when formulating the response, the psychology expert should use the sentence in the original statement; Finally, the response rules of psychology experts are summarized to form matching rules.

Images