CN113298366A - Tourism performance service value evaluation method - Google Patents
Tourism performance service value evaluation method Download PDFInfo
- Publication number
- CN113298366A CN113298366A CN202110515659.0A CN202110515659A CN113298366A CN 113298366 A CN113298366 A CN 113298366A CN 202110515659 A CN202110515659 A CN 202110515659A CN 113298366 A CN113298366 A CN 113298366A
- Authority
- CN
- China
- Prior art keywords
- characteristic
- emotion
- value
- word
- travel performance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 34
- 230000008451 emotion Effects 0.000 claims abstract description 90
- 238000000034 method Methods 0.000 claims abstract description 37
- 238000003062 neural network model Methods 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 18
- 238000004458 analytical method Methods 0.000 claims abstract description 16
- 238000010219 correlation analysis Methods 0.000 claims abstract description 11
- 239000003607 modifier Substances 0.000 claims description 24
- 230000008447 perception Effects 0.000 claims description 14
- 238000012216 screening Methods 0.000 claims description 14
- 230000009467 reduction Effects 0.000 claims description 11
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 9
- 230000002452 interceptive effect Effects 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 7
- 238000004140 cleaning Methods 0.000 claims description 6
- 230000002996 emotional effect Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 210000002569 neuron Anatomy 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 230000007935 neutral effect Effects 0.000 claims description 3
- 238000012935 Averaging Methods 0.000 claims description 2
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 description 10
- 238000011160 research Methods 0.000 description 6
- 238000003058 natural language processing Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000012552 review Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000013210 evaluation model Methods 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 230000001149 cognitive effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000007480 spreading Effects 0.000 description 2
- 206010013911 Dysgeusia Diseases 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 230000003796 beauty Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000981 bystander Effects 0.000 description 1
- 230000001364 causal effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000556 factor analysis Methods 0.000 description 1
- 230000008713 feedback mechanism Effects 0.000 description 1
- 238000003958 fumigation Methods 0.000 description 1
- 238000007654 immersion Methods 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000003340 mental effect Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 230000035807 sensation Effects 0.000 description 1
- 210000000697 sensory organ Anatomy 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000000859 sublimation Methods 0.000 description 1
- 230000008022 sublimation Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0282—Rating or review of business operators or products
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/14—Travel agencies
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Accounting & Taxation (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Finance (AREA)
- Marketing (AREA)
- Computational Linguistics (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biomedical Technology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Educational Administration (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
Abstract
The invention belongs to the technical field of travel performance value evaluation, and relates to a travel performance service value evaluation method, which comprises the following steps: step 1: obtaining the influence factor indexes of the service value of the tourism performance; step 2: obtaining comment data and comprehensive scores; and step 3: obtaining a characteristic word list of the value of the travel performance service; and 4, step 4: extracting a characteristic sentence; and 5: training an LSTM model; acquiring an emotion value; step 6: calculating the grey correlation degree; and 7: determining a topological structure of a grey neural network model; and 8: training a grey neural network model; and step 9: and obtaining the evaluation value of the specific travel performance drama service. The method and the system add culture elements to the assessment of the service value of the travel performance. And an LSTM fine-grained emotion analysis and grey neural network model are provided, so that the method for evaluating the service value of the tourism performance is enriched, and the evaluation is more intelligent and accurate. And the grey correlation analysis is used for determining an index system, so that subjective and objective combination is achieved, and the method is more scientific.
Description
Technical Field
The invention belongs to the technical field of tourism performance value evaluation, relates to a tourism performance service value evaluation method, and particularly relates to a tourism performance value evaluation method based on LSTM (long and short term memory artificial neural network) fine-grained emotion analysis and a gray neural network.
Background
The tourism performance is used as an important component of cultural tourism, is a carrier for 'activating' historical culture, deeply integrates culture and tourism experience, and is popular with people in a high-participation and high-experience tourism mode. The tourism performance service has economic value and wide social value. The tourism performance service can highlight and transmit the characteristic regional culture of the tourist site and create new cultural business cards; can improve the spiritual pursuit and cultural bound of people and arouse the people's true and beautiful pursuit. However, the major national policies such as "internet +" are gradually promoted, and the tourism performance is developed vigorously by means of on-line ticketing channels. However, as the evaluation feedback mechanism is still to be perfected, more and more tourism performance products are light in content and heavy in form, insufficient in subject innovation and lack in culture taste after prosperous appearance. Meanwhile, as the number of products of the tourism performance increases, the public praise of tourists on the tourism performance is different, but the requirement on the quality of the tourists is higher. Therefore, under the current situation, how to objectively measure and evaluate the value of the travel performance service is the key point for the product to fully mine the association and contribution of different factors of the service to the value and further to know the culture quality so that the travel performance product goes from 'extensive' to 'fine'; the method is also an important reference for ranking and recommending the performance products by the platform service provider; but also important reference indexes for quality supervision of third-party institutions and promotion of benign competition in performance markets.
Traditional tourist performance value assessment perspectives include: the tourist perception, the customer satisfaction and the service quality. The perception angle of the tourists is to utilize the perception quality and perception value theory to discuss the internal information acquisition of the sensory organs of the tourists on the performance objects and the performance environment. In addition, according to the customer satisfaction index theory taught by Fornell, an evaluation model of the customer satisfaction index of the tourist site is constructed by referring to the causal relationship model principle of the ACSI (American customer satisfaction index) model. In addition, due to the deep development of service quality research, key attributes and measurement indexes of the travel service quality are searched, and the development of travel destinations is accelerated to become a new research aspect. The value evaluation of the travel performance should not only be performed from the aspects of performance content and effect, performance site and environment, etc., but also be a factor of cultural value as a paradigm of text-travel fusion development, belonging to the category of cultural products. At present, the method for evaluating the value of the tourism performance mainly comprises statistical methods such as rooting theory, scale method, factor analysis and the like. The accuracy of such qualitative research methods is far from sufficient. Therefore, the intelligent quantitative research by using a text analysis method becomes a new trend, but the current research only analyzes the ratio of high-frequency words extracted by ROST (high-frequency vocabulary statistics software) and positive emotional sentences of the whole text, so that the value of the travel performance is evaluated quite rough and inaccurate. Therefore, the intelligent and accurate universality service value evaluation decision of enterprises and platforms can be met only by fully mining the comment text in the network platform and analyzing the emotional connotation of text data.
Disclosure of Invention
The invention aims to provide a method for evaluating the service value of a tourism performance based on LSTM fine-grained emotion analysis and a grey neural network, and a model for evaluating the service value of the tourism performance based on LSTM fine-grained emotion analysis and the grey neural network is constructed so as to solve the problems in the background technology.
The invention is realized by the following technical scheme:
a value evaluation method for a travel performance service is based on LSTM fine-grained emotion analysis and a grey neural network, and comprises the following steps:
step 1: analyzing influence factors of the travel performance service value on the basis of an interactive ceremony chain theory to obtain influence factor indexes of the travel performance service value;
step 2: obtaining comment data of the travel performance drama and comprehensive grading data corresponding to the travel performance drama;
and step 3: screening effective comments, and extracting characteristic words to obtain a feature word list of the value of the travel performance service;
and 4, step 4: extracting a characteristic sentence;
and 5: training an LSTM model; obtaining the emotion value of the index of the tourism performance drama;
step 6: calculating the grey correlation degree of a sequence formed by the sentiment value of each index sequence and the comprehensive scoring data by adopting a grey correlation analysis method, and removing the index sequence with the minimum correlation degree to form the index sequence after dimensionality reduction;
and 7: determining a topological structure of a grey neural network model;
and 8: initializing parameters of a grey neural network model, taking the index sequence subjected to dimensionality reduction as an input signal of the grey neural network model, taking corresponding comprehensive grading data as an output signal of the grey neural network model, and training the grey neural network model;
and step 9: aiming at a specific travel performance drama service, obtaining the emotion value of the index of the travel performance drama service by utilizing the steps 2-5 according to the index after dimensionality reduction;
and inputting the emotion value of the index of the travel performance drama service into the trained grey neural network model to obtain a predicted comprehensive score which is used as the evaluation value of the travel performance drama service.
On the basis of the technical scheme, the specific steps of the step 1 are as follows: the method is characterized in that the method is based on the interactive ceremony chain theory, and the method is based on the interactive ceremony chain theory. The tourism performance integrates culture elements in the whole process from creation to performance, forms instant shared sense of existence with audiences, spreads the emotion given by culture, obtains membership and emotion energy associated with cognitive symbols, and realizes efficient spreading of cultural connotation. The influence factors of the tourism performance service value are constructed from three dimensions of creation, perception and fitting, and comprise the following steps: 3 primary indexes and 8 secondary indexes;
the primary indicators include: an creation layer, a perception layer and a conjunction layer;
the authoring layer comprises the following two-level indicators: branding, originality, and coverage;
the perception layer comprises the following two-level indexes: artistry, serviceability, and emotions;
the conjunction layer comprises the following two-level indexes: satisfaction and loyalty;
the first-level index and the second-level index are both: and (4) influence factor indexes of the service value of the travel performance.
On the basis of the technical scheme, the brandiness refers to: whether the tourists choose to watch the tourism performance products or not according to the known name of the creation team;
the originality refers to: in the creation process of the performance product, whether innovative performance contents and performance forms are adopted or not and the degree of creativity attracting tourists;
the coverage means that: coverage degree of tourism products on regional knowledge, historical stories and folk-custom folk wind from the culture layer;
the artistry refers to: the value of performance products is measured from the aspects of stage backgrounds, lighting effects, clothing characteristics, dance arrangement, music effects, scientific and technological application, script scenarios, actor performance and the like;
the serviceability means that: the value of the additional service beyond the performance product to the audience is shown from the site environment, the basic service facility, the scenic spot supporting facility, the time arrangement, the entrance and exit, the ticket booking and taking, the traffic condition and the like;
the emotion is: the situation perception of tourists on the performance products, the subjective experience obtained through objective performance, and further the sublimation of mental levels caused by cultural fumigation and pottery;
the satisfaction refers to: whether the feeling of the tourist meets the expected expectation or not and whether the feeling of identity can be generated or not;
the loyalty refers to: whether the tourists have positive public praise or not and the desire of the tourists to obtain the tourism again.
On the basis of the technical scheme, the specific steps of the step 2 are as follows: and collecting comment data of the travel performance drama and comprehensive grading data corresponding to the travel performance drama, which are provided by the network platform, by using a crawler technology.
On the basis of the above technical solution, the network platform includes: a distance-carrying network.
On the basis of the technical scheme, the specific steps of the step 3 are as follows: firstly, carrying out data preprocessing work such as duplicate removal and data cleaning on the comment data to obtain effective comments;
the data cleaning is as follows: when the comment data under the user name is empty, deleting the empty comment data;
then, extracting feature words:
the method comprises the following steps: performing word segmentation on the effective comments obtained by the last step by using a jieba (jieba) tool to obtain entries of all the effective comments;
and then extracting key words by using a TF-IDF (word frequency-inverse document frequency) algorithm, wherein the method specifically comprises the following steps: the calculation is carried out by using the formulas (1), (2) and (3),
wherein, TFωThe term frequency of the entry omega;
wherein, the IDF is the reverse file frequency; if the number of the effective comments containing a certain entry is less, the IDF of the entry is larger, and the entry has good category distinguishing capability;
TFIDF＝TFω*IDF (3)
wherein, TFIDF is: word frequency-inverse document frequency; screening entries with high TFIDF values as keywords (for example, screening entries with TFIDF values greater than 20 as keywords); the screening tends to filter out common words and retain relatively important words;
then, carrying out word frequency statistics on the keywords by using a Counter library to obtain candidate characteristic words;
finally, according to the influence factor indexes of the travel performance service value, after manual screening and identification, classifying the candidate characteristic words in a grading way to obtain a travel performance service value characteristic word list;
the Counter library is a library in python, belonging to a subclass of dictionaries, elements are stored as keys of the dictionary, and the number of occurrences of a key is stored as a corresponding numerical value.
On the basis of the technical scheme, the characteristic sentence comprises: displaying the characteristic sentences and the implicit characteristic sentences;
the specific steps of the step 4 are as follows:
firstly, extracting an explicit characteristic sentence;
the explicit characteristic sentence extraction is as follows: traversing all the entries of the effective comments word by word, comparing the entries with a characteristic vocabulary of the value of the tourism performance service, taking the matched characteristic words as the characteristic attributes of the effective comments where the entries are located, and simultaneously extracting the effective comments with the characteristic attributes as explicit characteristic sentences;
the second step is that: extracting an implicit characteristic sentence;
performing dependency sentence pattern analysis on the extracted explicit characteristic sentence by using a Standfordcore NLP platform, and extracting a modifier of the explicit characteristic sentence;
the specific steps of extracting the modifiers of the explicit characteristic sentences are as follows: performing word-by-word traversal on the entries of the explicit characteristic sentences, comparing the entries with the modified words of the HowNet emotion dictionary, and taking the matched modified words as modified words of the explicit characteristic sentences where the entries are located;
the HowNet emotion dictionary comprises: adjectives, nouns, verbs, adverbs, and combinations thereof;
aiming at the explicit characteristic sentences matched with the modifiers, the following processing is carried out:
taking the feature words of the display feature sentences as leading words, taking the modifying words of the display feature sentences as emotion words, and constructing attribute feature-emotion word pairs so as to obtain attribute feature-emotion word-attribute emotion word pair weights;
the attribute characteristics are as follows: a dominant word;
and recording the attribute emotion word pair weight as: SQ, calculated according to equation (4),
for the characteristic sentences which are not matched with the characteristic words, performing word-by-word traversal on the entries of the characteristic sentences, and comparing the characteristic sentences with the modified words of the HowNet emotion dictionary;
when the characteristic sentence which is not matched with the characteristic word is not matched with the modifier, deleting the characteristic sentence;
when the characteristic sentences which are not matched with the characteristic words are matched with the modifiers, the matched modifiers are used as modifiers of the characteristic sentences of the entry, and the modifiers are used as emotion words;
then, according to the attribute feature-emotion word-attribute emotion word pair weight calculated by the formula (4), selecting the attribute feature with the maximum attribute emotion word pair weight as the feature word of the feature sentence which is not matched with the feature word according to the emotion word in the feature sentence which is not matched with the feature word;
taking the characteristic sentence of the obtained characteristic words which is not matched with the characteristic words as an implicit characteristic sentence;
the standard NLP platform is a natural language processing toolkit, and integrates a plurality of very practical functions, including word segmentation, part of speech tagging, syntactic analysis and the like; the Standford NLP platform is not a deep learning framework, but a trained model, and can be analogized to software; the stanford NLP platform is written in Java language and has a python interface.
On the basis of the technical scheme, the specific steps of the step 5 are as follows:
firstly, manually marking emotion polarity on a characteristic sentence;
the emotion polarities are labeled as: the positive emotion is marked as 1, the negative emotion is marked as-1, and the neutral emotion is marked as 0;
then, converting the characteristic sentence into a word vector by using word2 vec;
and finally, inputting the word vector, the characteristic words corresponding to the characteristic sentences, the secondary indexes of the characteristic words corresponding to the characteristic sentences, the primary indexes of the characteristic words corresponding to the characteristic sentences and the emotion polarity labels corresponding to the characteristic sentences into an LSTM model, and training the LSTM model to finally obtain the emotion values of the secondary indexes of the travel performance dramas.
On the basis of the technical scheme, the specific parameters of the LSTM model are set as follows: the activation function is tan h function; the word vector dimension value is set to 100; the batch processing amount of data is 32; the window size is 7; the training period is 4; the number of iterations is 1; the neuron discard rate was 0.5.
On the basis of the technical scheme, the specific steps of the step 6 are as follows:
calculating the gray correlation degree of each secondary index subsequence and the scoring parent sequence by adopting gray correlation analysis;
the comprehensive scoring data of m travel performance dramas provided by a network platform is collected through a crawler technology;
the scoring parent sequence is as follows: comprehensive scoring data of the m travel performance dramas;
the emotional values of the m tourism performance dramas of each secondary index form a secondary index subsequence;
the subsequence of each secondary index is as follows: the emotion value sequences of 8 secondary indexes formed by the m travel performance dramas,
carrying out non-dimensionalization operation on the secondary index subsequence and the scoring parent sequence by using an averaging method, specifically calculating according to the formula (5),
wherein x' (k) is: the original value of the kth element in each secondary index subsequence or score mother sequence; x (k) is: the value of the k-th element in each secondary index subsequence or scoring mother sequence after equalization; then, according to the formulas (6) and (7), the gray correlation coefficient xi of each secondary index subsequence and the scoring mother sequence is respectively calculatedijAnd degree of grey correlation Rij，
Wherein, | xi(k)-xj(k) The absolute difference value of the ith element in the secondary index subsequence after equalization and the jth element in the scoring mother sequence after equalization is | shown; miniminjRepresents: taking the minimum absolute difference, max, among all absolute differencesimaxjRepresents: taking out the maximum absolute difference value from all the absolute difference values; rho is a resolution coefficient, and is usually 0.5;
the specific calculation steps of the grey correlation coefficient are as follows: firstly, fixing the jth element in the scoring mother sequence, and then calculating a gray correlation coefficient according to a formula (6) by using all elements in the secondary index subsequence and the jth element; by analogy, replacing elements in the scoring mother sequence, and calculating m gray correlation coefficients; calculating the average value of the m gray correlation coefficients according to a formula (7) to serve as the gray correlation degree of the corresponding secondary index and the scoring mother sequence;
and sequencing the gray correlation degrees through gray correlation analysis, removing the service secondary index sequence with the minimum correlation degree to form a secondary index sequence after dimensionality reduction, and judging the key influence factors of the service value of the tourism performance.
On the basis of the technical scheme, the topological structure of the grey neural network model consists of an LA layer, an LB layer, an LC layer and an LD layer which are sequentially connected with one another, wherein the LA layer is provided with 1 node, the LB layer is provided with 1 node, the LC layer is provided with 8 nodes, and the LD layer is provided with 1 node; respectively corresponding 7 input nodes of the LC layer to emotion values of secondary indexes of brand, coverage, originality, loyalty, satisfaction, emotion and artistry;
on the basis of the technical scheme, the parameters of the gray neural network model are set as follows: the maximum training times is 50000 times; the learning rate is 0.05; target error is e-7。
The invention has the following beneficial technical effects:
the method creatively adds the culture elements into the evaluation system of the travel performance service value, and constructs the travel performance service value evaluation system under a network platform comprising 3 primary indexes and 7 secondary indexes of an creation layer, a perception layer and a fitting layer. The service value evaluation model based on LSTM fine-grained emotion analysis and grey neural network prediction is provided on the basis of enriching the influence factors of the service value of the travel performance, and enriches the service value evaluation method of the travel performance, so that the service value evaluation is more intelligent and accurate. In addition, the gray correlation analysis is utilized to determine the index system, so that the determination process of the index system is subjectively and objectively combined, and the method is more scientific.
Drawings
The invention has the following drawings:
FIG. 1 is a flow chart of the method for evaluating the value of the travel performance service according to the present invention.
FIG. 2 is a flowchart illustrating an embodiment of the method for evaluating a value of a travel performance service according to the present invention.
Detailed Description
Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
In order to solve the problems in the prior art, the invention provides a tourism performance service value evaluation model based on LSTM fine-grained emotion analysis and a gray neural network, an influence factor system of the tourism performance service value evaluation based on an interactive ceremony chain theory is constructed, the model adopts an LSTM fine-grained emotion analysis method to calculate corresponding emotion values for the crawled comment information, and after gray correlation screening indexes are carried out, the gray neural network model is used for training to obtain a tourism performance service value evaluation result.
As shown in fig. 1, a schematic flow chart of the service value evaluation of the tourism performance based on LSTM fine-grained emotion analysis and gray neural network is implemented, which includes the following steps:
step 1: analyzing influence factors of the travel performance service value on the basis of an interactive ceremony chain theory to obtain influence factor indexes of the travel performance service value;
step 2: obtaining comment data of the travel performance drama and comment data corresponding to the travel performance drama;
and step 3: screening effective comments, and extracting characteristic words to obtain a feature word list of the value of the travel performance service;
and 4, step 4: extracting a characteristic sentence;
and 5: training an LSTM model to obtain an emotion value of an index of a travel performance drama; obtaining influence factor sentiment values based on LSTM fine-grained sentiment analysis;
step 6: calculating the grey correlation degree of the sequence formed by the sentiment value of each index sequence and the comprehensive scoring data by adopting a Grey Correlation Analysis (GCA) method, and removing the index sequence with the minimum correlation degree to form the index sequence after dimensionality reduction; determining influence factors by adopting a grey correlation analysis method;
and 7: determining a topological structure of a grey neural network model;
and 8: initializing parameters of a gray neural network model, taking the index sequence after dimensionality reduction as an input signal of the gray neural network model, taking corresponding comprehensive scoring data (namely, a journey platform scoring shown in figure 2) as an output signal of the gray neural network model, and training the gray neural network model;
and step 9: aiming at a specific travel performance drama service, obtaining the emotion value of the index of the travel performance drama service by utilizing the steps 2-5 according to the index after dimensionality reduction;
and inputting the emotion value of the index of the travel performance drama service into the trained grey neural network model to obtain a predicted comprehensive score, wherein the predicted comprehensive score is used as the evaluation value of the travel performance drama service, namely the value of the travel performance drama service is predicted by using the grey neural network model.
The specific steps of the step 1 are as follows: the method is characterized in that the method is based on the interactive ceremony chain theory, and the method is based on the interactive ceremony chain theory. The tourism performance integrates culture elements in the whole process from creation to performance, forms instant shared sense of existence with audiences, spreads the emotion given by culture, obtains membership and emotion energy associated with cognitive symbols, and realizes efficient spreading of cultural connotation. And constructing a tourism performance service value influence factor system containing 3 first-level indexes and 8 second-level indexes from three dimensions of creation, perception and fitting. The first-level indexes are respectively: creation layer, perception layer and agree with the layer, the second grade index is respectively: branding, originality, coverage, artistry, serviceability, emotionality, satisfaction, and loyalty.
The specific steps of the step 2 are as follows: and collecting comment data of the travel performance drama and comprehensive grading data corresponding to the travel performance drama, which are provided by the travel network, by using a crawler technology. The travel is taken as a leading online ticketing service company in China, the number of users is large, the types of tourism performance are comprehensive, the platform is high in openness, and the users can make evaluations anytime and anywhere and express own attitudes and emotions. According to the entrance lists of the Chinese travel performance ticket house ranking lists in 2017 and 2018 of the track performance, selecting travel performance dramas with the number of reviews being more than 400 as research objects, and crawling the review time, the review user name and the review data content by using a web crawler technology. Finally, the data are shown in 'Re-Chang Ping-Yuan' (Chang-Yuan) ',' Re-Dunhuang '(Huang)', 'Huang-Liu Sanjie', 'impression-Dahongpao', 'impression-Putuo', 'impression-Wulong-pending' (Wu-Huang-Lizhong) ',' Song-Cheng Qiaogu ',' Sanqian-Gu ',' Lijiang-Qiaogu '(Wu-Lin-Gu', 'Jiuzhai-Gu' (Wu-Jian-Gu ',' Shang-Cheng-Gong '(Yunan-Wan)' Yunan impression, Chang-Lin-Shu-Wan-Tu-Wan-Tu-Su-Wan-Tu-Wan-Su-Wan-Su-Tu-Su-Wan.
The specific steps of the step 3 are as follows: firstly, data preprocessing work such as duplicate removal and data cleaning of the same text data is carried out on the comment data, and therefore effective comments are obtained.
The data cleaning is as follows: when the comment data under the user name is empty, deleting the empty comment data;
then, extracting feature words;
performing word segmentation on the effective comments obtained by the last step by using a jieba (jieba) tool to obtain entries of all the effective comments;
and then extracting key words by using a TF-IDF (word frequency-inverse document frequency) algorithm, wherein the method specifically comprises the following steps: the calculation is carried out by using the formulas (1), (2) and (3),
wherein, TFωThe term frequency of the entry omega;
wherein, the IDF is the reverse file frequency; if the number of the effective comments containing a certain entry is less, the IDF of the entry is larger, and the entry has good category distinguishing capability;
TFIDF＝TFω*IDF (3)
wherein, TFIDF is: word frequency-inverse document frequency; screening entries with high TFIDF values as keywords (for example, screening entries with TFIDF values greater than 20 as keywords); the above screening tends to filter out common words, leaving words of relative importance.
And finally, carrying out classification on the candidate characteristic words through manual screening and identification according to the influence factor indexes of the travel performance service value to obtain a travel performance service value characteristic word list.
The Counter library is a library in python, belonging to a subclass of dictionaries, elements are stored as keys of the dictionary, and the number of occurrences of a key is stored as a corresponding numerical value.
The branded feature words comprise: zhangyezhu, lao zi, director, zhanglead, drama, works, and guide;
the inventive feature words include: special effects, designs, themes, styles, modes, forms, expression forms, novelty, originality, highlight, uniqueness and characteristics;
the inclusive feature words include: story, story nature, love story, plot, story plot, drama, story, legend, tradition, knowledge, classic, local, native, original, folk custom, folk wind, ethnic, folk culture, geomantic affection, folk affection, minority ethnic, ethnic characteristics, culture, humanistic, cultural background implication, meaning, connotation, narration, fusion, combination;
the artistic characteristic words comprise: performances, actors, deductions, performances, actors' formation, acrobatics, matches, formation, population, performers, real persons, crowd actors, dancing actors, performances, staffs, science, technology, high technology, props, scenes, reality, backgrounds, pictures, scenes, stages, scenery, stage effects, stage colors, scenes, night scenes, beauty scenes, lighting, acoustoelectric, lighting effects, stage lighting, light shadow, dance design, actions, vision, appreciation, visual effects, audio-visual effects, viewing and hearing, performance, appreciation, sound effects, music, sound effects, singing sounds, sound, songs, breath, songs, clothing, and apparel;
the service characteristic words comprise: taking, buying, booking, ordering, booking, buying, drawing, sending, changing, booking, returning, checking, selling, taking, queuing, entering, leaving, scattering, environment, facility, whole scene, field number, outfield, park, field, theater, scenic spot, doorway, window, auditorium, VIP seat, service attitude, staff, tour guide, customer service, traffic, taxi taking, and pick-up;
the emotional characteristic words comprise: sense of participation, personal sensation, unclear sight, inaudible hearing, immersion, appeal, presence, experience, experiential style, feeling, exclamation, feeling, impact, look and feel, indecision, passing through, invisible, experience, aftertaste, understanding, substituting, explaining, bystander, narration, scene, situation, emotion, mood, meaning;
the satisfactory characteristic words comprise: expectation, surprise, credence, yes, no, good, bad, appraisal;
the loyalty feature words include: strong recommendation, introduction, must go, card punch, father mother, one family, old mother, old man, daughter, son, child, family, old man, child, parent.
The specific steps of the step 4 are as follows:
firstly, explicit characteristic sentence extraction.
The explicit characteristic sentence extraction is as follows: traversing all the entries of the effective comments word by word, comparing the entries with a characteristic vocabulary of the value of the tourism performance service, taking the matched characteristic words as the characteristic attributes of the effective comments where the entries are located, and simultaneously extracting the effective comments with the characteristic attributes as explicit characteristic sentences;
the second step is that: and (4) extracting the implicit characteristic sentences.
Performing dependency sentence pattern analysis on the extracted explicit characteristic sentence by using a Standard NLP platform, and extracting a modifier of the explicit characteristic sentence;
the specific steps of extracting the modifiers of the explicit characteristic sentences are as follows: performing word-by-word traversal on the entries of the explicit characteristic sentences, comparing the entries with the modified words of the HowNet emotion dictionary, and taking the matched modified words as modified words of the explicit characteristic sentences where the entries are located;
the HowNet emotion dictionary comprises: adjectives, nouns, verbs, adverbs, and combinations thereof;
aiming at the explicit characteristic sentences matched with the modifiers, the following processing is carried out:
taking the feature words of the display feature sentences as leading words, taking the modifying words of the display feature sentences as emotion words, and constructing attribute feature-emotion word pairs so as to obtain attribute feature-emotion word-attribute emotion word pair weights;
the attribute characteristics are as follows: a dominant word;
and recording the attribute emotion word pair weight as: SQ, calculated according to equation (4),
for the characteristic sentences which are not matched with the characteristic words, performing word-by-word traversal on the entries of the characteristic sentences, and comparing the characteristic sentences with the modified words of the HowNet emotion dictionary;
when the characteristic sentence which is not matched with the characteristic word is not matched with the modifier, deleting the characteristic sentence;
when the characteristic sentences which are not matched with the characteristic words are matched with the modifiers, the matched modifiers are used as modifiers of the characteristic sentences of the entry, and the modifiers are used as emotion words;
then, according to the attribute feature-emotion word-attribute emotion word pair weight calculated by the formula (4), selecting the attribute feature with the maximum attribute emotion word pair weight as the feature word of the feature sentence which is not matched with the feature word according to the emotion word in the feature sentence which is not matched with the feature word;
and taking the characteristic sentence which is not matched with the characteristic words and is obtained as the implicit characteristic sentence.
The specific steps of the step 5 are as follows: manually marking emotion polarity of each feature sentence aiming at the feature sentences extracted in the last step (positive emotion is marked as 1, negative emotion is marked as-1, and neutral emotion is marked as 0);
then, converting the characteristic sentence into a word vector by using word2 vec;
and finally, inputting the word vector, the characteristic words corresponding to the characteristic sentences, the secondary indexes of the characteristic words corresponding to the characteristic sentences, the primary indexes of the characteristic words corresponding to the characteristic sentences and the emotion polarity labels corresponding to the characteristic sentences into an LSTM model, and training the LSTM model to finally obtain the emotion values of the secondary indexes of the travel performance dramas.
Finally obtaining the emotional values of the secondary indexes of different tourism performance dramas shown in the table 1,
TABLE 1 Emotion value table of second-level indexes of different travel performance dramas
The specific parameters of the LSTM model are set as: the activation function is tan h function; the word vector dimension value is set to 100; the batch processing amount of the data is 32, namely 32 samples are selected as input each time; the window size is 7; the training period is 4; the number of iterations is 1; the neuron drop rate is 0.5, i.e., the probability of neurons being temporarily dropped from the network to weaken the joint fitness between neuron nodes is 0.5.
After the model is trained, the evaluation indexes of the detection experiment result are respectively represented by accuracy rate, recall rate and F1 value, wherein the accuracy rate refers to the accuracy of model classification for the prediction result, the recall rate refers to the proportion of correct classification for the original sample, the F1 is used for carrying out comprehensive evaluation on the accuracy rate, the recall rate and the F1 value are respectively: 89.70%, 85.83% and 87.23%.
Step 6 is further: the grey correlation analysis is used for calculating the grey correlation degree of the emotion value sequence and the scoring sequence of each secondary index, as shown in table 2,
TABLE 2 Grey relevance Table
From the above table, the influence degrees of the factors on the service value of the travel performance are ranked as artistry, satisfaction, loyalty, originality, coverage, emotion and brand, the service is provided, the influence degree on the service value of the travel performance is artistry, the minimum service is provided, the association degree is only 0.643, the difference between the influence degree and other influence factors is large, and the service is removed from the secondary index system.
Determining the topological structure of the grey neural network model as follows: 1-1-8-1, specifically, the LED node comprises an LA layer, an LB layer, an LC layer and an LD layer which are sequentially connected with one another, wherein the LA layer is provided with 1 node, the LB layer is provided with 1 node, the LC layer is provided with 8 nodes, and the LD layer is provided with 1 node; the 7 input nodes of the LC layer are respectively corresponding to the emotion values of the secondary indexes of brand, coverage, originality, loyalty, satisfaction, emotion and artistry, as shown in figure 2, y2(t)，…，yn(t) is: input of LC layer, omega11Is the connection weight of LA layer and LB layer, omega21,ω21,…,ω2nIs the connection weight of the LB layer and the LC layer, omega31,ω31,…,ω3nThe connection weight of the LC layer and the LD layer.
The parameters for the grey neural network model were set as follows: the maximum training times is 50000 times; the learning rate is 0.05; target error is e-7。
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and scope of the present invention are intended to be included therein.
Those not described in detail in this specification are within the knowledge of those skilled in the art.
Claims (10)
1. A method for evaluating the value of a travel performance service is characterized in that the method for evaluating the value of the travel performance service is based on LSTM fine-grained emotion analysis and a grey neural network, and comprises the following steps:
step 1: analyzing influence factors of the travel performance service value on the basis of an interactive ceremony chain theory to obtain influence factor indexes of the travel performance service value;
step 2: obtaining comment data of the travel performance drama and comprehensive grading data corresponding to the travel performance drama;
and step 3: screening effective comments, and extracting characteristic words to obtain a feature word list of the value of the travel performance service;
and 4, step 4: extracting a characteristic sentence;
and 5: training an LSTM model; obtaining the emotion value of the index of the tourism performance drama;
step 6: calculating the grey correlation degree of a sequence formed by the sentiment value of each index sequence and the comprehensive scoring data by adopting a grey correlation analysis method, and removing the index sequence with the minimum correlation degree to form the index sequence after dimensionality reduction;
and 7: determining a topological structure of a grey neural network model;
and 8: initializing parameters of a grey neural network model, taking the index sequence subjected to dimensionality reduction as an input signal of the grey neural network model, taking corresponding comprehensive grading data as an output signal of the grey neural network model, and training the grey neural network model;
and step 9: aiming at a specific travel performance drama service, obtaining the emotion value of the index of the travel performance drama service by utilizing the steps 2-5 according to the index after dimensionality reduction;
and inputting the emotion value of the index of the travel performance drama service into the trained grey neural network model to obtain a predicted comprehensive score which is used as the evaluation value of the travel performance drama service.
2. The method of claim 1, wherein the method comprises: the influence factors of the tourism performance service value are constructed from three dimensions of creation, perception and fitting, and comprise the following steps: 3 primary indexes and 8 secondary indexes;
the primary indicators include: an creation layer, a perception layer and a conjunction layer;
the authoring layer comprises the following two-level indicators: branding, originality, and coverage;
the perception layer comprises the following two-level indexes: artistry, serviceability, and emotions;
the conjunction layer comprises the following two-level indexes: satisfaction and loyalty;
the first-level index and the second-level index are both: and (4) influence factor indexes of the service value of the travel performance.
3. The travel performance service value evaluation method as set forth in claim 2, wherein: the specific steps of the step 2 are as follows: and collecting comment data of the travel performance drama and comprehensive grading data corresponding to the travel performance drama, which are provided by the network platform, by using a crawler technology.
4. The travel performance service value evaluation method as set forth in claim 3, wherein: the specific steps of the step 3 are as follows: firstly, carrying out the work of duplicate removal and data cleaning pretreatment on the same text data on the comment data so as to obtain effective comments;
the data cleaning is as follows: when the comment data under the user name is empty, deleting the empty comment data;
then, extracting feature words:
the method comprises the following steps: performing word segmentation on the effective comments obtained by the last step by using a jieba tool to obtain entries of all the effective comments;
and then extracting key words by using a TF-IDF algorithm, wherein the method specifically comprises the following steps: the calculation is carried out by using the formulas (1), (2) and (3),
wherein, TFωThe term frequency of the entry omega;
wherein, the IDF is the reverse file frequency;
TFIDF＝TFω*IDF (3)
wherein, TFIDF is: word frequency-inverse document frequency; screening entries with high numerical values of TFIDF as keywords;
then, carrying out word frequency statistics on the keywords by using a Counter library to obtain candidate characteristic words;
and finally, according to the influence factor indexes of the travel performance service value, carrying out manual screening and identification, classifying the candidate characteristic words in a grading way, and obtaining a travel performance service value characteristic word list.
5. The travel performance service value evaluation method as set forth in claim 4, wherein: the characteristic sentence comprises: displaying the characteristic sentences and the implicit characteristic sentences;
the specific steps of the step 4 are as follows:
firstly, extracting an explicit characteristic sentence;
the explicit characteristic sentence extraction is as follows: traversing all the entries of the effective comments word by word, comparing the entries with a characteristic vocabulary of the value of the tourism performance service, taking the matched characteristic words as the characteristic attributes of the effective comments where the entries are located, and simultaneously extracting the effective comments with the characteristic attributes as explicit characteristic sentences;
secondly, extracting implicit characteristic sentences;
performing dependency sentence pattern analysis on the extracted explicit characteristic sentence by using a Standfordcore NLP platform, and extracting a modifier of the explicit characteristic sentence;
the specific steps of extracting the modifiers of the explicit characteristic sentences are as follows: performing word-by-word traversal on the entries of the explicit characteristic sentences, comparing the entries with the modified words of the HowNet emotion dictionary, and taking the matched modified words as modified words of the explicit characteristic sentences where the entries are located;
aiming at the explicit characteristic sentences matched with the modifiers, the following processing is carried out:
taking the feature words of the display feature sentences as leading words, taking the modifying words of the display feature sentences as emotion words, and constructing attribute feature-emotion word pairs so as to obtain attribute feature-emotion word-attribute emotion word pair weights;
the attribute characteristics are as follows: a dominant word;
and recording the attribute emotion word pair weight as: SQ, calculated according to equation (4),
for the characteristic sentences which are not matched with the characteristic words, performing word-by-word traversal on the entries of the characteristic sentences, and comparing the characteristic sentences with the modified words of the HowNet emotion dictionary;
when the characteristic sentence which is not matched with the characteristic word is not matched with the modifier, deleting the characteristic sentence;
when the characteristic sentences which are not matched with the characteristic words are matched with the modifiers, the matched modifiers are used as modifiers of the characteristic sentences of the entry, and the modifiers are used as emotion words;
then, according to the attribute feature-emotion word-attribute emotion word pair weight calculated by the formula (4), selecting the attribute feature with the maximum attribute emotion word pair weight as the feature word of the feature sentence which is not matched with the feature word according to the emotion word in the feature sentence which is not matched with the feature word;
and taking the characteristic sentence which is not matched with the characteristic words and is obtained as the implicit characteristic sentence.
6. The travel performance service value evaluation method as set forth in claim 5, wherein: the specific steps of the step 5 are as follows:
firstly, manually marking emotion polarity on a characteristic sentence;
the emotion polarities are labeled as: the positive emotion is marked as 1, the negative emotion is marked as-1, and the neutral emotion is marked as 0;
then, converting the characteristic sentence into a word vector by using word2 vec;
and finally, inputting the word vector, the characteristic words corresponding to the characteristic sentences, the secondary indexes of the characteristic words corresponding to the characteristic sentences, the primary indexes of the characteristic words corresponding to the characteristic sentences and the emotion polarity labels corresponding to the characteristic sentences into an LSTM model, and training the LSTM model to finally obtain the emotion values of the secondary indexes of the travel performance dramas.
7. The travel performance service value evaluation method as set forth in claim 6, wherein: the specific parameters of the LSTM model are set as: the activation function is tan h function; the word vector dimension value is set to 100; the batch processing amount of data is 32; the window size is 7; the training period is 4; the number of iterations is 1; the neuron discard rate was 0.5.
8. The travel performance service value evaluation method as set forth in claim 7, wherein: the specific steps of the step 6 are as follows:
calculating the gray correlation degree of each secondary index subsequence and the scoring parent sequence by adopting gray correlation analysis;
the comprehensive scoring data of m travel performance dramas provided by a network platform is collected through a crawler technology;
the scoring parent sequence is as follows: comprehensive scoring data of the m travel performance dramas;
the emotional values of the m tourism performance dramas of each secondary index form a secondary index subsequence;
the subsequence of each secondary index is as follows: the emotion value sequences of 8 secondary indexes formed by the m travel performance dramas,
carrying out non-dimensionalization operation on the secondary index subsequence and the scoring parent sequence by using an averaging method, specifically calculating according to the formula (5),
wherein x' (k) is: the original value of the kth element in each secondary index subsequence or score mother sequence; x (k) is: the value of the k-th element in each secondary index subsequence or scoring mother sequence after equalization; then, according to the formulas (6) and (7), the gray correlation coefficient xi of each secondary index subsequence and the scoring mother sequence is respectively calculatedijAnd degree of grey correlation Rij，
Wherein, | xi(k)-xj(k) The absolute difference value of the ith element in the secondary index subsequence after equalization and the jth element in the scoring mother sequence after equalization is | shown; miniminjRepresents: taking the minimum absolute difference, max, among all absolute differencesimaxjRepresents: taking out the maximum absolute difference value from all the absolute difference values; rho is a resolution coefficient, and is usually 0.5;
the specific calculation steps of the grey correlation coefficient are as follows: firstly, fixing the jth element in the scoring mother sequence, and then calculating a gray correlation coefficient according to a formula (6) by using all elements in the secondary index subsequence and the jth element; by analogy, replacing elements in the scoring mother sequence, and calculating m gray correlation coefficients; calculating the average value of the m gray correlation coefficients according to a formula (7) to serve as the gray correlation degree of the corresponding secondary index and the scoring mother sequence;
and sequencing the gray correlation degrees through gray correlation analysis, removing the service secondary index sequence with the minimum correlation degree to form a secondary index sequence after dimensionality reduction, and judging the key influence factors of the service value of the tourism performance.
9. The travel performance service value evaluation method as set forth in claim 8, wherein: the topological structure of the grey neural network model consists of an LA layer, an LB layer, an LC layer and an LD layer which are sequentially connected with one another, wherein the LA layer is provided with 1 node, the LB layer is provided with 1 node, the LC layer is provided with 8 nodes, and the LD layer is provided with 1 node; the 7 input nodes of the LC layer are respectively corresponding to the emotion values of the secondary indexes of brand, coverage, originality, loyalty, satisfaction, emotion and artistry.
10. The travel performance service value evaluation method as set forth in claim 9, wherein: the parameters of the grey neural network model are set as follows: the maximum training times is 50000 times; the learning rate is 0.05; target error is e-7。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110515659.0A CN113298366B (en) | 2021-05-12 | 2021-05-12 | Travel performance service value assessment method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110515659.0A CN113298366B (en) | 2021-05-12 | 2021-05-12 | Travel performance service value assessment method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113298366A true CN113298366A (en) | 2021-08-24 |
| CN113298366B CN113298366B (en) | 2023-12-12 |
Family
ID=77321636
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110515659.0A Active CN113298366B (en) | 2021-05-12 | 2021-05-12 | Travel performance service value assessment method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113298366B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663046A (en) * | 2012-03-29 | 2012-09-12 | 中国科学院自动化研究所 | Sentiment analysis method oriented to micro-blog short text |
| WO2015105270A1 (en) * | 2014-01-10 | 2015-07-16 | 에스케이플래닛 주식회사 | User device, method for controlling same, and sensibility evaluation system comprising same |
| CN108108433A (en) * | 2017-12-19 | 2018-06-01 | 杭州电子科技大学 | A kind of rule-based and the data network integration sentiment analysis method |
| CN109034893A (en) * | 2018-07-20 | 2018-12-18 | 成都中科大旗软件有限公司 | A kind of tourist net comment sentiment analysis and QoS evaluating method |
| CN110019958A (en) * | 2017-12-28 | 2019-07-16 | Tcl集团股份有限公司 | A kind of generation method, device and the terminal device of films and television programs label |
| US20190287142A1 (en) * | 2018-02-12 | 2019-09-19 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method, apparatus for evaluating review, device and storage medium |
| CN110502744A (en) * | 2019-07-15 | 2019-11-26 | 同济大学 | A kind of text emotion recognition methods and device for history park evaluation |
| CN111753093A (en) * | 2020-07-02 | 2020-10-09 | 东北电力大学 | Method and device for evaluating level of network public opinion crisis |
| CN112287240A (en) * | 2020-09-23 | 2021-01-29 | 昆明理工大学 | Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network |
| CN112560464A (en) * | 2020-12-16 | 2021-03-26 | 中国联合网络通信集团有限公司 | Method and device for identifying implicit attribute of commodity, computer equipment and storage medium |
-
2021
- 2021-05-12 CN CN202110515659.0A patent/CN113298366B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663046A (en) * | 2012-03-29 | 2012-09-12 | 中国科学院自动化研究所 | Sentiment analysis method oriented to micro-blog short text |
| WO2015105270A1 (en) * | 2014-01-10 | 2015-07-16 | 에스케이플래닛 주식회사 | User device, method for controlling same, and sensibility evaluation system comprising same |
| CN108108433A (en) * | 2017-12-19 | 2018-06-01 | 杭州电子科技大学 | A kind of rule-based and the data network integration sentiment analysis method |
| CN110019958A (en) * | 2017-12-28 | 2019-07-16 | Tcl集团股份有限公司 | A kind of generation method, device and the terminal device of films and television programs label |
| US20190287142A1 (en) * | 2018-02-12 | 2019-09-19 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method, apparatus for evaluating review, device and storage medium |
| CN109034893A (en) * | 2018-07-20 | 2018-12-18 | 成都中科大旗软件有限公司 | A kind of tourist net comment sentiment analysis and QoS evaluating method |
| CN110502744A (en) * | 2019-07-15 | 2019-11-26 | 同济大学 | A kind of text emotion recognition methods and device for history park evaluation |
| CN111753093A (en) * | 2020-07-02 | 2020-10-09 | 东北电力大学 | Method and device for evaluating level of network public opinion crisis |
| CN112287240A (en) * | 2020-09-23 | 2021-01-29 | 昆明理工大学 | Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network |
| CN112560464A (en) * | 2020-12-16 | 2021-03-26 | 中国联合网络通信集团有限公司 | Method and device for identifying implicit attribute of commodity, computer equipment and storage medium |
Non-Patent Citations (6)
| Title |
|---|
| DWI INTAN AF\'IDAH等: "Long Short Term Memory Convolutional Neural Network for Indonesian Sentiment Analysis towards Touristic Destination Reviews", 《2020 INTERNATIONAL SEMINAR ON APPLICATION FOR TECHNOLOGY OF INFORMATION AND COMMUNICATION (ISEMANTIC)》, pages 630 - 637 * |
| HUI SONG 等: "Semantic Analysis and Implicit Target Extraction of Comments from E-Commerce Websites", pages 331 - 335 * |
| 吕家欣 等: "文旅品牌顾客契合价值测量——基于细粒度情感分析模型", vol. 34, no. 01, pages 162 - 164 * |
| 姚小云 等: "基于灰色关联分析的山水实景演出大学生感知评价研究 ———以张家界山水实景演出《天门狐仙》为例", no. 12, pages 59 - 63 * |
| 杨柳;张星;邓春林;隆征帆;: "基于舆情大数据的境外旅游影响因素的灰色关联分析模型", vol. 42, no. 02, pages 18 - 27 * |
| 钟佩等: "基于网络文本分析的广西三江民族文化旅游演艺产品游客体验感知研究", 《山西师范大学学报（自然科学版）》, vol. 34, no. 2, pages 54 - 60 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113298366B (en) | 2023-12-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110489541B (en) | Case element and BiGRU-based text summarization method for case public opinion related news | |
| CN110674410B (en) | User portrait construction and content recommendation method, device and equipment | |
| Shepherd | The Cambridge introduction to performance theory | |
| Barentsen | Emerging leadership in the Pauline mission: A social identity perspective on local leadership development in Corinth and Ephesus | |
| Crossley | Music worlds and body techniques: On the embodiment of musicking | |
| CN105895087A (en) | Voice recognition method and apparatus | |
| CN109948158A (en) | Emotional orientation analytical method based on environment member insertion and deep learning | |
| Somandepalli et al. | Computational media intelligence: Human-centered machine analysis of media | |
| CN113298367B (en) | Theme park perception value evaluation method | |
| O'Donnell | Exploring identity through Appraisal Analysis: A corpus annotation methodology. | |
| Sarin et al. | SentiSpotMusic: a music recommendation system based on sentiment analysis | |
| Neely et al. | Learning in parking lots: Affinity spaces as a framework for understanding knowledge construction in informal settings | |
| Kuang et al. | [Retracted] Popular Song Composition Based on Deep Learning and Neural Network | |
| Zhang et al. | Assessing technology adoption practices in Chinese theme parks: Text mining and sentiment analysis | |
| Bakshi et al. | Structure-to-text generation with self-training, acceptability classifiers and context-conditioning for the GEM shared task | |
| CN105095302B (en) | Public praise-oriented analysis and inspection system, device and method | |
| Wang et al. | What is the competence boundary of Algorithms? An institutional perspective on AI-based video generation | |
| CN113298366A (en) | Tourism performance service value evaluation method | |
| Wang | Analysis of multimedia teaching path of popular music based on multiple intelligence teaching mode | |
| Egger | Characterizing traveling Instagrammers: an exploration of socio-psychological concepts, travel typologies and posting motives | |
| Teklemicael et al. | Toward gamified personality acquisition in travel recommender systems | |
| Ekins et al. | The Politics of Authenticating: Revisiting New Orleans Jazz | |
| KR102599368B1 (en) | System for customized content providing based on big data and method the same | |
| Nimkulrat et al. | Reflections and connections: on the relationship between creative production and academic research | |
| Manikonda | Analysis and Decision-Making with Social Media |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |