Record scores of distillability and long article in UMA

components/dom_distiller/content/renderer/distillability_agent.cc

 // 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 "base/metrics/histogram.h"
 #include "base/strings/string_util.h"
 
 #include "components/dom_distiller/content/common/distillability_service.mojom.h"
 #include "components/dom_distiller/content/renderer/distillability_agent.h"
 #include "components/dom_distiller/core/distillable_page_detector.h"
@@ skipping 75 matching lines @@
   std::vector<double> derived = CalculateDerivedFeatures(
     features.openGraph,
     parsed_url,
     features.elementCount,
     features.anchorCount,
     features.formCount,
     features.mozScore,
     features.mozScoreAllSqrt,
     features.mozScoreAllLinear
   );
-  bool distillable = detector->Classify(derived);
-  bool long_article = long_page->Classify(derived);
+  double score = detector->Score(derived) - detector->GetThreshold();
+  double long_score = long_page->Score(derived) - long_page->GetThreshold();
+  bool distillable = score > 0;
+  bool long_article = long_score > 0;
   bool blacklisted = IsBlacklisted(parsed_url);
 
+  if (!features.isMobileFriendly) {
+    int score_int = std::round(score * 100);
+    if (score > 0) {
+      UMA_HISTOGRAM_COUNTS_1000("DomDistiller.DistillabilityScoreNMF.Positive",
+          score_int);
+    } else {
+      UMA_HISTOGRAM_COUNTS_1000("DomDistiller.DistillabilityScoreNMF.Negative",
+          -score_int);
+    }
+    if (distillable) {
+      // The long-article model is trained with pages that are
+      // non-mobile-friendly, and distillable (deemed by the first model), so
+      // only record on that type of pages.
+      int long_score_int = std::round(long_score * 100);
+      if (long_score > 0) {
+        UMA_HISTOGRAM_COUNTS_1000("DomDistiller.LongArticleScoreNMF.Positive",
+            long_score_int);
+      } else {
+        UMA_HISTOGRAM_COUNTS_1000("DomDistiller.LongArticleScoreNMF.Negative",
+            -long_score_int);
+      }
+    }
+  }
+
   int bucket = static_cast<unsigned>(features.isMobileFriendly) |
       (static_cast<unsigned>(distillable) << 1);
   if (is_last) {
     UMA_HISTOGRAM_ENUMERATION("DomDistiller.PageDistillableAfterLoading",
         bucket, 4);
   } else {
     UMA_HISTOGRAM_ENUMERATION("DomDistiller.PageDistillableAfterParsing",
         bucket, 4);
     if (!distillable) {
       UMA_HISTOGRAM_ENUMERATION("DomDistiller.DistillabilityRejection",
@@ skipping 68 matching lines @@
   render_frame()->GetServiceRegistry()->ConnectToRemoteService(
       mojo::GetProxy(&distillability_service));
   DCHECK(distillability_service);
   distillability_service->NotifyIsDistillable(
       IsDistillablePage(doc, is_last), is_last);
 }
 
 DistillabilityAgent::~DistillabilityAgent() {}
 
 }  // namespace dom_distiller