components/dom_distiller/core/distillable_page_detector.h

   1 // Copyright 2015 The Chromium Authors. All rights reserved.
   2 // Use of this source code is governed by a BSD-style license that can be
   3 // found in the LICENSE file.
   4
   5 #ifndef COMPONENTS_DOM_DISTILLER_CORE_DISTILLABLE_PAGE_DETECTOR_H_
   6 #define COMPONENTS_DOM_DISTILLER_CORE_DISTILLABLE_PAGE_DETECTOR_H_
   7
   8 #include <vector>
   9
  10 #include "base/memory/scoped_ptr.h"
  11 #include "components/dom_distiller/core/proto/adaboost.pb.h"
  12
  13 namespace dom_distiller {
  14
  15 // DistillablePageDetector provides methods to identify whether or not a page is
  16 // likely to be distillable based on a vector of derived features (see
  17 // dom_distiller::CalculateDerivedFeatures). It uses a simple AdaBoost-trained
  18 // model.
  19 class DistillablePageDetector {
  20  public:
  21   static const DistillablePageDetector* GetDefault();
  22   explicit DistillablePageDetector(scoped_ptr<AdaBoostProto> proto);
  23   ~DistillablePageDetector();
  24
  25   // Returns true if the model classifies the vector of features as a
  26   // distillable page.
  27   bool Classify(const std::vector<double>& features) const;
  28
  29   double Score(const std::vector<double>& features) const;
  30   double GetThreshold() const;
  31  private:
  32   scoped_ptr<AdaBoostProto> proto_;
  33   double threshold_;
  34   DISALLOW_COPY_AND_ASSIGN(DistillablePageDetector);
  35 };
  36
  37
  38 }  // namespace dom_distiller
  39
  40 #endif  // COMPONENTS_DOM_DISTILLER_CORE_DISTILLABLE_PAGE_DETECTOR_H_