Speech:Exps 0077

From Openitware
Jump to: navigation, search

Description

Author: Group A + D

Date: 4/9/2013

Purpose:

  • Create a new corpus from the first 5 hours of corpus data
  • Run a train and decode on the first 5 hours of corpus data

Details: Begin the process of running a full train on 5 hours worth of corpus data and run a test on the first 30 minutes of that data. After the corpus was created successfully there is a large amount of words to be added to the dictionary which are listed below. These words will need to be defined and added in order to progress to the next stage.

111 words to be added:

007
101
60
ABC
ABOUTS
ABRODE
ALBRIDGE
ALRIGHTY
ALWAY
AMA
ANANIMOUS
AOR
ATM
ATM'S
BABYSITTINGWISE
BASTROP
BSING
CAMPANY
CASSEROLES
CBS
CHAPE
CHOWPHERD
COGNIZITIVE
COMPOOTER
CONDITIONER'S
CONQUISTADORS
CPA
CULOTTES
DC
DC'S
DIESEL'S
DISEAL
DMC
DUCTWORK
EDS
ELLSWORTH'S
ENVIRONMENTALS
ESPN
EXERCISING'S
FEDERALDES
FINALIZATION
FREON'S
GESTALT
GM
GPA
GRE
GREENWARE
GROUND'S
GTE
HEARTRENDING
HM
HOUSEPAINTER
IBM
IBS
JFK
JUNKINS'
KGB
KLIF
LER
LEWISVILLE
LISTENABLE
LOA
LUCKED
MAHALS
MARKOV
MAY'VE
MOOSEWOOD
MOUNTAINEERING
MTV
NBC
NONSOLICITED
NOX
OGEN
PBS
PC
PCS
PHD
POSTGRADUATE
PRICEWISE
PTA'S
PURIFIES
RV
RVER
SCYENE
SHANKAR
SHEETROCK
SKORTS
SMUSH
SOCIOPATHIC
SOUTHBEND
SPOOKIER
SUPERVISOR'S
TACB
TENSIONED
TEXTELLER
THANKSGIVING'S
TI'S
TRANGED
TREE'S
TVS
UNAIR
UNSEASONAL
UNTYPICALLY
VACATION'S
VEGAN
VET'S
VOCALIZED
WANT'S
WEATHERWISE
YELLER
YUPPYISH

Results: TBD