Method for automatically judging cigarette product consumer satisfaction
Through natural language processing technology and custom dictionary, the consumer satisfaction score of cigarette products is automatically calculated, which solves the problems of low efficiency and high cost of traditional evaluation methods, and realizes efficient and low-cost automatic evaluation and continuous monitoring of consumer satisfaction .
Patent Information
- Application Number
- CN202411035604.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-11-08
AI Technical Summary
The traditional cigarette product consumer satisfaction evaluation method relies on manual collation and statistics, which is inefficient and cost-effective, making it difficult to collect dynamic data and track continuous tracking, and the analysis results lack scientific basis.
Natural language processing technology is adopted, and a custom emotional dictionary for cigarette consumer evaluation, degree adverb dictionary and negative word dictionary are used to automatically capture Internet comment data, calculate consumer satisfaction scores, and realize automatic judgment.
An automated consumer satisfaction evaluation process is realized, with short cycles, high efficiency and low cost, and can continuously track and monitor and provide scientifically based satisfaction evaluation results.
Smart Images

Figure CN118917544A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of tobacco, and in particular relates to a method for automatically judging consumer satisfaction with cigarette products based on Internet consumer evaluation data. Background Art
[0002] The goal of judging consumer satisfaction with cigarette products is to analyze and judge the consumer satisfaction and market popularity of own products by collecting and tracking consumer evaluation information of own products, identifying the strengths and weaknesses of own products, summarizing and proposing effective improvement suggestions, and guiding the development of product quality improvement and maintenance work. The traditional method of judging consumer satisfaction is to use focus interviews or intercept interviews to collect consumer satisfaction evaluation information on cigarette products using survey questionnaires. The organization process of survey activities is time-consuming, costly, and inefficient, and it is difficult to achieve dynamic data collection and continuous tracking and monitoring. Survey data mainly relies on manual collation and statistical summary, especially the descriptive language of consumers, which is unstructured data. Survey analysts need to read a large number of survey records and rely on subjective impressions in their brains to make general judgments. The workload is large and the analysis results lack scientific basis.
[0003] In recent years, with the vigorous development of e-commerce, social networking sites, blogs, microblogs and self-media, consumers have a wider space to express their opinions. People are more enthusiastic about participating in discussions and publishing their own opinions, attitudes and emotional information. Therefore, Internet information with subjective speech and comments is becoming more and more abundant. These massive, emotional, unstructured texts contain a lot of information. Enterprises can obtain users' opinions on products or services from them, and consumers can get more suggestions before consumption. Therefore, the analysis of consumers' opinions, views, evaluations and other tendency information has important practical value. Text analysis technology is used to mine consumers' subconscious consumption information (taste, taste, packaging, etc.), understand consumers' recognition and evaluation suggestions of cigarette products, find out the advantages and disadvantages of their own products, and explore improvement methods.
[0004] The present invention is proposed for this purpose. Summary of the invention
[0005] In order to solve the above technical problems, the present invention provides a method for automatically judging the consumer satisfaction of cigarette products based on Internet consumer evaluation data. The method of the present invention uses natural language processing technology, based on a self-defined special sentiment dictionary, degree adverb dictionary and negative word dictionary for cigarette consumer evaluation, to calculate the consumer satisfaction score for subjective text with emotional color on the Internet, thereby simulating the process of survey analysts summarizing, summarizing and analyzing by reading materials and relying on subjective impressions. The method of the present invention automatically captures consumer comment data on the Internet to construct a sample set of consumer satisfaction evaluation data, calls the special sentiment dictionary, degree adverb dictionary and negative word dictionary for cigarette consumer evaluation to calculate the consumer satisfaction score, and realizes the automatic judgment of consumer satisfaction of cigarette products. The whole process is completed automatically, with a short cycle, high efficiency and low cost, which can meet the needs of enterprises for large-scale product satisfaction evaluation and can continuously track and monitor.
[0006] The technical solution of the present invention is as follows:
[0007] A method for automatically judging consumer satisfaction of cigarette products comprises the following steps:
[0008] (1) Use web crawler tools to automatically capture consumer review data on the Internet and construct a sample set of consumer satisfaction evaluation data X = {x i}, i = 1 2 3 ... n;
[0009] (2) Construct a special sentiment dictionary for cigarette consumer evaluation and assign scores; construct a degree adverb dictionary and assign weights; construct a negative word dictionary and assign weights;
[0010] (3) Calculate consumer review data x i The corresponding consumer satisfaction score of cigarette products i ;
[0011] (4) Calculate the consumer satisfaction score s i The arithmetic mean of s i The arithmetic mean of the consumer satisfaction with cigarette products was used to evaluate the consumer satisfaction with cigarette products.
[0012] Preferably, step (1) utilizes web crawler technology to automatically capture consumer review data, including the following steps:
[0013] (11) Set the URL set to be crawled and create a URL queue;
[0014] (12) Read the URL addresses from the URL queue in sequence, crawl the element data of the corresponding web page, and cache them in the page database;
[0015] (13) Filter consumer evaluation data and construct a sample set X of evaluation data.
[0016] Preferably, step (2) constructs a special sentiment dictionary for cigarette consumer evaluation and assigns scores; constructs a degree adverb dictionary and assigns weights; constructs a negative word dictionary and assigns weights; and comprises the following steps:
[0017] (21) Constructing a special sentiment dictionary for cigarette consumer evaluation and assigning scores: Collecting the vocabulary of smokers' communication terms, cigarette evaluation professional terms, and Internet cigarette consumer evaluation corpus, and selecting common sentiment words from eight dimensions, namely packaging, taste, mouthfeel, smoke, aftertaste, aroma, price, and quality, according to the two sentiment polarities of positive and negative, to construct a special sentiment dictionary for cigarette consumer evaluation and assign scores;
[0018] (22) Construct a degree adverb dictionary and assign weights: collect degree adverbs and assign weights according to extreme, high, medium-high, medium, and low;
[0019] (23) Construct a negative word dictionary and assign weights: Collect negative words to construct a negative word dictionary and assign weights.
[0020] Preferably, step (3) calculates the consumer review data x i The corresponding consumer satisfaction score of cigarette products i , including the following steps:
[0021] (31) Consumer review data x i Perform word segmentation;
[0022] (32) Traverse x i For each word in , if the word belongs to the special sentiment dictionary for cigarette consumer evaluation, determine whether there is a degree adverb before the word; if so, multiply the degree adverb weight by the word score; if the degree adverb also contains a negative word, multiply the negative word weight and the degree adverb weight by the word score; get the score s ij ;
[0023] (33) Traverse to the entire x i Until there are no more sentiment words, add up the scores of each traversal process to get x i Consumer satisfaction scores i .
[0024] Preferably, in step (4), the consumer satisfaction score s i The arithmetic mean calculation formula is:
[0025] Preferably, the judging criteria for consumer satisfaction with cigarette products are: when s>0, it indicates that consumers are satisfied with the cigarette products; when s<0, it indicates that consumers are dissatisfied with the cigarette products; the absolute value of s indicates the intensity of consumers' satisfaction or dissatisfaction with the cigarette products.
[0026] Beneficial effects of the present invention:
[0027] 1. Compared with the traditional method of conducting consumer satisfaction evaluation of cigarette products through market research, the automatic evaluation method of consumer satisfaction of cigarette products of the present invention automatically completes the entire process from consumer comment data collection, analysis and calculation to satisfaction evaluation, with a short cycle, high efficiency and low cost, and can meet the needs of enterprises for large-scale product satisfaction evaluation and can continuously track and monitor.
[0028] 2. The traditional satisfaction evaluation method requires research analysts to read a large number of research records and make a general judgment based on subjective impressions in their brains. The workload is large and the analysis results lack scientific basis. The present invention is based on the constructed special sentiment dictionary, degree adverb dictionary and negative word dictionary for cigarette consumer evaluation, and automatically calculates the consumer satisfaction score of cigarette products through word segmentation and traversal, thereby realizing the automatic evaluation of consumer satisfaction of cigarette products.
[0029] 3. The method for automatically judging the consumer satisfaction of cigarette products of the present invention can continuously maintain and update the special sentiment dictionary, degree adverb dictionary and negative word dictionary for cigarette consumer evaluation to cope with the changes in consumer evaluation terms and emerging vocabulary; and can filter consumer evaluation data by setting conditions such as URL range and consumer information to achieve targeted analysis and judgment of cigarette product consumer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 This is a schematic diagram of the main steps of the method for automatically judging consumer satisfaction with cigarette products of the present invention. DETAILED DESCRIPTION
[0031] In order to make the purpose and technical scheme of the present invention clearer, detailed description is given in conjunction with the following examples, which are intended to illustrate the content of the present invention rather than to further limit the scope of protection of the present invention. The process, conditions, reagents, experimental methods, etc. implemented, except for the content specifically mentioned below, are common knowledge and common common sense in the art, and the present invention has no special restrictions. The experimental methods for which specific conditions are not indicated in each embodiment are usually in accordance with conventional conditions or according to the conditions recommended by the manufacturer. Unless otherwise indicated, the meanings of all professional terms and scientific terms used in this specification are the same as those generally understood by technicians in the technical field to which the present invention belongs. However, in case of conflict, the present specification including the definition shall prevail.
[0032] like Figure 1 As shown, a method for automatically judging consumer satisfaction of cigarette products of the present invention comprises the following steps:
[0033] (1) Use web crawler tools to automatically capture consumer review data on the Internet and construct a sample set of consumer satisfaction evaluation data X = {x i}, i = 1 2 3 ... n;
[0034] (2) Construct a special sentiment dictionary for cigarette consumer evaluation and assign scores; construct a degree adverb dictionary and assign weights; construct a negative word dictionary and assign weights;
[0035] (3) Calculate consumer review data x i The corresponding consumer satisfaction score of cigarette products i ;
[0036] (4) Calculate the consumer satisfaction score s i The arithmetic mean of s i The arithmetic mean of the consumer satisfaction with cigarette products was used to evaluate the consumer satisfaction with cigarette products.
[0037] Among them, step (1) uses web crawler technology to automatically crawl consumer review data, including the following steps:
[0038] (11) Set the URL set to be crawled and create a URL queue;
[0039] (12) Read the URL addresses from the URL queue in sequence, crawl the element data of the corresponding web page, and cache them in the page database;
[0040] (13) Filter consumer evaluation data and construct a sample set X of evaluation data.
[0041] Among them, step (2) constructs a special sentiment dictionary for cigarette consumer evaluation and assigns scores; constructs a degree adverb dictionary and assigns weights; constructs a negative word dictionary and assigns weights; and includes the following steps:
[0042] (21) Constructing a special sentiment dictionary for cigarette consumer evaluation and assigning scores: Collecting the vocabulary of smokers' communication terms, cigarette evaluation professional terms, and Internet cigarette consumer evaluation corpus, and selecting common sentiment words from eight dimensions, namely packaging, taste, mouthfeel, smoke, aftertaste, aroma, price, and quality, according to the two sentiment polarities of positive and negative, to construct a special sentiment dictionary for cigarette consumer evaluation and assign scores; as shown in Table 1:
[0043] Table 1 Sentiment dictionary for cigarette consumer evaluation
[0044]
[0045] (22) Construct a degree adverb dictionary and assign weights: Collect degree adverbs and assign weights according to extreme, high, medium-high, medium, and low; as shown in Table 2:
[0046] Table 2 Degree adverbs and weight settings
[0047]
[0048] (23) Construct a negation word dictionary and assign weights: For Chinese sentences containing multiple negations, when the number of times a negation word appears in a phrase is an odd number, it indicates a negative meaning; when the number of times a negation word appears in a phrase is an even number, it indicates a positive meaning; collect the negation words to form a negation word dictionary, and set the negation word weight to -1, as shown in Table 3 below:
[0049] Table 3 Negative words and weight settings
[0050]
[0051] Among them, step (3) calculates the consumer review data x i The corresponding consumer satisfaction score of cigarette products i , including the following steps:
[0052] (31) Consumer review data x i Perform word segmentation;
[0053] (32) Traverse x i For each word in , if the word belongs to the special sentiment dictionary for cigarette consumer evaluation, determine whether there is a degree adverb before the word; if so, multiply the degree adverb weight by the word score; if the degree adverb also contains a negative word, multiply the negative word weight and the degree adverb weight by the word score; get the score s ij ;
[0054] (33) Traverse to the entire x i Until there are no more sentiment words, add up the scores of each traversal process to get x i Consumer satisfaction scores i .
[0055] Among them, step (4) consumer satisfaction score s i The calculation formula for the arithmetic mean is:
[0056] Among them, the judging criteria for consumer satisfaction with cigarette products are: when s>0, it indicates that consumers are satisfied with the cigarette products; when s<0, it indicates that consumers are dissatisfied with the cigarette products; the absolute value of s indicates the intensity of consumers' satisfaction or dissatisfaction with the cigarette products.
[0057] Embodiment: This embodiment is a method for automatically judging the consumer satisfaction of the cigarette product "Yuxi (soft)", the main steps are as follows Figure 1 As shown; the specific steps are as follows:
[0058] Step (1) collects the netizen evaluation data about the cigarette product "Yuxi (soft)" from Yanyue.com, sets the URL set to be crawled, and establishes a URL queue (see Table 4); reads the URL addresses from the URL queue in sequence, crawls the element data of the corresponding web page and caches it in the page database (see Table 5); screens the consumer evaluation data and constructs an automatic judgment data sample set X (see Table 6).
[0059] Table 4 URL Queue
[0060] Serial number URL 1 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 1 2 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 2 3 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 3 4 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 4 5 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 5 6 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 6 7 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 7 8 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 8 …… …… 163 https: / / www.yanyue.cn / product / comment / productid / 148 / p / 163
[0061] Table 5 Page database cache
[0062]
[0063] Table 6 Data sample set X
[0064] Serial number variable Comment data 1 <![CDATA[x1]]> The taste is very smooth, the aroma is pure, and the packaging is not rough. 2 <![CDATA[x 2]]> This packaging always feels fake, the paper feels so bad 3 <![CDATA[x 3]]> I like this one, it's quite good and worth recommending. …… …… …… 1619 <![CDATA[x 1619 ]]> It's not as good as soft clouds. Maybe it's because I'm a novice and don't know how to evaluate it. I just think soft clouds are the best.
[0065] Step (2), according to Tables 1 to 3, construct a special sentiment dictionary for cigarette consumer evaluation and assign scores; construct a degree adverb dictionary and assign weights; construct a negative word dictionary and assign weights; calculate the corresponding cigarette product consumer satisfaction score starting from the first consumer comment data:
[0066] i. Segment the consumer review data x1 into words: {'taste', 'very', 'smooth', 'aroma', 'pure', 'packaging', 'no', 'rough'}.
[0067] ii. Traverse each word in x1: 'soft' belongs to the special sentiment dictionary for cigarette consumer evaluation (score +1), and there is a degree adverb 'very' (weight 2) before the word, and there is no negative word, then s 11 =1*2=2; 'pure' belongs to the special sentiment dictionary for cigarette consumer evaluation (score +1), there is no degree adverb or negative word before the word, then s 12 =1; 'rough' belongs to the special sentiment dictionary for cigarette consumer evaluation (score -1), and the word is preceded by the negative word 'not', then s 13 =-1*-1=1; the consumer satisfaction corresponding to x1 is s1=s 11 +s 12 +s 13 =4.
[0068] iii. Calculate the consumer satisfaction scores corresponding to all evaluation data using the same method.
[0069] Step (3), calculate the arithmetic mean of the consumer satisfaction scores for all 1619 review data, which is: There are 1,215 consumer reviews with satisfaction scores greater than 0, accounting for about 75.05% of all consumer reviews, indicating that the market consumer satisfaction with the cigarette product "Yuxi (soft)" is relatively high.
[0070] The whole process from data collection to analysis and calculation takes 2 minutes and 30 seconds. It is highly efficient and low cost. It can be completed automatically.
[0071] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.
Claims
1. A method for automatically judging consumer satisfaction of cigarette products, characterized in that: The steps include: (1) Use web crawler tools to automatically crawl consumer review data on the Internet and construct a sample set of consumer satisfaction evaluation data X = {x i }, i = 1 2 3 ... n; (2) Construct a special sentiment dictionary for cigarette consumer evaluation and assign scores; construct a degree adverb dictionary and assign weights; construct a negative word dictionary and assign weights; (3) Calculate consumer review data x i The corresponding consumer satisfaction score of cigarette products i ; (4) Calculate the consumer satisfaction score s i The arithmetic mean of s i The arithmetic mean of the consumer satisfaction with cigarette products was used to evaluate the consumer satisfaction with cigarette products.
2. The method according to claim 1, characterized in that Step (1) automatically crawls consumer review data using web crawler technology, including the following steps: (11) Set the URL set to be crawled and create a URL queue; (12) Read the URL addresses from the URL queue in sequence, crawl the element data of the corresponding web page, and cache them in the page database; (13) Filter consumer evaluation data and construct a sample set X of evaluation data.
3. The method according to claim 1, characterized in that Step (2) constructs a special sentiment dictionary for cigarette consumer evaluation and assigns scores; constructs a degree adverb dictionary and assigns weights; constructs a negative word dictionary and assigns weights; and includes the following steps: (21) Constructing a special sentiment dictionary for cigarette consumer evaluation and assigning scores: Collecting the vocabulary of smokers' communication terms, cigarette evaluation professional terms, and Internet cigarette consumer evaluation corpus, and selecting common sentiment words from eight dimensions, namely packaging, taste, mouthfeel, smoke, aftertaste, aroma, price, and quality, according to the two sentiment polarities of positive and negative, to construct a special sentiment dictionary for cigarette consumer evaluation and assign scores; (22) Construct a degree adverb dictionary and assign weights: collect degree adverbs and assign weights according to extreme, high, medium-high, medium, and low; (23) Construct a negative word dictionary and assign weights: Collect negative words to construct a negative word dictionary and assign weights.
4. The method according to claim 1, characterized in that: Step (3) Calculate the consumer review data x i The corresponding consumer satisfaction score of cigarette products i , including the following steps: (31) Consumer review data x i Perform word segmentation; (32) Traverse x i For each word in , if the word belongs to the special sentiment dictionary for cigarette consumer evaluation, determine whether there is a degree adverb before the word; if so, multiply the degree adverb weight by the word score; if the degree adverb also contains a negative word, multiply the negative word weight and the degree adverb weight by the word score; get the score s ij ; (33) Traverse to the entire x i Until there are no more sentiment words, add up the scores of each traversal process to get x i Consumer satisfaction scores i .
5. The method according to claim 1, characterized in that Step (4) Consumer satisfaction score s i The arithmetic mean calculation formula is:
6. The method according to claim 5, characterized in that The judging criteria for consumer satisfaction with cigarette products are: when s>0, it indicates that consumers are satisfied with the cigarette products; when s<0, it indicates that consumers are dissatisfied with the cigarette products; the absolute value of s indicates the intensity of consumers' satisfaction or dissatisfaction with the cigarette products.
