次の方法で共有


AssessmentSentiment クラス

AssessmentSentiment には、予測されたセンチメント、信頼度スコア、および特定のターゲットに関する評価に関するその他の情報が含まれます。 たとえば、"The food is good" という文では、ターゲットの 'food' の評価は 'good' です。

継承
azure.ai.textanalytics._dict_mixin.DictMixin
AssessmentSentiment

コンストラクター

AssessmentSentiment(**kwargs: Any)

メソッド

get
has_key
items
keys
update
values

get

get(key: str, default: Any | None = None) -> Any

パラメーター

名前 説明
key
必須
default
既定値: None

has_key

has_key(k: str) -> bool

パラメーター

名前 説明
k
必須

items

items() -> Iterable[Tuple[str, Any]]

keys

keys() -> Iterable[str]

update

update(*args: Any, **kwargs: Any) -> None

values

values() -> Iterable[Any]

属性

confidence_scores

"positive" ラベルと "negative" ラベルの評価に対する 0 から 1 のセンチメント信頼度スコア。 'neutral' のスコアは常に 0 になります

confidence_scores: SentimentConfidenceScores

is_negated

評価の値が否定されるかどうか。 たとえば、"The food is not good" では、評価 "good" は否定されます。

is_negated: bool

length

評価テキストの長さ。 この値は、元の要求で設定された string_index_type パラメーターの値 (既定では UnicodeCodePoints) によって異なります。

length: int

offset

ドキュメントの先頭からの評価テキスト オフセット。 値は、元の要求で設定された string_index_type パラメーターの値 (既定では UnicodeCodePoints) によって異なります。

offset: int

sentiment

評価の予測センチメント。 指定できる値は、"正"、"混合"、"負" です。

sentiment: str

text

評価テキスト。

text: str