VideoContentSearchShoppingOpinionsAnchorFeatures

AI Overview😉

  • The potential purpose of this module is to analyze video content and extract relevant information about products mentioned in the video, such as opinions, sentiments, and aspects. It appears to be designed for shopping-related videos and aims to provide a more detailed understanding of the video content.
  • This module could impact search results by providing more accurate and relevant information about video content, allowing users to better understand the content of a video before watching it. This could lead to more targeted and relevant search results, improving the overall user experience.
  • To be more favorable for this function, a website could ensure that its video content is accurately transcribed and annotated with relevant information such as product names, aspects, and sentiments. Additionally, the website could provide high-quality video content that is easy to analyze and understand, and ensure that the video metadata is accurate and complete. This could involve using techniques such as automatic speech recognition, natural language processing, and machine learning to analyze and extract relevant information from the video content.

Interesting Module? Vote 👇

Voting helps other researchers find interesting modules.

Current Votes: 3

GoogleApi.ContentWarehouse.V1.Model.VideoContentSearchShoppingOpinionsAnchorFeatures (google_api_content_warehouse v0.4.0)

Features for an individual Shopping Opinions Anchor. This file is used for video anchor use case. Next Id: 34

Attributes

  • asrForProConExtraction (type: String.t, default: nil) - The asr with sentence break that was used for pro/con extraction.
  • snippetWordCount (type: integer(), default: nil) - The number of words in the ASR for the selected segment window.
  • snippetQaScore (type: number(), default: nil) - The QA model score for the selected segment window against the question.
  • anchorLabelSentiment (type: number(), default: nil) - The sentiment score of the anchor label, with range: [-1, 1]. If using Lumin Pro/Con tags as the anchor labels, the "Pro" Lumin tag will have a score of 1 and Con Lumin tag will have a score of -1.
  • productNameFromTitle (type: String.t, default: nil) - The product name from title extracted by the grampus model.
  • snippetSubSegmentQaScore (type: number(), default: nil) - The QA model score for the best sub segment against the question.
  • anchorLabelFirstMentionPos (type: integer(), default: nil) - The first anchor_label mention position (word index, 0-based). It is computed from the snippet_sub_segment if exists. Otherwise it is computed from the snippet. It is not populated if there is no such mention.
  • babelSimilarityScore (type: number(), default: nil) - The Babel similarity score between the snippet and the anchor label.
  • conScore (type: number(), default: nil) - The classification score of the anchor being a con opinion.
  • isPro (type: boolean(), default: nil) - Whether the anchor is classified as a pro opinion.
  • luminAspectWordsMentions (type: integer(), default: nil) - The number of times words in Lumin aspect (that is not a stopword) being mentioned in the snippet.
  • mumScore (type: number(), default: nil) - The score from the MUM model if the pro/con anchor is extracted by MUM.
  • grampusScore (type: number(), default: nil) - The score from the Grampus model if the pro/con is extracted by Grampus.
  • snippetSubSegment (type: String.t, default: nil) - The ASR for the best matched sub segment inside the selected segment.
  • snippetSentimentScore (type: number(), default: nil) - The go/scarlett sentiment score of the selected segment window. Positive score represents positive sentiment. Negative score represents negative sentiment.
  • luminAspect (type: String.t, default: nil) - The lumin aspect of the Pro/Con Lumin tag. e.g. "weight".
  • question (type: String.t, default: nil) - The question used to score this video segment.
  • snippetSubSegmentSentimentScore (type: number(), default: nil) - The go/scarlett sentiment score of the best matched sub segment. Positive score represents positive sentiment. Negative score represents negative sentiment.
  • opinionsDolphinUsefulnessScore (type: number(), default: nil) -
  • snippetSubSegmentWordCount (type: integer(), default: nil) - The number of words in the ASR for the best matched sub segment.
  • mumProductAspect (type: String.t, default: nil) - The product aspect of the pro/con generated using the MUM model.
  • luminScore (type: number(), default: nil) - The Lumin model score for the anchor label against the segment.
  • isCon (type: boolean(), default: nil) - Whether the anchor is classified as a con opinion.
  • anchorOrAspectWordsMentions (type: integer(), default: nil) - The number of times words in anchor label or Lumin aspect (that is not a stopword) being mentioned in the snippet. If a word exists in both anchor label and Lumin aspect, it shall be only counted once for a mention in the snippet.
  • aspect (type: GoogleApi.ContentWarehouse.V1.Model.VideoContentSearchAspect.t, default: nil) - Product aspect being discussed by this Shopping Opinions.
  • proScore (type: number(), default: nil) - The classification score of the anchor being a pro opinion.
  • snippet (type: String.t, default: nil) - The ASR for the selected segment window.
  • anchorOrAspectFirstMentionPos (type: integer(), default: nil) - The smaller number of anchor_label_first_mention_pos and lumin_aspect_first_mention_pos.
  • isProConWhenExtractedFromMum (type: boolean(), default: nil) - True if the anchor is considered as pro or con when extracted from MUM.
  • luminAspectFirstMentionPos (type: integer(), default: nil) - The first Lumin aspect mention position (word index, 0-based). It is computed from the snippet_sub_segment if exists. Otherwise it is computed from the snippet. It is not populated if there is no such mention.
  • opinionsDolphinDescriptivenessScore (type: number(), default: nil) - Scores from Opinions Dolphin scorer. Opinions Dolphin scorer is built by finetuning the Dolphin-based Video Anchor Unified Scorer V2 on the Opinions anchors ratings. It outputs two scores, which are optimized for descriptiveness and usefulness ratings respectively. 'descriptiveness_score' measures how well the anchor label describes the video section. 'usefulness_score' measures how useful the anchor label is for jumping to an important section in the video. go/vs-opinions-migration-report
  • anchorLabel (type: String.t, default: nil) - The anchor label.
  • anchorLabelWordsMentions (type: integer(), default: nil) - The number of times words in anchor label (that is not a stopword) being mentioned in the snippet.

Summary

Types

t()

Functions

decode(value, options)

Unwrap a decoded JSON object into its complex fields.

Types

Link to this type

t()

@type t() ::
  %GoogleApi.ContentWarehouse.V1.Model.VideoContentSearchShoppingOpinionsAnchorFeatures{
    anchorLabel: String.t() | nil,
    anchorLabelFirstMentionPos: integer() | nil,
    anchorLabelSentiment: number() | nil,
    anchorLabelWordsMentions: integer() | nil,
    anchorOrAspectFirstMentionPos: integer() | nil,
    anchorOrAspectWordsMentions: integer() | nil,
    aspect:
      GoogleApi.ContentWarehouse.V1.Model.VideoContentSearchAspect.t() | nil,
    asrForProConExtraction: String.t() | nil,
    babelSimilarityScore: number() | nil,
    conScore: number() | nil,
    grampusScore: number() | nil,
    isCon: boolean() | nil,
    isPro: boolean() | nil,
    isProConWhenExtractedFromMum: boolean() | nil,
    luminAspect: String.t() | nil,
    luminAspectFirstMentionPos: integer() | nil,
    luminAspectWordsMentions: integer() | nil,
    luminScore: number() | nil,
    mumProductAspect: String.t() | nil,
    mumScore: number() | nil,
    opinionsDolphinDescriptivenessScore: number() | nil,
    opinionsDolphinUsefulnessScore: number() | nil,
    proScore: number() | nil,
    productNameFromTitle: String.t() | nil,
    question: String.t() | nil,
    snippet: String.t() | nil,
    snippetQaScore: number() | nil,
    snippetSentimentScore: number() | nil,
    snippetSubSegment: String.t() | nil,
    snippetSubSegmentQaScore: number() | nil,
    snippetSubSegmentSentimentScore: number() | nil,
    snippetSubSegmentWordCount: integer() | nil,
    snippetWordCount: integer() | nil
  }

Functions

Link to this function

decode(value, options)

@spec decode(struct(), keyword()) :: struct()

Unwrap a decoded JSON object into its complex fields.