/usr/include/libGIFTQuInvertedFile/include/CWeightingFunction.h is in libgnuift0-dev 0.1.14-12.1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 | /* -*- mode: c++ -*-
*/
/*
GIFT, a flexible content based image retrieval system.
Copyright (C) 1998, 1999, 2000, 2001, 2002, CUI University of Geneva
Copyright (C) 2003, 2004 Bayreuth University
2005 Bamberg University
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
/** -*- mode: c++ -*- */
/**
*
* CWEIGHTINGFUNCTION
*
*
*
* modification history:
*
* HM 090399 created some documentation
*
* @author Wolfgang Müller
*
* compiler defines used:
*
*
*/
#ifndef _CWEIGHTINGFUNCTION
#define _CWEIGHTINGFUNCTION
#include "libGIFTQuInvertedFile/include/uses-declarations.h"
#include <cmath>
#include "libMRML/include/TID.h"
#include "libGIFTAcInvertedFile/include/CDocumentFrequencyElement.h"
#include "libGIFTAcInvertedFile/include/CAcInvertedFile.h"
class CQueryNormalizer;
/** a universally usable class for weighting.
The rationale is, that we want to implement very flexible
feature weighting, so flexible that it does not necessarily
have anything to do with the initial concept of feature
weighting.
The scenario: We have query/feedback from the user. That is:
We have a list of images with either positive or negative
relevance for the user. Each of these images contains a list
of features.
Now we want to combine the different features. What we
usually need to know is: how many images where in
the positive feedback, and how many images where
in the negative feedback.
*/
class CWeightingFunction{
protected:
/** This value depends on all the elements of the query, which
have a positive user-assigned relevance.
i.e. not necessarily only on the elements which contain
this feature, so it has to be explicitly set using
the function setRelevanceSum from the outside.
If relevance=+-1 then this is the number of positive images in the query
*/
double mPositiveRelevanceSum;
/** This value depends on all the elements of the query, which
have a negative user-assigned relevance.
i.e. not necessarily only on the elements which contain
this feature, so it has to be explicitly set using
the function setRelevanceSum from the outside.
If relevance=+-1 then this is the number of negative images in the query
*/
double mNegativeRelevanceSum;
/* A pseudo term frequency calculated from the input
at present by calculating the mean.
mPositiveTermFrequency/mPositiveRelevanceSum is one term in a
a pseudo term frequency used for calculating query weights */
double mPositiveTermFrequency;
/* A pseudo term frequency calculated from the input
at present by calculating the mean.
mNegativeTermFrequency/mNegativeRelevanceSum is the other term in a
a pseudo term frequency used for calculating query weights */
double mNegativeTermFrequency;
/**
At present the feature description is a simple integer which expresses
if the feature is a histogram feature or not and if it is a texture feature
or not. We probably will use more elaborate descriptions in the future.
*/
int mFeatureDescription;
/** */
TID mID;
/** The accessor for the file on which the query is performed */
const CAcInvertedFile* mAccessor;
/** FIXME THERE IS A GOOD DESCRIPTION MISSING FOR THE NEXT
TWO VARIABLES */
mutable CQueryNormalizer* mQueryNormalizer;
/** A normalizer for apply on self
*/
mutable CQueryNormalizer* mThisNormalizer;
/**
The query usually will separate into a part
which is equal for all the query and a part which is to be
calculated for each document. The part which stays constant
*/
double mQueryFactor;
/** */
double mDocumentFactor;
public:
/**
*
* The query factor contains all informatio which
* depends only on the query. (for efficiency)
*
*/
double getQueryFactor()const;
/**
*
* Calculate the QueryFactor.
*
*/
virtual void preCalculate();
/**
*
* Constructor:
* @param inAccessor: the accessor containing this \
* (for getting information about single documents)
* @param inQueryNormalizer: FIXME
* @param inThisNormalizer: FIXME
*/
CWeightingFunction(const CAcInvertedFile* inAccessor=0,
CQueryNormalizer* inQueryNormalizer=0,
CQueryNormalizer* inThisNormalizer=0);
/**
*
* Set the accessor (should not be needed)
*
*/
void setAccessor(const CAcInvertedFile*);
/**
*
* Set the query normalizer (should not be needed);
*
*/
void setNormalizers(CQueryNormalizer* inQueryNormalizer,
CQueryNormalizer* inThisNormalizer);
/**
*
* Set the ID of this feature
*
*/
void setID(TID);
/**
*
* get the ID of this feature
*
*/
TID getID()const;
/**
*
* Setting the relevance sum,
* that is the sum of the absolute values of all
* relevance levels of all documents of the query
*
* for rocchio we need this sum to be once calculated
* for positive, once for negative feedback images
*
*/
void setRelevanceSum(double inPositiveRelevanceSum,
double inNegativeRelevanceSum);
/**
*
* Add a feature. A feature will be a document frequency element
* out of the feature file.
*
*/
virtual void addQueryFeature(double inRelevanceLevel,
const CDocumentFrequencyElement&
inQueryFeature);
/**
*
* Giving a pseudo term frequency for normalization
*
*/
virtual double getTermFrequency()const;
/**
*
* Apply this on a document frequency
*
*/
virtual double subApply(const double inDocumentFrequency,
const double inNormalizingFactor)const;
/**
*
* Preparing the right parameters for subApply: Apply this on another feature
*
*/
virtual double apply(const CDocumentFrequencyElement& inResultFeature)const;
/**
*
* Preparing the right parameters for subApply: Apply this on itself
*
*/
double applyOnThis()const;
/**For the "prototype pattern":
Kinda virtual copy constructor:
Gamma/Helm/Johnson/Vlissides p. 107*/
virtual CWeightingFunction* constructNew(TID inID)const;
/**For the "prototype pattern":
Kinda virtual copy constructor:
Gamma/Helm/Johnson/Vlissides p. 107
This is a new version of constructNew which
manages to live without the setID function
*/
virtual CWeightingFunction* clone()const;
/** This virtual destructor is here inorder to call the (inherited)
default constructors
*/
virtual ~CWeightingFunction();
friend class CSortByDFTimesLogICF_WF;
};
/**
*
*
*
*/
bool operator<(const CWeightingFunction&,
const CWeightingFunction&);
/** CWeightingFunction: Sort by FeatureID */
class CSortByFeatureID_WF:
public binary_function
<CWeightingFunction,CWeightingFunction,bool>{
/** */
inline bool operator()(const CWeightingFunction& l,
const CWeightingFunction& t){
return l.getID()<t.getID();
}
};
/**
*
* CWeightingFunction: Sort by DocumentFrequency/CollectionFrequency
*
*/
class CSortByDFTimesLogICF_WF:
public binary_function
<CWeightingFunction,CWeightingFunction,bool>{
/** */
inline bool operator()(const CWeightingFunction& l,
const CWeightingFunction& t){
return
l.getTermFrequency()
*
fabs(log(l.mAccessor->FeatureToCollectionFrequency(l.getID())))
<
t.getTermFrequency()
*
fabs(log(t.mAccessor->FeatureToCollectionFrequency(t.getID())));
};
};
/**
*
* CWeightingFunction: Sort by DocumentFrequency/CollectionFrequency
*
*/
class CSortByQueryFactor_WF:
public binary_function
<CWeightingFunction,CWeightingFunction,bool>{
public:
/** */
inline bool operator()(const CWeightingFunction& l,
const CWeightingFunction& t){
return
l.getQueryFactor()
<
t.getQueryFactor();
};
};
/**
CWeightingFunction: Sort by the absolute value
of the query factor. This is useful for pruning
when the feedback is negative.
*/
class CSortByAbsQueryFactor_WF:
public binary_function
<CWeightingFunction,CWeightingFunction,bool>{
public:
/** */
inline bool operator()(const CWeightingFunction& l,
const CWeightingFunction& t){
return
fabs(l.getQueryFactor())
<
fabs(t.getQueryFactor());
};
};
/**
*
*
*
*/
template<class CSortOp>
class CSortPointers_WF:
binary_function<CWeightingFunction*,CWeightingFunction*,bool>{
protected:
/** */
CSortOp mSorter;
public:
/** */
inline CSortPointers_WF(){};
/** */
inline bool operator()(const CWeightingFunction* l,
const CWeightingFunction* t){
return mSorter(*l,*t);
};
};
#include "libGIFTQuInvertedFile/include/CQueryNormalizer.h"
#endif
|