Ecological tourism resource assessment method and sustainable development planning system

Abstract

The invention relates to the technical field of tourism resource analysis, in particular to an ecological tourism resource evaluation method and a sustainable development planning system. Performing text vectorization expression, performing word segmentation processing and similarity calculation and judgment on a comment set of each time period, starting from a second time period, calculating cosine similarity CS between a comment semantic vector of a current time period and a comment semantic vector of a previous time period, and comparing the similarity CS with a preset threshold value theta, according to the method, on-line comment data is extracted to determine a calculation formula of a front attention trend score and a word-of-mouth stability score, and a text similarity-based emotion trend and word-of-mouth stability index intelligent adjustment method is provided by utilizing semantic similarity characteristics of the on-line comment data, so that rapid fine adjustment of the index when the data similarity is relatively high is realized; unnecessary repeated calculation is effectively reduced, and the analysis efficiency is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tourism resource analysis, and particularly to an ecological tourism resource evaluation method and a sustainable development planning system. Background Art

[0002] Traditional tourism resource evaluation methods usually adopt means such as questionnaires, expert evaluations, and on-site surveys. These methods have long cycles, high costs, and strong subjectivity, and cannot capture the actual demand changes and emotional feedback of tourists in real time and efficiently. In recent years, with the wide application of Internet technology, a large number of tourists have real-time published experience evaluations of tourist attractions on major online platforms (such as Ctrip, Dianping, Weibo, Mafengwo). These online review data reflect the real psychological demands and experience feelings of tourists. Therefore, tourism resource evaluation based on online reviews has gradually become a research hotspot.

[0003] In the existing online evaluation method for tourism regional resources (publication number CN118132818B), a tourism regional resource evaluation method based on semantic differential is proposed. This method first analyzes the positive attention trend and word-of-mouth stability of tourist attractions through text data collection and semantic space construction, so as to achieve tourism resource evaluation. However, this method still has the following obvious deficiencies in practical applications:

[0004] (1) The analysis process adopted in this method is to independently conduct sentiment analysis and trend calculation on the online review texts of each day. This daily independent analysis method does not consider the continuity and repeatability characteristics of the review content, resulting in a large amount of calculation. Especially when the amount of review data is huge, it will cause serious waste of computing resources and time;

[0005] (2) Online review data often shows the characteristic of highly similar content within a continuous time period. Traditional methods fail to effectively utilize this similarity for simplification processing and optimization calculation, resulting in a large amount of repetitive labor and low efficiency of sentiment trend analysis and word-of-mouth stability evaluation;

[0006] (3) During the peak tourist season, holidays, or hot events, the review data shows an explosive growth. The completely recalculated mode of the existing method is likely to cause the system to have an excessive computing load and poor real-time performance, and cannot meet the actual needs of real-time operation management and rapid decision-making of scenic spots;

[0007] (4) In addition, although the existing method proposes the concepts of positive attention trend and word-of-mouth stability, it does not clarify how to accurately fine-tune the indicators in the case of a high degree of repetition of review content and cannot adapt to the actual complex application scenarios.

[0008] In summary, the current online evaluation method for tourism regional resources urgently needs to solve the following technical problems:

[0009] How to effectively reduce duplicate calculations and improve analysis efficiency while ensuring the accuracy of evaluation results;

[0010] How to utilize the content similarity features of online reviews to achieve precise fine-tuning of indicators, thereby quickly and effectively evaluating tourists' emotional trends and the stability of scenic area word-of-mouth;

[0011] How to effectively alleviate the computing load during peak data bursts and ensure the real-time response speed and decision-making support capabilities of the system.

[0012] Therefore, an evaluation method that can solve the above problems and improve the efficiency of online evaluations of tourist attractions is needed. Summary of the Invention

[0013] The purpose of the present invention is to provide an ecological tourism resource evaluation method and a sustainable development planning system to solve the above technical problems.

[0014] The present invention proposes an ecological tourism resource evaluation method and a sustainable development planning system, including the following steps:

[0015] (1) Data collection and preprocessing. Through web crawler technology, regularly obtain online text review data of tourists on tourist attractions from a tourism evaluation platform, and divide the review data by fixed time periods to form a text set for each time period;

[0016] (2) Text vector representation. Perform word segmentation on the review set for each time period divided in step (1). After removing stop words, use the Word2Vec model to calculate the average of the vector representations of all words within that time period to obtain a high-dimensional semantic vector a representing the overall semantics of that time period;

[0017] (3) Similarity calculation and judgment. Starting from the second time period, calculate the cosine similarity CS between the review semantic vector of the current time period and the review semantic vector of the previous time period. The calculation formula is as follows:

[0018]

[0019] Where is the review semantic vector of the current time period; is the review semantic vector of the previous time period;

[0020] (4) Compare the size of the similarity CS and the preset threshold θ to determine the calculation formulas for the positive attention trend score and the word-of-mouth stability score.

[0021] (5) According to the positive attention trend score and the word-of-mouth stability score, calculate the virtual correlation quotient for tourism scenic area operation risk assessment.

[0022] Furthermore, the calculation formula for the similarity CS is as follows:

[0023]

[0024] where is the comment semantic vector for the current time period; is the comment semantic vector for the previous time period.

[0025] Furthermore, if the similarity CS is less than the preset threshold θ, a complete sentiment analysis is performed on the current time period, the proportion of positive comments is calculated, and the sentiment trend change amplitude Δ is calculated in combination with the proportion of positive comments in the previous time period. Then, the positive attention trend score and the word-of-mouth stability score for the current time period are calculated. If the similarity CS is greater than or equal to the preset threshold θ, the positive attention trend score and the word-of-mouth stability score are fine-tuned according to the sentiment trend change amplitude Δ and the similarity CS, without having to recalculate completely.

[0026] Furthermore, the calculation formula for the proportion of positive comments is:

[0027]

[0028] In the formula, N is the total number of comments in the current time period, and N positive is the number of positive comments in the current time period.

[0029] Furthermore, the calculation formula for the sentiment trend change amplitude Δ is:

[0030] Δ = P t - P t-1

[0031] In the formula, P t is the proportion of positive comments in the current time period, and P t-1 is the proportion of positive comments in the previous time period;

[0032] If Δ > 0, it indicates an upward sentiment trend;

[0033] If Δ < 0, it indicates a downward sentiment trend;

[0034] If Δ = 0, it indicates a stable sentiment trend.

[0035] Furthermore, the adjustment formula for the positive attention trend score is:

[0036] T t = T t-1 + (1 - CS) × Δ

[0037] In the formula, T t is the positive attention trend score for the current time period, and T t-1 is the positive attention trend score for the previous time period.

[0038] Furthermore, the formula for adjusting the word-of-mouth stability score is as follows:

[0039] S t = S t-1 × cs

[0040] In the formula, S t is the word-of-mouth stability score for the current time period, S t-1 is the word-of-mouth stability score for the previous time period, and CS is the cosine similarity between the texts of the current time period and the previous time period.

[0041] The present invention also provides a scenic area sustainable development planning system that applies the above ecological tourism resource evaluation method.

[0042] Furthermore, an evaluation report on the operation risk of the scenic area is generated according to the virtual correlation quotient, to assist in the management decision-making of the scenic area.

[0043] Furthermore, the specific criteria for classifying the operation risk level of the scenic area are as follows:

[0044] When the positive trend score < 0.3 and the stability score < 0.4, it is defined as a high risk;

[0045] When the positive trend score is between 0.3 and 0.6 and the stability score is between 0.4 and 0.7, it is defined as a medium risk;

[0046] When the positive trend score > 0.6 and the stability score > 0.7, it is defined as a low risk.

[0047] The beneficial effects achieved by the present invention using the above method are as follows:

[0048] (1) The present invention makes full use of the semantic similarity characteristics of online review data, and proposes an intelligent adjustment method for emotional trends and word-of-mouth stability indicators based on text similarity, realizing rapid fine-tuning of indicators when the data similarity is high, effectively reducing unnecessary repeated calculations, and significantly improving the analysis efficiency.

[0049] (2) During the peak tourist season or holiday peak period, in the face of the explosive growth of review data, the similarity rapid judgment and adjustment mechanism adopted by the present invention can effectively reduce the computing load, greatly improve the real-time response speed of the system, and ensure the rapid decision-making ability of the scenic area operation management;

[0050] (3) The present invention accurately calculates the change range Δ of the emotional trend and the word-of-mouth stability indicator through a clear indicator adjustment formula, ensuring the accuracy and sensitivity of the indicator calculation, being able to promptly capture the subtle differences in tourists' emotional changes, and improving the reliability and accuracy of the online evaluation indicators;

[0051] (4) Based on the virtual correlation quotient calculation method and risk level evaluation system proposed by the present invention, it is possible to quickly and accurately generate a tourism scenic area operation risk assessment report, effectively assist scenic area management personnel in making scientific decisions, and thus improve the accuracy and sustainability of scenic area resource management. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flowchart of the ecological tourism resource evaluation method of the present invention;

[0053] Figure 2 It is a flowchart of the fine-tuning calculation steps of the positive attention trend score and word-of-mouth stability score of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0055] In order to make the content of the present invention more clearly understood, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention.

[0056] Embodiment 1

[0057] Step 1, data collection and preprocessing;

[0058] Through web crawler technology, regularly obtain online text comment data of tourists on tourism scenic areas from tourism evaluation platforms, divide the comment data by fixed time periods, and form a text set for each time period.

[0059] Step 2, text vectorization representation. For the text set of comment texts obtained by dividing each time period, perform text vectorization processing;

[0060] Use a word segmentation algorithm to perform word segmentation on each comment and remove stop words;

[0061] Use the Word2Vec model to calculate the average of the vector representations of all words in the comments of each time period to obtain a high-dimensional semantic vector a representing the overall semantics of the comment set for that day;

[0062] Store the obtained semantic vector a for future use.

[0063] Step 3, similarity calculation and judgment;

[0064] Starting from the second time period, calculate the similarity between the comment semantic vector a(t) in the current time period and the comment semantic vector a(t - 1) in the previous time period. The calculation formula is as follows:

[0065]

[0066] Step 4, if CS is less than θ, use the Word2Vec model to output the word vector of the current time period, calculate the virtual correlation quotient, and conduct an operation risk assessment;

[0067] In addition, it should be noted that when the number of consecutive times that CS is less than θ is greater than 5 times, a full calculation is forced.

[0068] Based on the above risk assessment content, determine the risk level by combining the positive attention trend score and the word-of-mouth stability score;

[0069] Preset the risk levels as shown in the following table;

[0070]

[0071] Generate a risk assessment report according to the risk content to help decision-makers adjust the scenic area policy.

[0072] Embodiment 2

[0073] This embodiment specifically describes how to perform text vectorization representation on the online review text of a tourist scenic area, use similarity indicators to achieve the optimization of sentiment trend and word-of-mouth stability analysis, and fuse the acquisition methods of trend direction and amplitude Δ for index adjustment.

[0074] Step 1, data collection and preprocessing;

[0075] Through web crawler technology, regularly obtain the online text review data of tourists on the tourist scenic area from the tourism evaluation platform, divide the review data by a fixed time period to form a text set for each time period.

[0076] Step 2, text vectorization representation. For the text set of review texts obtained by dividing each time period, perform text vectorization processing;

[0077] Use a word segmentation algorithm to segment each comment and remove stop words;

[0078] Use the Word2Vec model to calculate the average of the vector representations of all words in the comments of each time period to obtain a high-dimensional semantic vector a representing the overall semantics of the comment set for that day;

[0079] Store the obtained semantic vector a for future use.

[0080] Step 3, similarity calculation and judgment;

[0081] Starting from the second time period, calculate the similarity between the comment semantic vector a(t) in the current time period and the comment semantic vector a(t - 1) in the previous time period.

[0082] If CS is less than θ, perform a complete sentiment analysis, trend, and stability calculation;

[0083] If the similarity CS is greater than or equal to θ, fine-tune the indicators according to the above formula without completely recalculating.

[0084] Step 4, CS is greater than or equal to θ, obtain the trend direction and amplitude Δ;

[0085] Classify the comments using a sentiment dictionary: Each comment is classified into three categories: positive, neutral, and negative

[0086] Calculate the proportion of positive comments in the current time period and save it for the calculation of the next time period. The formula for calculating the proportion of positive comments is as follows:

[0087]

[0088] Among them, N is the total number of comments in the current time period, N positive is the number of positive comments in the current time period;

[0089] Calculate the trend change amplitude Δ based on the proportion of positive comments in the current window and the previous window:

[0090] Δ = P t - P t-1

[0091] Among them, P t-1 is the proportion of positive comments in the previous time period;

[0092] If the proportion of positive comments in the current time period is greater than the proportion in the previous time period, the trend direction is upward and Δ is a positive number;

[0093] If the proportion of positive comments in the current time period is less than the proportion in the previous time period, the trend direction is downward and Δ is a negative number;

[0094] If the proportion of positive comments in the current time period is equal to the proportion in the previous time period, then Δ is 0;

[0095] Step 5, calculate the positive attention trend score and the word-of-mouth stability score, and obtain the virtual correlation quotient.

[0096] Positive attention trend score adjustment formula:

[0097] T t = T t-1 +(1 - cs)×Δ

[0098] Among them, Tt is the positive attention trend score for the current time period, T t-1 Score the positive attention trend in the previous time period;

[0099] Word of mouth stability score adjustment formula:

[0100] S t =S t-1 ×cs

[0101] Where S t is the positive attention trend score for the current time period, S t-1 Score the positive attention trend in the previous time period;

[0102] The positive attention trend score and word-of-mouth stability score are processed together to obtain the virtual relevance quotient.

[0103] Step 6, forming a scenic spot operation signal according to the virtual correlation quotient of each scenic spot;

[0104] The scenic spots are ranked according to the virtual relevance quotient and the virtual resilience index is obtained, which is then compared with the risk threshold.

[0105] If the virtual resilience index is greater than or equal to the risk threshold, a stable signal is issued;

[0106] If the virtual resilience index is less than or equal to the risk threshold, a risk signal is issued.

[0107] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. An ecotourism resource assessment method, characterized in that: The steps include: (1) Data collection and preprocessing: Using web crawler technology, we regularly obtain online text review data of tourists on tourist attractions from the tourism evaluation platform, and divide the review data into fixed time periods to form a text collection for each time period; (2) Text vectorization: perform word segmentation on the comment set for each time period divided in step (1), remove stop words, and use the Word2Vec model to average the vector representations of all words in the time period to obtain a high-dimensional semantic vector a that represents the overall semantics of the time period; (3) Similarity calculation and judgment: Starting from the second time period, the cosine similarity CS between the comment semantic vector of the current time period and the comment semantic vector of the previous time period is calculated; (4) Compare the similarity CS and the preset threshold θ to determine the calculation formula for the positive attention trend score and the word-of-mouth stability score. (5) Based on the positive attention trend score and the word-of-mouth stability score, a virtual correlation quotient is calculated to conduct an operational risk assessment of the tourist attraction.

2. The ecotourism resource evaluation method according to claim 1, characterized in that: The calculation formula of similarity CS is as follows: in is the semantic vector of comments in the current time period; is the semantic vector of comments in the previous period.

3. The ecotourism resource evaluation method according to claim 1, characterized in that: If the similarity CS is less than the preset threshold θ, a complete sentiment analysis is performed on the current time period to calculate the proportion of positive comments, and the sentiment trend change amplitude Δ is calculated in combination with the proportion of positive comments in the previous time period, and then the positive attention trend score and word-of-mouth stability score of the current time period are calculated; If the similarity CS is greater than or equal to the preset threshold θ, the positive attention trend score and the word-of-mouth stability score are fine-tuned according to the emotional trend change amplitude Δ and the similarity CS without completely recalculating.

4. The ecotourism resource evaluation method according to claim 1, characterized in that: The formula for calculating the positive review ratio is: Where N is the total number of comments in the current time period, N positive The number of positive comments in the current time period.

5. The ecotourism resource evaluation method according to claim 1, characterized in that: The calculation formula for the emotional trend change amplitude Δ is: Δ=P t -P t-1 Where P t is the proportion of positive comments in the current time period, P t-1 is the proportion of positive comments in the previous time period; If Δ>0, it means the sentiment trend is rising; If Δ<0, it means the sentiment trend is decreasing; If Δ=0, it means the sentiment trend is stable.

6. The ecotourism resource evaluation method according to claim 3, characterized in that: The adjustment formula for the positive attention trend score is: T t =T t-1 +(1-CS)×Δ Where, T t is the positive attention trend score for the current time period, T t-1 Score for the positive attention trend in the previous time period.

7. The ecotourism resource evaluation method according to claim 3, characterized in that: The word-of-mouth stability score adjustment formula is: S t =S t-1 ×cs In the formula, S t is the word-of-mouth stability score for the current period, S t-1 is the word-of-mouth stability score of the previous time period, and CS is the cosine similarity between the texts in the current time period and the previous time period.

8. A scenic spot sustainable development planning system, characterized in that: Use the ecotourism resource assessment method described in any one of claims 1-7.

9. The sustainable development planning system according to claim 8, characterized in that: Generate a scenic spot operation risk assessment report based on the virtual correlation quotient to assist scenic spot management decision-making.

10. The sustainable development planning system according to claim 9, characterized in that: The method according to claim 6 is characterized in that the scenic area operation risk level classification standard is specifically: When the positive trend score is <0.3 and the stability score is <0.4, it is defined as high risk; When the positive trend score is between 0.3 and 0.6 and the stability score is between 0.4 and 0.7, it is defined as medium risk; Low risk was defined when the positive trend score was >0.6 and the stability score was >0.7.

Images