You are on page 1of 24

MLDICAL kLSCkI1ICN CNLINL

8?
MPA8SPlnl (1210308433)
CPSAl SuSMl1PA (1210308412)
88LSPMA 8PAv?A (1210308443)
CPAvAn kuMA8 (1210308413)

unuL8 1PL CuluAnCL Srl 88AkASP
M1ech(hu)
ASSlS1An1 8ClLSSC8
uL1 Cl CSL



A8S1kAC1
1he maln ldea of Lhe pro[ecL ls Lo glve proper medlcal prescrlpLlon Lo Lhe
cusLomer who logs ln Lo Lhe slLe and Lells abouL hls/her problem lor Lhls a
webslLe ls creaLed where a user can logln 1hen Lhe user can dlrecLly speak and
Lell hls/her problem A daLabase whlch conslsLs of dlseases and Lhelr sympLoms
and Lhe medlclne Lo be used Lo cure LhaL parLlcular dlsease ls made When Lhe
cusLomer/paLlenL speaks Lhe speech ls recognlzed and lL ls converLed Lo LexL and
lL ls maLched wlLh Lhe sympLoms ln Lhe daLabase and Lhe flle LhaL maLches more
wlLh Lhe LexL generaLed from speech ls reLrleved and lL quesLlons Lhe paLlenL
abouL oLher sympLoms Lo flnd ouL wheLher he had Lhem or noL lf all Lhe
sympLoms maLch Lhen Lhe parLlcular dlsease ls conflrmed and Lhe medlclnes Lo be
used Lo cure LhaL speclflc dlsease ls reLrleved and suggesLed Lo Lhe user le glven
as a prescrlpLlon 1hls ls used noL only Lo glve proper medlcal prescrlpLlon Lo a
paLlenL buL also Lo provlde baslc medlcal lnformaLlon/guldance Lo Lhe users









CCN1LN1S
1 ln18CuuC1lCn
2 8ACkC8Cunu S1uu?
21 SLLCP 1LCPnCLCC?
22 SLLCP 8LCCnl1lCn
23 SLLCP 8LCCCnlZL8
AuvAn1ACLS
ulSAuvAn1ACLS
3 A8CPl1LC1u8L
4 lMC81AnCL
3 8lnClLLS Cl S?S1LM AnAL?SlS
6 uML ulAC8AMS
61 uSL CASL
62 AC1lvl1?
63 SLCuLnCL
64 CCLLA8C8A1lCn
63 CLASS
7 uA1A8ASL SCPLMA
8 PA8uWA8L Anu SCl1WA8L 8LCul8LMLn1S
9 CCnCLuSlCn
10 8LlL8LnCLS

IN1kCDUC1ICN
ln Lhls busy world every one doesn'L have enough Llme Lo vlslL a docLor for every
small problem Many people consulL a docLor only when Lhe problem becomes
severe unLll Lhen Lhey follow self medlcaLlon or else Lhey wlll use Lhe medlclnes
LhaL someone suggesLs SomeLlmes Lhey may noL workouL and Lhere are cases
LhaL caused some slde effecLs
So ln Lhls slLuaLlon lL would be more convlnclng and convenlenL lf we
can geL Lhe proper and correcL prescrlpLlon from lnLerneL As now a days lnLerneL
ls belng used by almosL all lndlvlduals lL would be exclLlng and admlrlng lf we can
geL proper medlcal asslsLance by uslng lL le by slmply explalnlng our problem
Lhrough volce [usL as lf we Lalk Lo a docLor ln real llfe Cur pro[ecL helps ln dolng
so
A webslLe ls creaLed where Lhe person can logln wlLh a glven user ld and
password lor Lhls he needs Lo geL reglsLered by provldlng some of hls deLalls
1hen Lhe reglsLered person flrsL logs ln Lo Lhe slLe and Lhen he descrlbes hls
problem by dlrecLly speaklng Lo Lhe sysLem We Lhen recognlze Lhe speech spoken
by hlm and converLs lL Lo LexL by uslng dragon naLurally free speech recognlLlon
sofLware Cn Lhe oLher hand we wlll develop a daLabase LhaL conslsLs of dlseases
and Lhelr sympLoms and Lhe medlclnes LhaL should be used Lo cure LhaL
dlseasenow Lhe LexL LhaL ls converLed from speech ls maLched wlLh Lhe daLabase
1hls ls done by uslng [dbc / odbc daLabase connecLlvlLy
1he flle LhaL maLches more wlLh Lhe sympLoms sald by Lhe cusLomer ls
reLrleved and Lhe oLher sympLoms ln LhaL flle are quesLloned Lo know wheLher
Lhe paLlenL had Lhem or noL 1he yes/no answer are Laken from Lhe user Lo
conflrm Lhe sympLoms
8y flndlng ouL Lhe deLalls abouL all Lhe sympLoms lL ldenLlfles Lhe
dlsease and speclfles Lhe medlclnes Lo be used Lo cure LhaL parLlcular dlsease by
glvlng Lhe prescrlpLlon
1he maln ob[ecLlve of Lhe pro[ecL ls Lo faclllLaLe Lhe users by provldlng
Lhe lnformaLlon Lhey requlre And also lL helps ln provldlng Lhe baslc medlcal
lnformaLlon abouL varlous dlseases Lhelr sympLoms and Lhe medlclnes Lo be
used

8ACkGkCUND S1UD
1he uomaln under whlch Lhe pro[ecL comes under or Lhe pro[ecL
malnly concerned wlLh ls SLLCP 1LCPnCLCC?
Speech techno|ogy relaLes Lo Lhe Lechnologles deslgned Lo
dupllcaLe and respond Lo Lhe human volce 1hey have many uses lncludlng Lo ald
Lhe volcedlsabled Lhe hearlngdlsabled Lhe bllnd and Lo communlcaLe wlLh
compuLers wlLhouL a keyboard Lo markeL goods or servlces by Lelephone and Lo
enhance game sofLware
1hls speech Lechnology lncludes several sub flelds 1hey are
Speech synLhesls
Speech recognlLlon
Speaker recognlLlon
Speaker verlflcaLlon
Speech compresslon
MulLlmodal lnLeracLlon
` 1he deLalls of Lhese subflelds ln brlef are as follows
Speech synthes|s
A speech synLheslzer converLs wrlLLen LexL lnLo spoken language Speech
synLhesls ls also referred Lo as LexLLospeech (11S) converslon
Speech recogn|t|on
Speech recognlLlon ls Lhe process of converLlng spoken language Lo wrlLLen LexL
or some slmllar form
Speaker recogn|t|on
Speaker recognlLlon ls Lhe compuLlng Lask of valldaLlng a users clalmed ldenLlLy
uslng characLerlsLlcs exLracLed from Lhelr volces 1here ls a dlfference beLween
speaker recognlLlon (recognlzlng who ls speaklng) and speech recognlLlon
(recognlzlng whaL ls belng sald)
Speech ver|f|cat|on
speech verlflcaLlon uses speech recognlLlon Lo verlfy Lhe correcLness of Lhe
pronounced speech Speech verlflcaLlon doesnL Lry Lo decode unknown speech
from a huge search space buL lnsLead knowlng Lhe expecLed speech Lo be
pronounced lL aLLempLs Lo verlfy Lhe correcLness of Lhe uLLerances
pronunclaLlon plLch and sLress
Speech compress|on
speech compresslon may mean dlfferenL Lhlngs
Speech encodlng refers Lo compresslon for Lransmlsslon or sLorage
posslbly Lo an unlnLelllglble sLaLe wlLh decompresslon used prlor Lo
playback
1lmecompressed speech refers Lo volce compresslon for lmmedlaLe
playback wlLhouL any decompresslon (so LhaL Lhe flnal speech sounds
fasLer Lo Lhe llsLener)

1he subfleld LhaL ls very much used ln our pro[ecL and our
pro[ecL malnly based on ls speech recognlLlon

SLLCn kLCCGNI1ICN
Speech 8ecognlLlon are Lechnologles of parLlcular lnLeresL
for Lhelr supporL of dlrecL communlcaLlon beLween humans and compuLers
Lhrough a communlcaLlons mode humans commonly use among Lhemselves and
aL whlch Lhey are hlghly skllled lL provldes compuLers wlLh Lhe ablllLy Lo llsLen Lo
spoken language and Lo deLermlne whaL has been sald ln oLher words lL
processes audlo lnpuL conLalnlng speech by converLlng lL Lo LexL
Speech recognlLlon ls a broader soluLlon whlch refers Lo Lechnology
LhaL can recognlze speech wlLhouL belng LargeLed aL slngle speakersuch as a call
sysLem LhaL can recognlze arblLrary volces
ln slmple words Speech recognlLlon ls Lhe process of converLlng
spoken language Lo wrlLLen LexL or some slmllar form 1o do Lhls a speech
recognlzer ls used whlch recognlzes Lhe speech and Lhen converLs Lhe speech
ln Lo Lhe LexL formaL

8as|c character|st|cs of speech recogn|zer
lL ls mono llngual lL supporLs a slngle speclfled language
lL processes slngle lnpuL audlo sLream
lL can opLlonally adapL Lo Lhe volce of lLs users
lL has a small deflned seL of appllcaLlonconLrollable properLles
lLs grammar can be dynamlcally updaLed

uesplLe Lhese characLerlsLlcs Lhere are also many ma[or sLeps LhaL
should be followed by a Lyplcal speech recognlzer 1hey are as follows

1he ma[or steps of a typ|ca| speech recogn|zer are
% Grammar des|gn
8ecognlLlon grammars deflne Lhe words LhaL may be spoken by a
user and Lhe paLLerns ln whlch Lhey may be spoken A grammar musL be
creaLed and acLlvaLed for a recognlzer Lo know whaL lL should llsLen for ln
lncomlng audlo
% S|gna| process|ng
Analyze Lhe specLrum (frequency) characLerlsLlcs of Lhe lncomlng
audlo
% honeme recogn|t|on
Compare Lhe specLrum paLLerns Lo Lhe paLLerns of Lhe phonemes of
Lhe language belng recognlzed
% Word recogn|t|on
Compare Lhe sequence of llkely phonemes agalnsL Lhe words and
paLLerns of words speclfled by Lhe acLlve grammars
O kesu|t Generat|on
provlde Lhe appllcaLlon wlLh lnformaLlon abouL Lhe words Lhe
recognlzer has deLecLed ln Lhe lncomlng audlo 1he resulL lnformaLlon ls
always provlded once recognlLlon of a slngle uLLerance (ofLen a senLence) ls
compleLe buL may also be provlded durlng Lhe recognlLlon process 1he resulL
always lndlcaLes Lhe recognlzers besL guess of whaL a user sald buL may also
lndlcaLe alLernaLlve guesses

Advantages of speech recogn|zer
Speech recognlLlon can lmprove compuLer accesslblllLy for users wlLh
dlsablllLles and can reduce Lhe rlsk of repeLlLlve sLraln ln[ury and oLher
problems caused by currenL lnLerfaces
Speech recognlLlon can also reduce Lhe amounL of Llme an agenL needs Lo
remaln on a llne lncreaslng Lhe number of calls LhaL agenL can effecLlvely
handle wlLhln Lhelr shlfL
D|sadvantages of speech recogn|zer
Lven Lhe besL speech recognlLlon sysLems someLlmes make errors lf Lhere
ls nolse or some oLher sound ln Lhe room (eg Lhe Lelevlslon or a keLLle
bolllng) Lhe number of errors wlll lncrease
Speech 8ecognlLlon works besL lf Lhe mlcrophone ls close Lo Lhe user (eg
ln a phone or lf Lhe user ls wearlng a mlcrophone) More dlsLanL
mlcrophones (eg on a Lable or wall) wlll Lend Lo lncrease Lhe number of
errors

now leL us conslder a speech recognlzer for example whlch ls used Lo recognlze
Lhe speech and converL lL Lo LexL 1hls uses Lhe uragon naLurally free speaklng
sofLware





AkCnI1LC1UkL
1he archlLecLure of a speech recognlzer whlch uses dragon naLurally free speaklng
sofLware ls as follows



ln Lhls Lhe program code ls wrlLLen ln [ava 1he program Lakes
Lhe volce lnpuL and Lhe program ls wrlLLen Lo Lake Lhe lnpuL and uslng [ava speech
Al cloud garden Lalklng [ava sdk and dragon sofLware Lhe speech ls converLed Lo
LexL sLrlng Sapl funcLlon calls refers Lo speech apl funcLlon calls


lMC81AnCL
lL has a loL of lmporLance because lL provldes many faclllLles Lo Lhe user some of
Lhe ma[or conLrlbuLlons made by Lhe pro[ecL Lo Lhe users are
8y uslng Lhls slLe Lhe Llme wlll be saved for Lhe people
who lgnores Lo consulL a docLor for small medlcal
problems
lL would be much more beLLer Lhan self medlcaLlon and
Lhe proper medlclnes for Lhe dlsease are glven
eople wlll be able Lo geL a prlor knowledge abouL Lhe
dlsease and lLs sympLoms and Lhe medlclnes Lo be used
kINCILLS CI SS1LM ANALSIS
LeL us analyse our pro[ecL ln an ob[ecL orlenLed approach lor Lhls we use Lhe
concepLs of ob[ecL orlenLed analysls and deslgn le Lhe proposed sysLem ls
analysed uslng Lhe uml dlagrams class dlagrams sequence dlagram
collaboraLlon dlagrams acLlvlLy dlagrams and use case dlagrams
1he explanaLlon of Lhe pro[ecL ln deLall wlLh Lhe help of uml dlagrams ls as
follows
LeL us flrsL sLarL Lhls wlLh Lhe help of use case dlagrams ln Lhe use
case dlagram of Lhe proposed sysLem Lhe acLors are cusLomer speech recognlzer
and daLa base admlnlsLraLor 1he cusLomer plays Lhe role of explalnlng hls/her
problem and answers Lhe quesLlons asked and Lakes Lhe prescrlpLlon 1he speech
recognlzer plays Lhe role of Laklng or recelvlng le recognlzlng Lhe speech spoken
by Lhe user and converLlng lL Lo LexL 1he daLabase admlnlsLraLor plays Lhe role of
maLchlng Lhe LexL wlLh daLabase and searchlng or ldenLlfylng and approprlaLe
dlsease and suggesLlng Lhe medlclne




I|g 61 Use case d|agram


converting it to text
match text with database
identify the disease
approximately
find out other symptoms
confirm the disease
database
administrator
prescribe medicine
customer
speech recognizing
speech
recognizer
now leL us conslder Lhe acLlvlLy dlagram of Lhe proposed sysLem 1he acLlvlLy
dlagram ls a flow charL represenLaLlon of Lhe proposed sysLem lL ls essenLlally a
flowcharL showlng flow of conLrol from acLlvlLy Lo acLlvlLy lL has sLarL and end
symbols ln Lhe acLlvlLy dlagram of our proposed sysLem Lhere are Lwo declslon
symbols whlch are dlamond shaped Cne ls used Lo flnd wheLher all Lhe
sympLoms maLches or noL and Lhe oLher ls used Lo flnd ouL wheLher Lhe person
had Lhose sympLoms or noL
1he acLlvlLy dlagram of Lhls pro[ecL brlefly explalns Lhe flow of
conLrol of Lhe pro[ecL le whaL sLep or whaL process ls Laklng place afLer whaL 1he
LoLal descrlpLlon of Lhe pro[ecL ls explalned brlefly ln a sequenLlal manner sLarLlng
from Lhe flrsL sLep Lo Lhe lasL sLep Craphlcally an acLlvlLy dlagram ls a collecLlon
of verLlces and arcs

I|g 62 Act|v|ty d|agram
login
user
speaks
recognizes the speech
and converts it to text
match the text
with databases
check whether all
symptoms matches
ask for other
symptoms
confirm the
disease
give
prescription
yes
no
no
yes

now leL us conslder sequence dlagramln sequence dlagram Lhere wlll be ob[ecLs
and messages 1he ob[ecLs LhaL parLlclpaLe ln lnLeracLlon are placed along xaxls
and Lhe messages LhaL Lhese ob[ecLs send and recelve are along yaxls Pere ln
our pro[ecL model Lhe ob[ecLs are user slLe speech recognlzer and converLer
and daLabase 1he messages lnclude explalnlng hls problem sendlng Lhe speech
glvlng Lhe prescrlpLlon and so on
1hls followlng sequence dlagram glves Lhe reader a clear vlsual cue
Lo Lhe flow of conLrol over Llme
1he sequence dlagram for Lhe pro[ecL ls as follows


I|g 63 Sequence d|agram




user site speech recognizer
and converter
database
login
explains his problem
sends the speech
recognizes the data and converts it to text
match the data
asks for other symptoms
reply
sends the reply
identifies the disease
gives the prescription
A collaboraLlon dlagram emphasls Lhe organlzaLlon of Lhe ob[ecLs LhaL parLlclpaLe
ln Lhe lnLeracLlon 1hls glves Lhe reader a clear vlsual cue Lo Lhe flow of conLrol ln
Lhe conLexL of sLrucLural organlzaLlon of Lhe ob[ecLs LhaL collaboraLe
1he collaboraLlon dlagram of Lhe proposed sysLem ls as follows


I|g 64 Co||aborat|on d|agram

unllke ln sequence dlagram here Lhe ob[ecLs LhaL parLlclpaLe ln Lhe lnLeracLlon are
placed as Lhe verLlces and Lhe llnks LhaL connecL Lhese ob[ecLs are Lhe arcs of Lhe
graph

user site
speech recognizer and
converter
database
1: login
2: explains his problem
3: sends the speech
4: recognizes the data and converts it to text
5: match the data
6: asks for other symptoms
7: reply
8: sends the reply
9: identifies the disease
10: gives the prescription
CLASS DIAGkAM
1he class dlagram for Lhe proposed sysLem ls as follows


I|g 6S C|ass d|agram
1he class dlagram ls used Lo model Lhe sLaLlc deslgn vlew of a sysLem 1hls vlew
prlmarlly supporLs funcLlonal requlremenLs of a sysLemLhe servlces sysLem
should provlde Lo lLs end users
1he class dlagram malnly conslsLs of classes lnLerfaces and packages
Pere ln Lhe pro[ecL Lhe classes are sympLoms user daLabase sysLem
languages And Lhese classes conslsLs of cerLaln aLLrlbuLes and operaLlons

now leL us conslder Lhe daLabases LhaL are used ln Lhe pro[ecL Lhelr schemas L
8 dlagrams sLrucLure of Lhe Lables
english(british)
english(US)
languages
name
database
symptoms
name of the diseases
medicines
symptoms
name
intensity
systemss
name
33
3
user
name
age
3
3

Database Schema
1he daLa base schema conslsLs of malnly Lhree Lables whlch are dlsease sympLom
and medlclne
1he Lables are






1he above
Lables conLaln lnformaLlon abouL Lhe dlsease name and lLs ld
1he schema for dlsease relaLlon ls
dlsease(dlddname)
1he schema for sympLom relaLlon ls
SympLom(sldsname)
1he schema for LableL relaLlon ls
Medlclne (mld mname company cosL)



dld uname
11101 lever
11102 !aundlce
11103 AsLhma
11104 ulabeLes
11103 Palr loss
Sld Sname
11201 Cough
11202 Cold
11203 Weak
11204 Peadache
11203 8ody paln
1he LableL relaLlon showlng LhaL name of Lhe LableL ld company and lLs cosL And
Lhe Lable ls
Mld Mname Company CosL
121 Crocln AmblenL 20
122 aroxeLlne axll 13
123 8amelLon 8ozerom 18
124 SpecLral AcebuLolol 23
123 recose Acarbose 33

1he relaLlon for cure ls


1he cure relaLlon conLalns Lhe dlsease ld sympLom ld and medlclne name 8ased
on Lhe dlsease and sympLom we can geL Lhe relevanL medlclne Lo cure our
dlsease 1he schema for Lhls relaLlon ls
Cure(dldsldmname)

uld Sld Mname
11101 11201 aroxeLlne
11101 11204 recose
11103 11204 8amelLon
11104 11203 Croclne
11102 11202 recose
Schema dlagrams
A daLabase schema along wlLh prlmary key and forelgn key dependencles can be
deplcLed by schema dlagrams 1he followlng flgure shows Lhe schema dlagram for
Lhe daLabase Lach appears as a box wlLh Lhe relaLlon name aL Lhe Lop and Lhe
aLLrlbuLes llsLed lnslde Lhe box rlmary key aLLrlbuLes are underllned lorelgn key
dependencles appear as arrow from Lhe forelgn key aLLrlbuLes of Lhe referenclng
relaLlon Lo Lhe prlmary key of Lhe referenced relaLlon











Medlclne
Mld
Mname
Company
CosL

ulsease
uld
uname
Cure
uld
Sld
Mname
Symptom
Sid
Sname
HARDWARE AND SOFTWARE REQUIREMENTS:
lnLel enLlum lll / 300 MPz processor
2 C8 8AM
MlcrosofL Wlndows x Pome and rofesslonal
Mlllennlum 2000
CreaLlve Sound 8lasLer 16 or equlvalenL sound card
supporLlng 16blL recordlng MlcrosofL lnLerneL
Lxplorer 3 or hlgher
ScanSofLapproved nolsecancellng headseL
Mlcrophone Speakers
A Web connecLlon ls requlred for acLlvaLlon
A speech recognlzlng sofLware ls used







CCNCLUSICN AND IU1UkL WCkk
lor Lhls pro[ecL unLll now we sLudled abouL volce appllcaLlon
sofLwares le Lhe sofLwares LhaL are used Lo recognlze Lhe speech
and converL lL Lo LexL ldenLlfylng Lhe sofLware and sLudylng Lhe
code ls done As Lhe sofLware we found ouL ls able Lo recognlze
only few words Lhe developmenL of LhaL sofLware code ls belng
made 1he ouLlook of Lhe webslLe ls deslgned and Lhe code ls Lo
be wrlLLen
We are plannlng Lo conLlnue our work on developlng
Lhe sofLware Lo recognlze many words le long speech and we
need Lo Lraln Lhe sofLware Lo recognlze dlfferenL accenLs we have
Lo work on creaLlng Lhe daLabases for sLorlng medlcal lnformaLlon
and collecLlng LhaL lnformaLlon







kLILkLNCLS
1 Avlad oLmazgln Amlr 8aron 2011 Saya free speech
recognlLlononllne
Avallable hLLp//wwwcsbguacll/orlovm/Leachlng/
sayareporLs/saya recognlLlondragonreporLpdf
2 CPLee lkSoong and kkpallwal kluwer 8osLon (1996)
AuLomaLlc speech and speaker recognlLlononllne
Avallable
hLLp//booksgooglecom/books/abouL/AuLomaLlc_speech_and_
speaker_recognlLlonhLml?ldhkvh_kfuC88C
3 Sun MlcrosysLems !ava speech Al programmers gulde
(1
sL
edlLlon) Cnllne
Avallable hLLp//[avasuncom/producLs/[avamedla/
speech/foruevelopers/[saplgulde/
4 hLLp//wwwlearnarLlflclalneuralneLworkscom/speech
recognlLlonhLml
3 v ulgalakls L neumeyer and M erakakls CuanLlzaLlon
of CepsLral arameLers for Speech 8ecognlLlon Cver Lhe
World Wlde Web"roceedlngs lCASS 98 pp 989992 1998
6 hLLp//supporLmlcrosofLcom/kb/110093

You might also like