Summary of the invention
According to the above-mentioned problems in the prior art, a kind of recommended method of prepackaged food is now provided, it is intended to be based onThe prepackaged food recommended method of marking system and label system, server-side recommend to tie according to the recommendation request of user to client feeds backFruit helps user's more relaxedly and rapidly shopping goods.
Above-mentioned technical proposal specifically includes:
A kind of recommended method of prepackaged food, is arranged a server-side, and the server-side is remotely connect with multiple respectivelyUser terminal is communicated, and the quality element standard of multiple prepackaged foods is provided in the server-side；
The original composition of each prepackaged food is obtained in advance and according to the preset quality element standard screeningMultiple quality elements are obtained, a scoring and at least one are then generated to the prepackaged food respectively according to the quality elementLabel；
The recommended method includes a process for recommending the prepackaged food to user by scoring, specifically:
Step A1, user send a recommendation request to the server-side by the user terminal, wrap in the recommendation requestInclude the prepackaged food of a pre-set categories pointed by user；
Step A2, the server-side according to the recommendation request to corresponding first recommendation results of the client feeds back,Include the prepackaged food of multiple and different manufacturers production in first recommendation results, includes in first recommendation resultsThe prepackaged food according to it is described scoring arrange from high to low；
The recommended method further includes a process for recommending the prepackaged food to user by marking, specifically:
Step B1, user send a recommendation request to the server-side by the user terminal, wrap in the recommendation requestInclude pointed by user the prepackaged food an of pre-set categories and the recommended requirements of user；
Step B2, the server-side parse to obtain the recommended requirements according to the recommendation request, and according to the recommendationDemand includes that multiple and different manufacturers produce to corresponding second recommendation results of the client feeds back, in second recommendation resultsThe prepackaged food, the prepackaged food, which has, is associated with the labels of the recommended requirements.
Preferably, to include that user is pre-set select the prepackaged food phase with user to the quality element standardAll original compositions closed.
Preferably, the scoring generating process of the prepackaged food specifically includes:
The quality element of step S1, server-side prepackaged food described in every class are screened, and are determined sameThe marking element of all prepackaged foods under classification；
Step S2, the server-side is according to a preset product standards of grading, to all described of each prepackaged foodMarking element is given a mark, and the first triggering for being associated with each marking element is previously provided in the product standards of gradingScore after condition and triggering；
Step S3, the server-side is according to each marking score of element and pre- for each marking elementIf weight, to each prepackaged food carry out it is described marking element weighted calculation, to obtain each pre- packetFill the scoring of food.
Preferably, the step A2 is specifically included:
Step A21, the server-side are screened and are exported corresponding to the institute under same category pointed by the recommendation requestThere is the prepackaged food；
Step A22, the server-side to the prepackaged food of output according to the scoring score value from high to low intoRow sequence, to form first recommendation results；
First recommendation results are fed back to user terminal by step A23, the server-side.
Preferably, the label generating process of the prepackaged food specifically includes:
Step M1, the server-side according to by the quality element of the prepackaged food respectively with a preset productEvaluation criterion is compared, and exports comparison result, includes corresponding each quality element in the product evaluation standardThe label after second trigger condition, and triggering；
Step M2, the server generate the label of the corresponding prepackaged food according to the comparison result.
Preferably, multiple keywords are also previously stored in the server-side, and are associated with the multiple of the keywordThe label；
The label generating process of the prepackaged food further include:
Step N1, the server-side obtain the related content of the prepackaged food by web crawlers from internet；
Step N2, the server-side carry out content reading to the related content and know according to a preset interpretation modelThe keyword for including in the not described related content；
Step N3, the server-side generate the label of the corresponding prepackaged food according to the keyword.
Preferably, the related content includes the news and/or microblogging and/or wechat of the prepackaged food.
Preferably, the server-side is obtained about the prepackaged food and/or about the science popularization text of the quality elementChapter；
First recommendation results and/or second recommendation results include the scientific popular article.
Preferably, the server-side saves evaluation of the user to the prepackaged food, the evaluation with it is described pre-packagedFood is associated；
First recommendation results and/or second recommendation results include the evaluation.
The beneficial effect of above-mentioned technical proposal is: the prepackaged food recommended method based on marking system and label system, serviceEnd, to client feeds back recommendation results, helps user's more relaxedly and rapidly shopping goods according to the recommendation request of user.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, completeSite preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based onEmbodiment in the present invention, those of ordinary skill in the art without creative labor it is obtained it is all itsHis embodiment, shall fall within the protection scope of the present invention.
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phaseMutually combination.
The present invention will be further explained below with reference to the attached drawings and specific examples, but not as the limitation of the invention.
Based on the above-mentioned problems in the prior art, the present invention provides a kind of recommended method of prepackaged food, settingOne server-side, server-side are communicated with multiple user terminals remotely connecting respectively, and multiple pre-packaged foods are provided in server-sideThe quality element standard of product；
The original composition of each prepackaged food is obtained in advance and is obtained according to preset quality element standard screening multipleQuality element then generates a scoring and at least one label to prepackaged food respectively according to quality element；
Recommended method includes a process for recommending prepackaged food to user by scoring, as shown in Figure 1, specifically:
Step A1, user send a recommendation request to server-side by user terminal, include pointed by user in recommendation requestA pre-set categories prepackaged food；
Step A2, server-side is according to recommendation request to corresponding first recommendation results of client feeds back, the first recommendation resultsIn include the production of multiple and different manufacturers prepackaged food, the prepackaged food for including in the first recommendation results is according to scoring by heightTo low arrangement；
Recommended method further includes a process for recommending prepackaged food to user by label, as shown in Fig. 2, specifically:
Step B1, user send a recommendation request to server-side by user terminal, include pointed by user in recommendation requestA pre-set categories prepackaged food and user recommended requirements；
Step B2, server-side parse to obtain recommended requirements according to recommendation request, and according to recommended requirements to client feeds backCorresponding second recommendation results include the prepackaged food of multiple and different manufacturers production, prepackaged food in the second recommendation resultsWith the label for being associated with recommended requirements.
In preferred embodiment of the invention, quality element standard, which includes that user is pre-set, selects pre- packet with userFill the relevant all original compositions of food.
Specifically, in specific embodiments of the present invention, the raw information of " soda " generally comprise " sodium bicarbonate "," white granulated sugar " or " sweetener ", " water " and other supplementary materials and additive, nutritional ingredient include " heat ", " protein ", " rougeFat " and " carbohydrate " etc. can filter out " energy from the raw information of " soda " according to the quality element standardAmount ", " sugar content " and " additive " these three quality elements.
In preferred embodiment of the invention, as shown in figure 3, the scoring generating process of prepackaged food specifically includes:
Step S1, server-side screen the quality element of every class prepackaged food, determine all pre- under same categoryThe marking element of packaged food；
Step S2, server-side is according to a preset product standards of grading, to all marking elements of each prepackaged foodIt gives a mark, after the first trigger condition and triggering for being associated with each marking element are previously provided in product standards of gradingScore；
Step S3, server-side are right according to the score of each marking element and for the preset weight of each marking elementEach prepackaged food carries out the weighted calculation of marking element, to obtain the scoring of each prepackaged food.
Specifically, in specific embodiments of the present invention, quality element of the server-side to all products under " milk " classificationIt is screened, determines that the marking element of " milk " category product includes " technique ", " raw material ", " additive " and " protein ".
It includes above-mentioned marking element that first trigger condition, which is in the quality element of the product, then beats all of the productSub-element is given a mark, wherein the average value of " protein content " of the milk of same type is set as to the threshold value of marking element,And threshold value is the NRV=5% of 100 grams of products, " protein content " of certain milk is the NRV=6% of 100 grams of products, i.e., highIn threshold value, then at " protein content ", this marking element scoring is positive point this milk.It is similar, by the ox of same type" additive " situation of milk is divided into " no added ", " added with the additive other than essence, fragrance, sweetener " and " contains perfumeEssence, fragrance and/or sweetener "." additive " of this milk belongs to " containing essence, fragrance and/or sweetener ", then this oxThis marking element scoring of " additive " of milk is negative point.
A weighted value is preset to each marking element, weighted value adduction is 100%, and weighted value is according to different pre-packaged foodsThe marking element for including under category is other, is preset with different weight proportion distribution methods, in specific implementation of the invention, BiscuitsPrepackaged food marking element include " energy ", " sodium content ", " additive ", " raw material " and " dietary fiber ", wherein " energyIt is the weight of 25%, " additive " is 20% that the weight of amount ", which is the weight of 35%, " sodium content ", " raw material " and " dietary fiber "Weighted value be respectively 10%；
In preferred embodiment of the invention, as shown in figure 4, step A2 is specifically included:
Step A21, server-side are screened and are exported corresponding to all pre-packaged under same category pointed by recommendation requestFood；
Step A22, server-side are ranked up the prepackaged food of output according to the score value of scoring from high to low, to be formedFirst recommendation results；
First recommendation results are fed back to user terminal by step A23, server-side.
Specifically, meet the default sort mode of the prepackaged food of recommendation request for descending arrangement, i.e., according to pre-packagedThe sequence arrangement of the scoring of food from high to low；The sortord for being supplied to user's selection further includes ascending order arrangement, i.e., according to pre-The sequence arrangement of the scoring of packaged food from low to high.
In preferred embodiment of the invention, as shown in figure 5, the label generating process of prepackaged food specifically includes:
Step M1, server-side are carried out with a preset product evaluation standard respectively according to by the quality element of prepackaged foodCompare, and export comparison result, includes the second trigger condition of corresponding each quality element, and triggering in product evaluation standardLabel afterwards；
Step M2, server generate the label of corresponding prepackaged food according to comparison result.
Specifically, the second trigger condition is that prepackaged food includes the quality element in product evaluation standard, then pre- to thisThe average value of " protein content " of " milk " of same type in above-described embodiment, is set as threshold by packaged food production markValue, and threshold value is the NRV=5% of 100 grams of products, " protein content " of this milk is the NRV=6% of 100 grams of products, i.e.,Higher than threshold value, then this " plain chocolate " is " protein content is high " in the label that " protein content " this quality element generates.However, " additive " situation of the milk of same type is divided into " no added ", " other than essence, fragrance, sweetenerAdditive " and " containing essence, fragrance and/or sweetener " three kinds of situations, " additive " of this milk be " containing essence,Fragrance and/or sweetener ", the then label that this quality element of " additive " of this milk generates are " to contain essence, fragranceAnd/or sweetener ".
In preferred embodiment of the invention, multiple keywords are also previously stored in server-side, and be associated with keyMultiple labels of word；
As shown in fig. 6, the label generating process of prepackaged food further include:
Step N1, server-side obtain the related content of prepackaged food by web crawlers from internet；
Step N2, server-side carry out content reading to related content and identify mutually inside the Pass according to a preset interpretation modelThe keyword for including in appearance；
Step N3, server-side generate the label of corresponding prepackaged food according to keyword.
Specifically, it according to preset interpretation model, is interpreted to from the related content of internet crawl prepackaged foodAnalysis, identifies the keyword in related content；Further according to the keyword in related content by such product automatic clustering, and according toThis intelligent Matching is corresponding with the classification to mark create-rule, and then generates the label of corresponding prepackaged food, label and passKeyword association saves.
In specific embodiments of the present invention, web crawlers crawls a entitled " classical Italian type flavor from internet automaticallyThe related content of the prepackaged food of espresso drink ", related content include news, microblogging, wechat and product assessment articleEtc. contents；Server-side carries out keyword identification to the related content crawled, specifically includes and judges name of product for " * coffee drinkMaterial ", and has " * water " in list of ingredients, but meet except " solid beverage " condition etc..To be determined as " coffee drinkMaterial ".According still further to the other label create-rule of " coffee beverage " this foodstuff, as whether there is or not " Arabica " or " spreaded out on labelThe kinds name such as this tower " or " Libiee's benefit card "；Whether there is or not " Jamaica " or " Vietnam " or " Indonesia " or " Ethiopia " or " Yunnan "Equal places of production name etc., it is final to determine to generate respective markers, such as " Italian type flavor ", " the brand place of production ", " well-known kind coffee bean "Keyword, label are saved with crucial word association.
In the information that web crawlers can not be crawled automatically, system manager imports the related content of prepackaged food extremelyServer-side generates corresponding prepackaged food further according to keyword according to interpreting model progress interpretive analysis and identifying keywordLabel.
In preferred embodiment of the invention, related content includes the news and/or microblogging and/or micro- of prepackaged foodLetter.
Specifically, the preset daily auto-browsing related web site of intelligent program in user terminal, using artificial intelligence identification andAutomatically the contents such as news, microblogging, the wechat of every class prepackaged food are searched for, keyword, such as intelligent program auto-browsing are generatedTo the wechat push article of entitled " 11 sections of mineral water evaluating results, this 5 sections of mineral waters are worth coming one bottle again ", automatic identification mineThe various microelement type and contents of spring, and generate the label that keyword generates mineral water.
In preferred embodiment of the invention, server-side is obtained about prepackaged food and/or about the section of quality elementGeneral article；
First recommendation results and/or the second recommendation results include scientific popular article.
Specifically, the related scientific popular article of all kinds of prepackaged foods is stored in server-side in advance, intelligent program is from correlationAutomatically the scientific popular article about the quality element in prepackaged food and/or prepackaged food is obtained on website and is stored to serviceIt include scientific popular article in the first recommendation results and/or the second recommendation results, scientific popular article, which is set to, meets recommendation request in endBelow prepackaged food, it is supplied to user's reading.
Scientific popular article is divided into news comment, product testing and existence general knowledge according to the type of content.Specific reality of the inventionApply in example, intelligent program with " soda " be keyword, obtain as " 12 ounces of soda probably have 10 tea spoons sugar youDetermine that you do not mind ", the articles such as " weight-reducing cannot pretend to make great efforts also to take frequently the soda of a big bottle height heat ", and arrange markAfter topic and link, show in the lower section of recommendation results.
In preferred embodiment of the invention, server-side saves evaluation of the user to prepackaged food, evaluation with it is pre-packagedFood is associated；
First recommendation results and/or the second recommendation results include evaluation.
Specifically, in above-described embodiment, user, can be to the pre- packet after having purchased prepackaged food by recommender systemDress food is evaluated, and evaluation information can be associated to the label of prepackaged food, browse the pre-packaged food in next bit userWhen the label of product, system is automatically pushed to user.
During recommending prepackaged food to user by scoring, it is previously provided in server-side related to unfavorable ratingsKeywords database, multiple keywords in keywords database are associated with a marking element of prepackaged food；Evaluation to userKeyword in information is identified, if identifying the relevant keyword of unfavorable ratings, then determines that this evaluation information is negativeFace evaluation.
In specific embodiments of the present invention, the marking element of prepackaged food " milk tea " includes " mouth of user experience levelFeel greasy ".If identifying the relevant keyword such as " sweet " unfavorable ratings from evaluation information, then this evaluation information is determinedFor unfavorable ratings.
A default first threshold is negatively commented if it is more than first threshold that unfavorable ratings, which account for the ratio all evaluated, describedPreset first score value is lowered in the scoring of the associated marking element of the keyword that valence includes；A default second threshold, if negativelyIt is more than second threshold, the then associated marking element of keyword for including to the unfavorable ratings that evaluation, which accounts for the ratio all evaluated,Preset second score value is lowered in scoring；A default third threshold value, if it is more than third threshold that unfavorable ratings, which account for the ratio all evaluated,The scoring of value, the then associated marking element of keyword for including to the unfavorable ratings is adjusted downward to a preset third score value.UnderScoring after tune is saved into server-side.
In above-described embodiment, the marking element of prepackaged food " milk tea " includes " mouthfeel is greasy " of user experience level.If it is more than 60% that the unfavorable ratings of " mouthfeel is greasy ", which account for the ratio all evaluated, the scoring of " mouthfeel is greasy " lowers 15 points；IfIt is more than 80% that the unfavorable ratings of " mouthfeel is greasy ", which account for the ratio all evaluated, then the scoring of " mouthfeel is greasy " lowers 30 points；If " mouthFeel greasy " unfavorable ratings to account for the ratio all evaluated be more than 90%, then be adjusted to 0 point under the scoring of " mouthfeel is greasy ".
During recommending prepackaged food to user by label, it is previously provided in server-side related to unfavorable ratingsKeywords database, multiple keywords in keywords database are associated with a label of prepackaged food, close to evaluation informationKeyword identification, if identifying the relevant keyword of unfavorable ratings, then determines this comment information for unfavorable ratings.Systemic presuppositionOne threshold value, if it is more than the threshold value that unfavorable ratings, which account for the ratio all evaluated, system is automatically deleted mark associated by the keywordNote.
The foregoing is merely preferred embodiments of the present invention, are not intended to limit embodiments of the present invention and protection modelIt encloses, to those skilled in the art, should can appreciate that all with made by description of the invention and diagramatic contentEquivalent replacement and obviously change obtained scheme, should all be included within the scope of the present invention.