blob: a54d1cdfebf598a2d2858746e7e7848fe27b53b1 [file] [log] [blame]
// Copyright 2015 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include <vector>
#include "base/values.h"
#include "url/gurl.h"
class GURL;
namespace dom_distiller {
// The length of the derived features vector.
extern int kDerivedFeaturesCount;
// The distillable page detector is a model trained on a list of numeric
// features derived from features of a webpage (like body's number of elements
// ). This derives the numeric features form a set of core features.
// Note: It is crucial that these features are derived in the same way and are
// in the same order as in the training pipeline. See //heuristics/distillable
// in the external DomDistiller repo.
std::vector<double> CalculateDerivedFeatures(bool isOGArticle,
const GURL& url,
double numElements,
double numAnchors,
double numForms,
const std::string& innerText,
const std::string& textContent,
const std::string& innerHTML);
// Calculates the derived features from the JSON value as returned by the
// javascript core feature extraction.
std::vector<double> CalculateDerivedFeaturesFromJSON(
const base::Value* stringified_json);
std::vector<double> CalculateDerivedFeatures(
bool openGraph,
const GURL& url,
unsigned elementCount,
unsigned anchorCount,
unsigned formCount,
double mozScore,
double mozScoreAllSqrt,
double mozScoreAllLinear);
} // namespace dom_distiller