VideoContentSearchCaptionEntityAnchorFeatures

AI Overview😉

  • The potential purpose of this module is to analyze video captions and identify entities mentioned in the video, such as people, places, or products. It extracts features from the captions, including the mention text, transcript offset, and confidence scores, to help determine the relevance and importance of each entity.
  • This module could impact search results by influencing the ranking of videos based on their caption content. For example, if a video has a high-confidence entity mention that matches the search query, it may be ranked higher in the search results. The module's features, such as duration coverage and entity connectedness, may also affect how videos are grouped or clustered in search results.
  • To be more favorable for this function, a website could:
    • Provide high-quality, accurate, and comprehensive video captions that include relevant entity mentions.
    • Optimize video captions for search queries by including targeted keywords and phrases.
    • Use structured data and schema markup to help search engines understand the content and context of the video captions.
    • Ensure that video captions are easily accessible and crawlable by search engines.

Interesting Module? Vote 👇

Voting helps other researchers find interesting modules.

Current Votes: 0

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

Features and debug info for individual caption entity video anchors.

Attributes

  • asrMentionText (type: String.t, default: nil) - If the description anchor has been recognized as an entity and that entity has also been mention in the ASR, this is the mention text from the ASR.
  • asrMentionTranscriptOffset (type: integer(), default: nil) - The start index of the ASR entity mention from the beginning of the ASR transcript. The index is included so that individual mentions that share the same mid and time can be distinguished which happens when an entity is mentioned multiple times in an ASR sentence. Each mid/asr_mention_transcript_offset will be unique.
  • asrSentence (type: String.t, default: nil) - The nearest ASR sentence.
  • asrStartTime (type: integer(), default: nil) - The begin time in ms of the ASR sentence.
  • bertScores (type: list(number()), default: nil) - The result of the BERT inference using the title, ASR sentence and entity mention text.
  • broadness (type: number(), default: nil) - A score to help determine how specific this entity is.
  • durationCoverage (type: number(), default: nil) - The percentage of the video covered by the span of the first mention to the last mention.
  • entityConnectedness (type: number(), default: nil) - The webref connectedness score of the entity.
  • entityDescription (type: String.t, default: nil) - A short text describing the entity.
  • entityInfo (type: GoogleApi.ContentWarehouse.V1.Model.VideoContentSearchCaptionEntityDocInfo.t, default: nil) - Information about how many documents the entity occurred in and how many total mentions the entity has across the entire video corpus.
  • entityMentionInDescription (type: boolean(), default: nil) - Whether or not the entity mention text appears in the description of the video.
  • estimatedMentionTime (type: integer(), default: nil) - The estimated begin time in ms of the entity mention using the text offset divided by the ASR duration.
  • groupCohesion (type: number(), default: nil) - Average similarity between this anchor and other anchors in the set.
  • hypernymConfidence (type: number(), default: nil) - The confidence of the hypernym used as the set label.
  • hypernymCount (type: number(), default: nil) - Number of hypernyms used for calculating similarity.
  • hyperpediaSalientTermsSimilarity (type: number(), default: nil) - The cosine similarity between the document salient terms and the hyperpedia hypernyms for a given entity.
  • inWebrefEntities (type: boolean(), default: nil) - Whether or not this entity is in the webref entities.
  • isOracleEntity (type: boolean(), default: nil) - True if the given entity appears as an Oracle followup query.
  • isProduct (type: boolean(), default: nil) - Whether this entity counts as a 'product' for the purpose of dividing entities between the 'related topics' and 'products in this video' features.
  • maxMentionConfidence (type: number(), default: nil) - The maximum confidence of all of the entity mentions in the transcript.
  • mentionConfidence (type: number(), default: nil) - The confidence that the ASR mention matches the given mid.
  • mentions (type: integer(), default: nil) - Number of times an entity is mentioned in the ASR transcript.
  • msFromLastAnchor (type: integer(), default: nil) - The duration in ms between this anchor and the previous anchor or the beginning of the video if this is the first anchor.
  • nextAsrSentence (type: String.t, default: nil) - The ASR sentence after asr_sentence or "" if it is the last sentence.
  • otherAsrMentionText (type: list(String.t), default: nil) - The ASR text of each mention of the entity.
  • otherEstimatedMentionTimes (type: list(integer()), default: nil) - Each time the entity was mentioned.
  • previousAsrSentence (type: String.t, default: nil) - The ASR sentence before asr_sentence or "" if it is the first sentence.
  • trustedNameConfidence (type: number(), default: nil) - The confidence that the ASR mention is a trusted name.
  • webrefEntityTopicality (type: number(), default: nil) - The webref entity topicality score if the entity is a webref entity and 0 if not.

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.VideoContentSearchCaptionEntityAnchorFeatures{
    asrMentionText: String.t() | nil,
    asrMentionTranscriptOffset: integer() | nil,
    asrSentence: String.t() | nil,
    asrStartTime: integer() | nil,
    bertScores: [number()] | nil,
    broadness: number() | nil,
    durationCoverage: number() | nil,
    entityConnectedness: number() | nil,
    entityDescription: String.t() | nil,
    entityInfo:
      GoogleApi.ContentWarehouse.V1.Model.VideoContentSearchCaptionEntityDocInfo.t()
      | nil,
    entityMentionInDescription: boolean() | nil,
    estimatedMentionTime: integer() | nil,
    groupCohesion: number() | nil,
    hypernymConfidence: number() | nil,
    hypernymCount: number() | nil,
    hyperpediaSalientTermsSimilarity: number() | nil,
    inWebrefEntities: boolean() | nil,
    isOracleEntity: boolean() | nil,
    isProduct: boolean() | nil,
    maxMentionConfidence: number() | nil,
    mentionConfidence: number() | nil,
    mentions: integer() | nil,
    msFromLastAnchor: integer() | nil,
    nextAsrSentence: String.t() | nil,
    otherAsrMentionText: [String.t()] | nil,
    otherEstimatedMentionTimes: [integer()] | nil,
    previousAsrSentence: String.t() | nil,
    trustedNameConfidence: number() | nil,
    webrefEntityTopicality: number() | nil
  }

Functions

Link to this function

decode(value, options)

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

Unwrap a decoded JSON object into its complex fields.