prebuilt.journalism_.models¶

Models for the Journalism AI Assistant.

This module defines all Pydantic models used by the journalism assistant for structured outputs, validation, and data handling throughout the fact-checking, analysis, and review workflows.

Examples

>>> from journalism_assistant.models import FactCheckStatement, ArticleSummary
>>> fact_check = FactCheckStatement(
...     statement="The economy grew by 3%",
...     status="confirmed",
...     explanation="Verified by official statistics"
... )

Note

All models use Pydantic v2 with comprehensive field descriptions for documentation and validation.

Classes¶

`ArticleChunk`	Chunk of article text for processing.
`ArticleSummary`	Structured summary of an article.
`BiasIndicator`	Potential bias indicator in the text.
`ComprehensiveReport`	Complete journalism analysis report.
`ExtractedQuote`	Individual quote extracted from an article.
`FactCheckResult`	Complete fact-checking results for an article.
`FactCheckStatement`	Individual fact-check result for a single statement.
`GrammarBiasReview`	Complete grammar and bias review results.
`GrammarIssue`	Individual grammar or style issue.
`JournalismAction`	User action request for journalism tasks.
`QuoteExtractionResult`	Complete quote extraction results.
`SearchResult`	Web search result for fact-checking.
`ToneAnalysis`	Analysis of article tone and sentiment.

Module Contents¶

class prebuilt.journalism_.models.ArticleChunk(/, **data: Any)¶

Bases: pydantic.BaseModel

Chunk of article text for processing.

Represents a portion of the article for processing within token limits.

chunk_id¶: Unique identifier for the chunk

text¶: The actual text content

start_position¶: Starting character position

end_position¶: Ending character position

word_count¶: Number of words in chunk

validate_positions() → ArticleChunk¶: Ensure end position is after start position.

chunk_id: int = None¶

end_position: int = None¶

start_position: int = None¶

text: str = None¶

word_count: int = None¶

class prebuilt.journalism_.models.ArticleSummary(/, **data: Any)¶

Bases: pydantic.BaseModel

Structured summary of an article.

Provides a comprehensive summary focusing on key events, people, and statistics.

main_points¶: List of main points from the article

key_people¶: Important people mentioned

key_statistics¶: Important numbers and data

word_count¶: Original article word count

summary_text¶: Complete summary in paragraph form

key_people: List[str] = None¶

key_statistics: List[str] = None¶

main_points: List[str] = None¶

summary_text: str = None¶

word_count: int = None¶

class prebuilt.journalism_.models.BiasIndicator(/, **data: Any)¶

Bases: pydantic.BaseModel

Potential bias indicator in the text.

Represents language or framing that may indicate bias.

bias_type¶: Type of bias detected

example_text¶: Text showing the bias

explanation¶: Why this indicates bias

severity¶: How severe the bias is

suggestion¶: How to make it more neutral

bias_type: Literal['political', 'cultural', 'gender', 'racial', 'economic', 'confirmation', 'selection', 'framing'] = None¶

example_text: str = None¶

explanation: str = None¶

severity: Literal['subtle', 'moderate', 'strong'] = None¶

suggestion: str | None = None¶

class prebuilt.journalism_.models.ComprehensiveReport(/, **data: Any)¶

Bases: pydantic.BaseModel

Complete journalism analysis report.

Combines all analysis results into a comprehensive report.

article_title¶: Title of the analyzed article

analysis_timestamp¶: When the analysis was performed

summary¶: Article summary

fact_check_results¶: Fact-checking results

tone_analysis¶: Tone and sentiment analysis

quotes¶: Extracted quotes

grammar_bias_review¶: Grammar and bias review

overall_assessment¶: Overall article assessment

recommendations¶: Recommendations for the article

to_markdown() → str¶

Convert report to markdown format.

Returns:: Formatted markdown report
Return type:: str

analysis_timestamp: datetime.datetime = None¶

article_title: str | None = None¶

fact_check_results: FactCheckResult | None = None¶

grammar_bias_review: GrammarBiasReview | None = None¶

overall_assessment: str = None¶

quotes: QuoteExtractionResult | None = None¶

recommendations: List[str] = None¶

summary: ArticleSummary | None = None¶

tone_analysis: ToneAnalysis | None = None¶

class prebuilt.journalism_.models.ExtractedQuote(/, **data: Any)¶

Bases: pydantic.BaseModel

Individual quote extracted from an article.

Represents a direct quote with attribution and context.

quote_text¶: The exact quoted text

speaker¶: Person who said the quote

context¶: Context or situation of the quote

position¶: Position/title of the speaker if mentioned

significance¶: Why this quote is important

context: str = None¶

position: str | None = None¶

quote_text: str = None¶

significance: str | None = None¶

speaker: str = None¶

class prebuilt.journalism_.models.FactCheckResult(/, **data: Any)¶

Bases: pydantic.BaseModel

Complete fact-checking results for an article.

Contains all fact-check statements and overall statistics about the verification process.

statements¶: List of individual fact-check results

total_claims¶: Total number of claims analyzed

confirmed_count¶: Number of confirmed claims

refuted_count¶: Number of refuted claims

unverifiable_count¶: Number of unverifiable claims

vague_count¶: Number of vague claims

overall_credibility¶: Overall credibility score

calculate_statistics() → FactCheckResult¶: Calculate statistics from statements.

confirmed_count: int = None¶

overall_credibility: confloat(ge=0.0, le=1.0) = None¶

refuted_count: int = None¶

statements: List[FactCheckStatement] = None¶

total_claims: int = None¶

unverifiable_count: int = None¶

vague_count: int = None¶

class prebuilt.journalism_.models.FactCheckStatement(/, **data: Any)¶

Bases: pydantic.BaseModel

Individual fact-check result for a single statement.

Represents the verification status of a claim or statement extracted from an article.

statement¶: The original claim being fact-checked

status¶: Verification status (confirmed/refuted/unverifiable/vague)

explanation¶: Detailed explanation of the verification result

suggested_keywords¶: Keywords for further research if needed

confidence¶: Confidence score in the fact-check result

sources¶: Supporting sources used for verification

classmethod validate_keywords(v: List[str]) → List[str]¶: Ensure keywords are unique and non-empty.

confidence: confloat(ge=0.0, le=1.0) = None¶

explanation: str = None¶

sources: List[Dict[str, str]] = None¶

statement: str = None¶

status: Literal['confirmed', 'refuted', 'unverifiable', 'vague'] = None¶

suggested_keywords: List[str] = None¶

class prebuilt.journalism_.models.GrammarBiasReview(/, **data: Any)¶

Bases: pydantic.BaseModel

Complete grammar and bias review results.

Comprehensive review covering grammar issues and potential biases.

grammar_issues¶: List of grammar/style issues found

bias_indicators¶: List of potential biases detected

overall_quality_score¶: Overall writing quality score

readability_score¶: Readability score

bias_score¶: Overall bias score

recommendations¶: General recommendations for improvement

bias_indicators: List[BiasIndicator] = None¶

bias_score: confloat(ge=0.0, le=1.0) = None¶

grammar_issues: List[GrammarIssue] = None¶

overall_quality_score: confloat(ge=0.0, le=1.0) = None¶

readability_score: confloat(ge=0.0, le=1.0) = None¶

recommendations: List[str] = None¶

class prebuilt.journalism_.models.GrammarIssue(/, **data: Any)¶

Bases: pydantic.BaseModel

Individual grammar or style issue.

Represents a specific grammar, spelling, or style problem found in the text.

issue_type¶: Type of issue found

text¶: The problematic text

suggestion¶: Suggested correction

severity¶: Severity level of the issue

location¶: Approximate location in the text

issue_type: Literal['grammar', 'spelling', 'punctuation', 'style', 'clarity'] = None¶

location: str | None = None¶

severity: Literal['minor', 'moderate', 'major'] = None¶

suggestion: str = None¶

text: str = None¶

class prebuilt.journalism_.models.JournalismAction(/, **data: Any)¶

Bases: pydantic.BaseModel

User action request for journalism tasks.

Represents what actions the user wants to perform on the article.

actions¶: List of requested actions

priority¶: Priority level for the request

actions: List[Literal['summarization', 'fact-checking', 'tone-analysis', 'quote-extraction', 'grammar-and-bias-review', 'full-report', 'no-action-required', 'invalid']] = None¶

priority: Literal['low', 'medium', 'high'] = None¶

class prebuilt.journalism_.models.QuoteExtractionResult(/, **data: Any)¶

Bases: pydantic.BaseModel

Complete quote extraction results.

Contains all extracted quotes with metadata about the extraction process.

quotes¶: List of extracted quotes

total_quotes¶: Total number of quotes found

unique_speakers¶: Number of unique speakers

has_attribution¶: Whether all quotes have proper attribution

calculate_quote_stats() → QuoteExtractionResult¶: Calculate statistics from quotes.

has_attribution: bool = None¶

quotes: List[ExtractedQuote] = None¶

total_quotes: int = None¶

unique_speakers: int = None¶

class prebuilt.journalism_.models.SearchResult(/, **data: Any)¶

Bases: pydantic.BaseModel

Web search result for fact-checking.

Represents a single search result used for verification.

title¶: Title of the search result

url¶: URL of the source

snippet¶: Brief excerpt from the page

relevance_score¶: How relevant this result is

credibility_score¶: Source credibility score

credibility_score: confloat(ge=0.0, le=1.0) | None = None¶

relevance_score: confloat(ge=0.0, le=1.0) = None¶

snippet: str = None¶

title: str = None¶

url: str = None¶

class prebuilt.journalism_.models.ToneAnalysis(/, **data: Any)¶

Bases: pydantic.BaseModel

Analysis of article tone and sentiment.

Evaluates the overall tone, sentiment, and potential biases present in the article.

overall_tone¶: Primary tone of the article

sentiment_score¶: Sentiment score (-1 to 1)

tone_examples¶: Specific examples supporting the analysis

detected_biases¶: List of potential biases found

objectivity_score¶: Score for journalistic objectivity

detected_biases: List[str] = None¶

objectivity_score: confloat(ge=0.0, le=1.0) = None¶

overall_tone: Literal['neutral', 'positive', 'negative', 'critical', 'opinionated', 'balanced'] = None¶

sentiment_score: confloat(ge=-1.0, le=1.0) = None¶

tone_examples: List[Dict[str, str]] = None¶