KISAN MITRA : AN INTELLIGENT CHATBOT

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN:2395-0072

KISAN MITRA

AN INTELLIGENT CHATBOT

Agriculture occupies an critical role within side the Indian economy. Indian farmers nowadays are dealing with the trouble of low profits because of the dearthofrecords approximately authorities schemes, fertilizers, farminggadget etc. Agriculture hired 50% of the Indian paintings pressure and contributed 17 18% to country’s GDP. In 2016. Agriculture and allied sectors like animal husbandry, forestry and fisheries accounted for 15.4% of the GDP (gross home product) with approximately 31% of the body of workers in 2014 This paintings pursuits to implement, a Chatbot that promotes far flung interplay of the users/farmers to the agriculture surroundings the usage of Natural language processing. We need to construct a talk bot which could solution the queries of the farmers and also can offer a probable records and answer associated with agriculture. There also are such a lot of authorities schemes which may be now no longer regarded to the farmers that are very useful for the farmers, the chat bot also can offer records as mentioned. Some smallholders and marginalized farmers have low recognition as maximum of them stay in far flung regions and do not have get right of entry to records approximately soil properties, seeds, currently used tools, fertilizers, etc. The report proposes an intelligent, transportable gadget that makes use of Natural language processing strategies to assist farmers use exclusive farming strategies, andinadditionassist them to reply their queries and remedy their doubts the usage of chatbot a good way to shop their time. To meet all of the necessities of farmers, a chatbot is proposed the usage of Natural language processing technology. The gadget will act as an interactive digital assistant for farmers, answering all queries associated with agriculture. This paper will undergo the implementation of the chatbot the usage of the chatterbot libraries and Android Studio

effective be beneficial to farmers to offer relational developments.Thereislikewiseafewsoftwareprogramto train farmers approximately technology. But maximum of those technology aren't viable as they do now no longer supplycorrectsolutionstothequeriesrequestedthroughthe farmers[2].Therearefarmerswhostayinfaroffregionsof the country, who do now no longer have any statistics associated with new technological developments and agricultural practices, who do now no longer have get admission toto enough statisticson crops,soil properties, seeds,moderndayequipmentused,fertilizers,etc.Lackofget admission to agriculture understanding and statistics approximatelythemoderndayfarmingpracticesendsinbad crop and farm animals productivity. To conquer the drawbacks, a chatbot is proposed the usage of the sample matching approach that offers an interface, in which the users, inthis example farmers,can speak with the chatbot efficiently.ThismissionKisanMitraisanomnibuscellular appadvancedtoassistfarmersthroughpresentingapplicable statisticstothemquickly.Withclickonofabutton,theyare able to get the statistics on climate of the day, Nutrient Management, Government Schemes, Market Information, Providing statistics approximately the seeds, planting materials,areastatisticsandculturalpracticesetc.Andthat too of their Regional Languages to empower farmers withinsidethenicefeasiblemanner.UsingMachinegettingto knowalgorithmsquestionresponsesaregenerated[1,2,5].If the proposed machine does now no longer reply to the farmers’ queries, those queries are forwarded to experts. Apartfromfarmers,it'lladditionallyassistthefolkswhoare novicesandinterestedbyfarming[1].Indestinysentiment evaluationandlanguagetranslationmaybeimplementedto thisapplication[3,7].TheChatbothasafewprecisefunctions like including the query and retraining, updating the prevailingquery,mailingtheunansweredqueriestoenhance the ones question to better degree through expert. These operationmakesthechatbotrespondefficiently.

2. LITERATURE SURVEY

1.INTRODUCTION

Farmingperformsaprimefunctionwithinsidetheforthe improvement of the country. It gives employment possibilitiestoabigvarietyofhumansacrosstheworld.Day through Day the generation is growing withinside the agriculture area [1]. Nowadays the authorities is accumulating information concerning rainfall and crop manufacturing however it's miles of no need for farmers. Analyzing and accumulating the information can most

In the paper titled " E AGRO: Intelligent Chat Bot. IoT and Artificial Intelligence to Enhance Farming Industry," the writer stated Agriculture is the principle enterprise withinsidetheruralregionsofSriLankaandtheagriculture contributesaconsiderablequantitytothefinancialsystemof SriLanka.However,theagricultureisturningintomuchless worthwhileenterpriseandsubsequently,there'sanwilling fashion that the agricultural network continues far from agriculture.Therearemanymotivesforthisdevelopment.

Key Words: Chatterbot, NLTK, android studio, Farming Assistant,NLP,Voiceassistant

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN:2395-0072

Among them having access to records at proper time approximatelyagricultureisamainissue.Hence,thegoalsof thisventurearetocommunityfarmersandofferanswersto farmers’troublesatpropertime.Tothatend,weadvanceda talkroomtoproportionthefarmers’storiesandknowledge. Further, we advanced an wise Chat Bot, which presents answers for positive troubles of farmers via an internet gadget.[1]

Inthepapertitled“Agribot:ANaturalLanguageGenerative Neural Networks Engine for Agricultural Applications”. OpenWeatherApianddifferentgadgetmasteringalgorithms .WeatherpredictiontheuseofAgribotgivenregionasenter Mostlyfarmerscommunicateoftheirnearbylanguages.We wantdevicewhichcouldinter relateourgadgetandtheir language.Theclimatepredictionmodulemaybechangedin orderthatitaccessestheregionwithoutuserinterruption.In ordertoenhancetheoverallperformanceinfuture,authentic length photographs taken in actual international environment[2]

Inthepapertitled“LanguageIdentificationfromanIndian MultilingualDocumentUsingProfileFeatures.”thewriteron thispaper,ansetofrulesforlanguageidentityofKannada, HindiandEnglishtextualcontentstrainsfrompublishedfiles isproposed.Thetechniqueisprimarilybasedtotallyatthe evaluationofthepinnacleandbacksideprofilesofcharacter textualcontentstrainsandsubsequentlydoesnownolonger requireanymanorwomanorphrasesegmentation.Further studies to enhance the set of rules thinking about extraordinary font kind and length and additionally to paintingsonhandwrittenfiles.[3]

In the paper titled " Language Identification for MultilingualMachineTranslation,"thewriteronthispaper provides a short assessment of the want and demanding situations concerned in automated language identity for gadget translation task. Language identity and machine translationmaybeverycriticaltomakemovelingualrecords tobehadtomass.AsMarathi,HindiandSanskritarevery intently associated languages, getting the distinguishing capabilitiesthatclassifythemisatoughtask.Asthesetypes oflanguagesproportiontheidenticalscript,typeofnamed entitiesislikewisetough.[4]

Inthepapertitled“AgriBot Anwiseinteractiveinterface to help farmers in agricultural activities.” the writer completedwholepaintingschangedintobasedtotallyongoal of supplying farmers with a accessible and transportable digital interactive farming assistant that may talk with farmers.Theappchangedintodesignedtobeextraordinarily farmer specific. Agribot, now no longer simplest allows farmers get high quality crop pointers however it additionally allows them betternurture their plants, shop theirplantsasawaytogrowthshelfexistenceofplants,in additiontoassistthemdiscoverhigh qualityrates.[5].

3. PROBLEM STATEMENT

Nowadays,thesmartphonesarebitnotunusualplaceor eventhefarmersarecapableofuseitbasically.Theylocatea few obligations a piece hard to go looking or accumulate recordsapproximatelytheirsimplenecessitieswiththeaidof usingtypingandlookingorwiththeaidofusingthelanguage barriertothey'regoingthroughafewproblemstoserveat thenettoaccumulaterecordssowehaddeterminedtocreate ansoftwareforthemtolessenalltheirobstaclesandtoserve themwiththedesiredrecordswiththeaidofusingimparting thefunctionsinoursoftwaretoeasetheirafewtypeofInour venturewe'reseekingtolessentheobstaclesofthefarmers togetuptodatewiththetoday'sandmedicalstuffstogrowth theirproductivity

3.1 DATA SOURCE

For Working ofanychatbotthe information is vital for schoolingthechatbot.Therefore,theprimarystepofgrowing the chatbot is information.The information base for the chatbot is received from the information this is collected. Grouped information from the maximum requested query andsolutionsapproximatelyfarmingoragriculturefromthe net. This accrued information is given as a dataset for schoolingthemodel.Thedatasetconsistsofattributes:query (query)andresponse(answer)inCSV.

Table 1: ExampleofLiveDataSet

3.2 DATA PREPROCESSING

Pre processing refers back to the differences implementedtoourinformationearlierthanfeedingittothe algorithm. Data Preprocessing is a method this is used to transform the uncooked information right into a smooth information set. In our device the voice enter is first off transformedintotextualcontentformat.Whenrunningwith textualcontentinformation,wewanttocarryoutnumerous preprocessing at the information earlier than we make a chatbot examine. Based at the necessities we want to use numerousoperationstopreprocesstheinformation.

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN:2395-0072

Tokenizing is the maximum simple and primary factor youmaydoontextualcontentinformation.Tokenizingisthe systemofbreakingthecompletetextualcontentintosmall elements like phrases. Here we iterate thru the styles and tokenize the sentence the usage of nltk. word_tokenize() characteristic and append every phrase withinside the phraseslisting.Weadditionallycreatealistingoflessonsfor ourtags.Nowweareabletolemmatizeeveryphraseandget ridofreplicaphrasesfromthelisting.

Lemmatizingisthesystemofchangingaphraseintoits lemmashapeafterwhichdevelopinga pickledocumentto keep the Python gadgets which we are able to use whilst predicting than we had created the schooling information wherein we are able to offer the enter and the output information for the chatbot to examine and deliver the outputaccordingly.

4. PROPOSED WORK

4.1 DATASET PREPARATION

As our chatbot is multilingual weought to educate our device different language datasets. So we had created the datasets in 3 languages i.e. English, Hindi and Marathi languages. We had created many queries primarily based totallyatthefarmingprimarilybasedtotallytoeducateour deviceinorderthatitcouldrecognizehowtheenterterms canbe.ToeducateourversionwehadusedNLTKbecauseit hasmanyfunctionalitiestotalk withthedevicevia wayof meansoftheusageofNaturallanguage.

a. NLTK

NLTKisatoolkitassembleforrunningwithNLPinPython.It providesusseveraltextprocessinglibrarieswithmassesof testdatasets.ItgoesUnderProcess:Tokenization,Lowercase conversion,StopWordsremoval,Stemming,Lemmatization, Parse tree or, Syntax Tree generation, POS Tagging. Tokenization: The idea is to break up each word and assembleavocabularysuchthatwecanrepresentallterms uniquely in a list. Lower case conversion: We convert all termsintothelowercasetoavoidredundancywithinsidethe tokenlist.StopWordsremoval:Removednoiseinadvance thanprocessingforcleanserprocessingwithinsidethemodel. Stemming: Extract the premise word and put off the rest. Herethepremisewordlong establishediscalled‘stem’andit isnotconstantlythatstemwantstoexistandcharacteristica meaning.Lemmatization:Wewanttoextractthelowestform of the word here. Which is available withinside the dictionary. We have the WordNet corpus and the lemma generated may be available in this corpus. Parse tree or, SyntaxTreegeneration:Wecandefinegrammarandthenuse NLTKRegexpParsertoextractallfactorsofspeechfromthe sentence.weoffereachwordaparticulartagandapproach them.

b. Architecture

Fig 1:ArchitectureofKisanMitra

Theusercanprovidetheinputinformoftextorspeech, if the user give input in the form of speech, the input is converted in to text and the process of natural language processingtakesplace.WehadusedNLTKmoduleforinput processing and generating output based on the language selectedbytheuser.Thenaturallanguageprocessstartafter theconversionprocess,theprocesssuchaslexicalanalysis, LexicalAnalysisisthefirstphaseofthecompileralsoknown asascanner.ItconvertstheHighlevelinputprogramintoa sequence of Tokens. Lexical Analysis can be implemented with the Deterministic finiteAutomata. The output is a sequence of tokens that is sent to the parser for syntax analysis.Syntacticanalysisisdefinedasanalysisthattellsus thelogicalmeaningofcertainlygivensentencesorpartsof thosesentences.Wealsoneedtoconsiderrulesofgrammarin ordertodefinethelogicalmeaningaswellasthecorrectness ofthesentencesIfthesystemInotabletogivetheansweror itdoesnotunderstandthequeryitprocessitandtrytolearn howtogiveanappropriateanswerforthatquerynexttime. For reasoning we are using nltk as it has functionality to understand the phrases and it provide the best suitable answer. And for further functions we had provided in our application we had used api’s to pull the live data from governmentofficialwebsites.Forapplicationdevelopment weareusingandroidstudioasanplatform.Itworksusing javaandflutter.OurGUIandotherfunctionalityweremade usingjavabutourcorefunctionalityofchatbotwasnotbeen able to develop using python and android studio does not supportpythondirectly.Sotousepythoninandroidstudio we have to use chaquopy which is an library used to add pythoninterpreterinandroidstudio.Wehavetoaddsome dependenciesinthemanifestfilesbutithassomelimitations aswecannotperformmajororheavytaskusingchaquopy.

Fig 2: FlowchartofKisanMitra

Oncetheuserstarttheapplicationthelanguageselection appears we had provided 3 languages English Hindi and Marathi. The next page depends on the user choice , if the user choose the English language the content will be in EnglishlanguageandifMarathithecontentwillbeinMarathi sameasforHindi.Accordingtothechosenlanguagetheuser cantalkwiththebotandcanusetheapplicationinthesame language.Ifthebotisnotabletoanswerthequerythebot will directly search it on the Google for this we had used beautiful soup library in python which is used for web scrapingusingthiswehadremovedalimitationastheusing oursystemtheusercanaskfurthermorequeriesapartfrom farmbasedanditwillalsotrytolearniftheuserwillaskthat questionagainsothebotwillbeabletoanswerthequery.We hadalsoimplementedotherfeaturesinourapplicationsuch as weather information, crop information, market information,governmentschemesandmuchmore.Weare extractinglivedataforallfeaturesfromgovernmentbased officialwebsites.

4. Result

Ifafarmerwantstoknowanythingforanypurposeandifhe don'tknowhowtooperatewithGooglethanitbecomesat most not possible to get information from other end so to overcome this problem we had created an application to assistfarmersfortheirqueries.Usingourapplicationnamed KisanMitrathefarmerswillbeabletogetthesolutionforthe queriestheyasked.Andifthefarmerisnotknownaboutthe English language we had also provide other regional languagessuchasHindiandMarathiforthecommunication. So there will be no barrier for language. We had provided other features so that the user can directly access that featuresfromthemainscreenonly.Thefeaturesare

5.1 Weather

Fig -3:WeatherPage

The fig no.8.2 shows the gateway request for get post serviceforourweatherapplication.Thisisusedinthispartis theAPIforliveweatherfetchingfunctionalityaccordingto thecity.ThiskeyisfetchingdatafromgooglebyusingJason dependencies in our java based application. The user can easily use our feature by using our bot or the specified featurepresentintheapplication.

Fig

:GovernmentSchemesPage

Market Information

Fig 6:GovernmentSchemesAPI

Thefigno.6showsthegatewayrequestforgetpostservice forourgovernmentschemeapplication.TheAPIurlistheapi key from the website vikaspedia which is an government managing website for the schemes for farmers and other workers like fisherman etc, which has all the updated governmentschemesandwearefetchinginformationfrom thissiteanddisplayingitonourapplicationscreen.Theuser can access this feature from the navigation panel and the featurenamedasGovernmentScheme.

MarketInformation

:MarketInformationAPI

Thefigno.8showsthegatewayrequestforgetpostservice forourmarketapplication.TheAPIurlistheapikeyfromthe website data.gov.in which has all the updated market informationwearefetchinginformationfromthatsiteand displayingitonourapplicationscreen.Theusercanaccess this feature from the navigation panel named as Market Information.

5.4 Chatbot System

We had provided chatbot in our application in three languages as English, Hindi and Marathi. We had also implementedwebscrapingforthiswehadusedbs4library whichisalsoknownasbeautifulsouplibraryinpython.As wehadmadeapplicationonandroidstudioplatformsowe hadusedChaquopytointegratepythoninandroidstudio.For EnglishchatbotsectionwehadcreatedDatasettotrainout bottoanswerqueriesaccordingtothephrasesastheuser

helps in

as

CONCLUSION

Project concludes that creating such a framework is conceivable. Farming is the spine of the economy. It gives nourishment,crudematerialsandindeedbusinessopenings to so numerous individuals. It's been practiced for over thousandsofalongtimepresently.Overthelongtime,dueto theheadwayandmostrecentpatternsininnovation,agrarian strategieshavemoreoveradvanced,butinIndia,thereare littleandnegligiblerancherswhohoneancient,conventional strategiesduetoneedofassets.Thus,thisframeworkworks asavirtualrighthandwhichguidestheranchersbyreplying alltheirinquirieswithrespecttoruralhones,inthismanner makingadifferencethemcreatehigherbenefits.Infuturethe chatmaybeempoweredtosendinquirieswithintheframeof

picturesandvideosandgetthecorrectreaction.Futurescope would be empowering sound and video calling highlights withthemaster,empoweringvideoandpicturereactionfrom thechatbotandGenerally,thisactivityofactualizingchatbots willhelpranchersandsubsequentlyboosttheeconomy.

6. FUTURE WORK

Infuture,wecanupgradeourapplication,aswecanaddthe otherregionallanguagestomakeitmoreefficienttotheusers ofotherreligions,sothattheywillbealsoabletosearchthe information and will be able o use our features more efficiently without hesitating. We can also add some more featuresastherequirementoftheapplicationusersinthe futuretomakeourapplicationmorereliableonthebasisof problemsfacedduringtheusageandotherreliablefeatures accordingtothefeedback&etc.Infuturewecanalsomakeit handierandmoreefficientsothelessliterateuserscanalso useitmoreefficiently,effectivelyandeasily.

REFERENCES

[1] Ramya .C, Shreya .R, Sowmiya .R ,” Virtual Conversational Assistant The FARMBOT”, International Journal of Engineering Technology Science and Research IJETSR,Volume5,Issue3,March2022.

[2]Prof.Yashaswini.D,Hemalatha.,Niveditha.(2019),”Smart ChatbotforAgriculture“,InternationalJournalofEngineering Science and Computing, Volume 9 Issue No. 5 , May 2021. [2]Proc.ACMInteract.Mob.WearableUbiquitousTechnol.,Vol. 2,No.4,Article170,December2021.

[3]AakashGRatkal,GangadharAkalwadi,VinayNPatiland KaviMahesh,(2019),“Farmer’sAnalyticalAssistant”,IEEE InternationalConference on Cloud Computing in Emerging [4]P.Jothimurugan,J.MuthuSaravanan,R.Sushanth,V.Suresh, H.SivaSubramaniam,S.Vasantharaj,S.Yogeswaran,SriEshwar College of Engineering, Coimbatore,”Solar E Bot for Agriculture”, 2018 Texas Instruments India Educators' Conference,2018IEEE

[5]K.D.Patel,”Review on Techniques in Natural Language Processing”,International Journal of Scientific Research in Research Paper.Computer Science and Engineering,Vol.7, Issue.5,pp.01 04,October(2018).

[6]Sweta P. Lende and M M Raghuwanshi, "Question AnsweringSystemonEducationActsusingNLPTechniques", IEEE sponsored word conference on futuristic trends in ResearchandInnovationforSocialWelfare,2017.

[7]S. Pudumalar, E. Ramanujam, R. Harine Rajashreen, C. Kavyan,T.KiruthikanandJ.Nishan,"CropRecommendation SystemforPrecisionAgriculture",IEEEEighthInternational ConferenceonAdvancedComputing,March2016.nn

International Research Journal of Engineering and Technology (IRJET) e ISSN:2395 0056

09

2022 www.irjet.net p ISSN:2395 0072

[8]Gourish Malage,Kiran Patil,”Raita Snehi A Voice Based Farmer Information System”,International Journal of ScientificresearchinResearchPaper.ComputerScienceand Engineering,Vol7,Issue6,pp.347 352,Jun 2016.n\

[9]VanditaMathad,GreeshmaR.R.,HarshithaJ.V.,DeepikaS., SnigdhaSen,”QualityAssessmentofCropsThroughDisease DetectionU

[10]M. C. Padma and P. A. Vijaya, “Identification and SeparationofTextWordsofKannada,Telugu,Tamil,Hindi, EnglishLanguagesthroughVisualDiscriminatingFeatures”, [2016][12]Inproc.ofInternationalconferenceonAdvances in Computer Vision and Information S.Pudumalar, E.Ramanujam,R.HarineRajashreen,C.Kavyan,T.Kiruthikan, J.NishanThiagarajarCollegeofEngineering,[2015]

[11]”Crop Recommendation System for Precision Agriculture”, IEEE Eighth International Conference on AdvancedComputing,March2016

[12]J.Browne.(2008,May23).“FromToilettoTapDiscover. [Online]Available: http:// discover magazine. com/ 2008/ may/23 from toilet to tap

[13]. A. Mulrine. (2006, March). “To the Rescue.” Prism. [Online]. Available: www.prism magazine.org/ mar 06/ feature_incredibles.cfm

[14]. (2005) “Provide Access to Clean Water.” National AcademyofEngineeringGrandChallengesforEngineering. [Online].