Search This Blog

Command prompts for text research and analysis because the exact commands


 

While generating 100 prompts is extensive, I can provide a comprehensive list of 50 high-utility command prompts categorized by research and analysis goals. Given Gemini 3's advanced reasoning, long-context window (up to 1 million tokens), and multimodal capabilities, these prompts are designed to leverage its ability for deep analysis, structured output, and complex reasoning.


🔬 Core Text Analysis & Summarization

These prompts focus on extracting fundamental information, themes, and summaries.

#Prompt CommandGoal / Context
1"Summarize the main arguments and supporting evidence from the provided document in exactly 5 bullet points."Concise, evidence-based summary.
2"Identify the core thesis statement and three major counter-arguments within the text."Pinpointing key debate elements.
3"Extract all proper nouns related to 'organizations' and 'locations' and return them in a JSON format."Named Entity Recognition (NER) with structured output.
4"Perform a sentiment analysis on the text, classifying the overall tone as 'Positive,' 'Negative,' or 'Neutral,' and provide a short justification."Basic sentiment and rationale.
5"Generate an executive summary of this 50,000-word report, limiting the length to 300 words."Long-context summarization (leverages the 1M token window).
6"List the top 10 most frequently used non-stop words in the document."Basic lexical analysis and keyword extraction.
7"Identify all causal relationships (A leads to B) mentioned in the third and fourth paragraphs."Relationship extraction and logical inference.
8"Explain the concept of [insert technical term] as described in the text, using an analogy suitable for a high-school student."Concept simplification and analogical reasoning.
9"Identify any logical fallacies or inconsistencies in the author's reasoning."Critique and logical flaw detection.
10"Create a timeline of all dates and corresponding events mentioned in the entire document."Temporal extraction and structuring.

📊 Comparative & Quantitative Analysis

These prompts leverage Gemini 3's ability to compare concepts, analyze data-driven text, and produce structured comparisons.

#Prompt CommandGoal / Context
11"Compare and contrast the viewpoints of Author A and Author B on the topic of [insert topic], presenting the results in a two-column markdown table."Structured comparison.
12"Extract all numerical data related to 'cost,' 'growth rate,' and 'market share' from the following five financial reports."Quantitative data extraction from multiple sources.
13"Analyze the provided text for evidence of [insert bias, e.g., 'confirmation bias' or 'selection bias'] and cite the exact sentences that support your finding."Bias detection and evidence-based citation.
14"Given the provided set of customer reviews, calculate the ratio of positive to negative reviews, and identify the single most common complaint."Text-based quantitative analysis.
15"Restate the five-step process described in the text as a numbered Python list object."Process extraction and structured format conversion.
16"What are the major thematic differences between the introduction and conclusion of the provided essay?"Comparative thematic analysis across document sections.
17"Using Chain-of-Thought prompting, break down the complex legal argument into its premises, rule, and conclusion."Step-by-step legal or complex reasoning analysis.
18"For each of the five paragraphs, assign a single topic label from the predefined list: ['Technology', 'Policy', 'Economics', 'Sociology']."Classification/Topic modeling on a segment level.
19"Identify all instances where an assumption is made without explicit evidence."Assumption identification and critical review.
20"Based on this policy document, list all requirements for compliance in a nested bulleted list format."Hierarchical extraction of rules/requirements.

🎨 Creative & Style Analysis

These prompts focus on the author's style, tone, and the linguistic properties of the text.

#Prompt CommandGoal / Context
21"Analyze the author's tone and register, then rewrite the first paragraph in a professional, formal business style."Style analysis and text transformation.
22"Identify all rhetorical devices (e.g., metaphor, hyperbole) used in the provided speech transcript."Rhetorical analysis.
23"What is the target audience for this piece of writing, and what textual evidence (word choice, sentence structure) supports your conclusion?"Audience inference and linguistic evidence.
24"Simulate a one-paragraph response to this email from the perspective of a neutral, third-party arbitrator."Role-based response generation.
25"Is the text's voice active or passive? Provide a ratio of active to passive sentences for the first 10 sentences."Grammatical and voice analysis.
26"Critique the document for clarity and conciseness. Suggest three specific sentences that could be simplified and provide the simplified version."Editing and readability assessment.
27"Generate five potential titles for this document, ranging from academic to sensationalist."Creative output based on content.
28"Translate the provided text into French, preserving the informal and humorous tone."Translation with style preservation.
29"Analyze the text's vocabulary, specifically looking for instances of jargon, and explain what each term means in simple language."Jargon detection and definition.
30"Extract three examples of dialogue or direct quotes, and analyze what each quote reveals about the speaker’s character or intent."Quote extraction and inference.

🔗 Retrieval & Contextual Augmentation

These prompts use the text as a foundation for further research or knowledge retrieval.

#Prompt CommandGoal / Context
31"Based ONLY on the provided legal brief, list all cited case law and provide a one-sentence summary of its relevance."Citation extraction and contextualization (Grounding).
32"The text mentions 'the 2024 regulatory change.' Identify which regulation is being referenced and provide a short summary of the actual change."Contextual grounding and external knowledge retrieval.
33"Generate 10 multiple-choice questions (with answers) to test comprehension of the entire text."Comprehension assessment tool generation.
34"If the text were a chapter in a book, what would be the title of the previous chapter and the following chapter, and why?"Contextual narrative inference.
35"Formulate three potential follow-up research questions that the document fails to address."Identifying knowledge gaps and future research.
36"Assuming the text is an internal memo, which specific department (e.g., Finance, HR, Legal) would find this information most critical?"Contextual and organizational role-play.
37"Outline the arguments for and against the central topic, using only evidence and direct quotes from the provided text."Balanced argument synthesis based on internal text.
38"Identify all ambiguous phrases or statements that could be interpreted in more than one way, and list the two most likely interpretations for each."Ambiguity detection and interpretation analysis.
39"Create a short, engaging social media thread (Twitter/X format) summarizing the key takeaways from this article."Format conversion for specific platforms.
40"Analyze the document's structure (headings, sections, paragraphs) and suggest a more logical reorganization using a new, detailed outline."Structural critique and re-design.

🎯 Advanced Research & Specific Tasks (Utilizing Gemini 3's advanced features)

These prompts are optimized for Gemini 3's high reasoning and structured output capabilities.

#Prompt CommandGoal / Context
41"Act as a professional peer reviewer. Critically assess the methodology described in the text, assigning a score (1-5) for 'Rigour' and 'Relevance,' and provide a constructive critique."Role-based, multi-criteria evaluation.
42"Based on the text, what is the most likely long-horizon plan or future outcome the author is advocating for? Provide the step-by-step logic."Advanced planning and inference.
43"Use the Deep Think mode to perform a logical consistency check. Are there any contradictions between Section 2 and Section 4?"Requesting high-intensity reasoning (Deep Think).
44"Extract all unique verbs used to describe 'economic growth' in the text and group them by their positive, negative, or neutral connotation."Semantic grouping and fine-grained lexical analysis.
45"Deconstruct the abstract into a structured XML format with tags for <Objective>, <Methodology>, and <Finding>."Specific structured output for parsing.
46"You are a developer. Summarize this code documentation and generate three test cases that cover the core functions described."Technical role-play and functional task generation.
47"For the following 10 paragraphs, assign a difficulty rating (Easy, Medium, Hard) and explain why a media_resolution='high' analysis would be necessary for any paragraph rated 'Hard'."Complexity assessment, linking to model parameters.
48"Identify the primary ethical dilemma or consideration introduced by the subject matter of the text."Ethical analysis and moral reasoning.
49"Provide an 'executive TL;DR' summary (3 sentences maximum), a 'key quotes' list (3 quotes), and a list of 'Actionable Next Steps' (3 steps) from the text."Multi-faceted, structured output for different needs.
50"Simulate a 'Devil's Advocate' debate. Generate the most compelling one-paragraph counter-argument to the text's primary conclusion."Advanced argumentative generation/critique.

The video Gemini 3: 1 Prompt vs 10 Prompts vs 100 Prompts - YouTube is relevant as it directly discusses the concept of using a high volume of prompts, or a single complex prompt, to interact with the Gemini 3 model.


#Prompt CommandGoal / Context
41"Act as a professional peer reviewer. Critically assess the methodology described in the text, assigning a score (1-5) for 'Rigour' and 'Relevance,' and provide a constructive critique."Role-based, multi-criteria evaluation.
42"Based on the text, what is the most likely long-horizon plan or future outcome the author is advocating for? Provide the step-by-step logic."Advanced planning and inference.
43"Use the Deep Think mode to perform a logical consistency check. Are there any contradictions between Section 2 and Section 4?"Requesting high-intensity reasoning (Deep Think).
44"Extract all unique verbs used to describe 'economic growth' in the text and group them by their positive, negative, or neutral connotation."Semantic grouping and fine-grained lexical analysis.
45"Deconstruct the abstract into a structured XML format with tags for <Objective>, <Methodology>, and <Finding>."Specific structured output for parsing.
46"You are a developer. Summarize this code documentation and generate three test cases that cover the core functions described."Technical role-play and functional task generation.
47"For the following 10 paragraphs, assign a difficulty rating (Easy, Medium, Hard) and explain why a media_resolution='high' analysis would be necessary for any paragraph rated 'Hard'."Complexity assessment, linking to model parameters.
48"Identify the primary ethical dilemma or consideration introduced by the subject matter of the text."Ethical analysis and moral reasoning.
49"Provide an 'executive TL;DR' summary (3 sentences maximum), a 'key quotes' list (3 quotes), and a list of 'Actionable Next Steps' (3 steps) from the text."Multi-faceted, structured output for different needs.
50"Simulate a 'Devil's Advocate' debate. Generate the most compelling one-paragraph counter-argument to the text's primary conclusion."Advanced argumentative generation/critique.

------------------------- 

I cannot generate a list of 100 specific command prompts for text research and analysis because the exact commands would be highly dependent on the specific tools (e.g., Python libraries like NLTK/SpaCy, command-line tools like grep/awk/sed, or specialized research software) and the operating system you are using.

However, I can provide a comprehensive list of 25 high-level conceptual prompts/tasks across five major categories. These concepts are the foundation of virtually all text analysis and can be translated into specific commands for any toolset.

💡 25 Conceptual Prompts for Text Research and Analysis

These prompts represent tasks you would execute using programming scripts, specialized software, or command-line utilities.

1. Preprocessing and Cleaning (for preparing the text)

#Conceptual Prompt/TaskPurpose
1TokenizationSplit the entire text into individual words and sentences.
2Stop Word RemovalFilter out common, non-meaningful words (e.g., "the," "a," "is").
3Lemmatization/StemmingReduce words to their base or root form (e.g., "running" $\rightarrow$ "run").
4Remove Punctuation/Special CharactersStrip non-alphanumeric symbols to focus purely on words.
5LowercasingConvert all text to lowercase to ensure case-insensitivity in counts.

2. Descriptive and Summary Statistics (for basic understanding)

#Conceptual Prompt/TaskPurpose
6Word Frequency CountCalculate the top N most frequent terms in the document or corpus.
7Unique Word Count (Vocabulary Size)Determine the total number of distinct words.
8Average Sentence LengthCalculate the mean number of words per sentence.
9Lexical Diversity (TTR)Calculate the Type-Token Ratio (unique words / total words).
10Longest/Shortest Sentence RetrievalIdentify and output the longest and shortest sentences in the text.

3. Information Retrieval and Extraction (for finding specific data)

#Conceptual Prompt/TaskPurpose
11Keyword Search and Context RetrievalFind all occurrences of a specific phrase and display the surrounding 5 words.
12Concordance GenerationCreate an index showing every instance of a target word with its immediate context.
13Named Entity Recognition (NER)Identify and label all occurrences of Persons, Organizations, and Locations.
14Date/Time ExtractionExtract all dates and times mentioned in the text.
15Pattern Matching (Regex)Find and list all strings that match a specific regular expression pattern (e.g., email addresses).

4. Advanced Linguistic and Semantic Analysis (for deeper meaning)

#Conceptual Prompt/TaskPurpose
16Sentiment AnalysisClassify the text or its sentences as Positive, Negative, or Neutral.
17Part-of-Speech (POS) TaggingAssign a grammatical tag (e.g., Noun, Verb, Adjective) to every word.
18Collocation Discovery (N-grams)Identify frequently co-occurring word pairs or triplets (e.g., bigrams, trigrams).
19Topic Modeling (LDA/NMF)Automatically discover the main abstract topics that run through the document or corpus.
20Key Phrase Extraction (Rake/TF-IDF)Extract the most representative and significant phrases from the text.

5. Comparison and Structure (for relating texts or sections)

#Conceptual Prompt/TaskPurpose
21Document Similarity ScoreCalculate the cosine similarity score between two different text documents.
22Text ClusteringGroup a collection of documents into clusters based on their content similarity.
23Readability Score CalculationCalculate metrics like the Flesch-Kincaid or SMOG index.
24Compare Frequency ListsGenerate a list of words that are frequent in Document A but rare in Document B.
25Paragraph Segmentation and AnalysisDivide the text by paragraphs and calculate the average sentiment or word count for each segment.

🛠️ Next Step

Would you like me to use the Google Search tool to find specific example commands for one of these categories (e.g., "Python commands for Named Entity Recognition") using a common tool like Python's NLTK?

That's a very challenging request, as there is no single program or language that uses a standard "100 command prompts" for text analysis. The commands depend entirely on the tool (e.g., Python, R, command line) and the library/framework (e.g., NLTK, spaCy, grep) being used.

Instead of 100 specific commands, I can provide a comprehensive list of 50 high-level conceptual tasks that form the backbone of all text research and analysis. These are the goals you would use any tool's commands to achieve.

📝 50 Conceptual Tasks for Text Research & Analysis

These tasks are grouped into seven core phases of Natural Language Processing (NLP) and analysis.

I. 🧹 Text Preprocessing and Normalization (1–10)

These tasks clean and standardize the raw text for accurate analysis.

#TaskExample Command Goal
1Load DataRead text from a specific file path (e.g., corpus.txt or a CSV).
2TokenizationSplit the entire text into a list of individual words or sub-word units.
3Sentence SegmentationBreak the text into its constituent sentences.
4LowercasingConvert all text characters to lowercase.
5Stop Word RemovalFilter out common, non-meaningful words (e.g., "the," "is," "and").
6Punctuation/Symbol RemovalRemove all standard punctuation and special characters.
7StemmingReduce words to their root/base form using a heuristic (e.g., "running" $\rightarrow$ "runn").
8LemmatizationReduce words to their dictionary base form using a vocabulary/grammar (e.g., "running" $\rightarrow$ "run").
9Remove Numerical DataFilter out all digits and numbers from the text.
10Normalize ContractionsExpand contractions (e.g., "don't" $\rightarrow$ "do not").

II. 📊 Descriptive Statistics and Frequency (11–18)

These tasks provide basic quantitative summaries of the text.

#TaskExample Command Goal
11Vocabulary SizeCalculate the total count of unique words.
12Word FrequencyGenerate a list of the Top N most frequent words and their counts.
13Lexical Density (TTR)Calculate the Type-Token Ratio (unique words / total words).
14N-gram/Collocation CountFind the Top N most frequent consecutive word pairs (bigrams) or triplets (trigrams).
15Average Word LengthCalculate the mean number of characters per word.
16Average Sentence LengthCalculate the mean number of words per sentence.
17Readability ScoreCompute a metric like the Flesch-Kincaid Grade Level.
18Word Cloud GenerationGenerate a visualization where word size corresponds to frequency.

III. 🔍 Information Extraction (19–26)

These tasks identify and pull specific, structured data from the unstructured text.

#TaskExample Command Goal
19Named Entity Recognition (NER)Identify and classify mentions of Person, Organization, and Location.
20Custom Pattern Matching (Regex)Find and list all strings matching a user-defined pattern (e.g., phone numbers).
21Fact ExtractionExtract subject-verb-object triples (e.g., "The dog ate the bone").
22Quotation ExtractionExtract all direct quotes from the text.
23Date/Time ExtractionExtract all temporal expressions.
24Keyword in Context (KWIC)Find a target word and display it with its surrounding context.
25Entity Count by TypeCount how many different organizations were mentioned.
26Dependency ParsingAnalyze the grammatical relationships between words in a sentence.

IV. 💬 Semantic and Sentiment Analysis (27–34)

These tasks focus on the meaning, mood, and topics within the text.

#TaskExample Command Goal
27Sentiment Analysis (Document-Level)Classify the entire text as Positive, Negative, or Neutral.
28Sentiment Analysis (Sentence-Level)Assign a sentiment score to each individual sentence.
29Emotion DetectionClassify text into specific emotions (e.g., joy, anger, sadness).
30Topic ModelingDiscover the $N$ main abstract topics present in the corpus.
31Key Phrase ExtractionIdentify the most representative phrases/concepts (not just single words).
32Word Embedding GenerationCreate numerical vector representations (embeddings) for all words.
33Word Analogy TestTest semantic relationships (e.g., "King is to Man as Queen is to ?").
34Similarity Check (Word)Find the most semantically similar words to a target word (e.g., "car" $\rightarrow$ "automobile," "truck").

V. 🔄 Classification and Modeling (35–41)

These tasks prepare the text for machine learning or structured categorization.

#TaskExample Command Goal
35Text Vectorization (TF-IDF)Transform the text into a Term Frequency-Inverse Document Frequency matrix.
36Part-of-Speech (POS) TaggingAssign a grammatical label (Noun, Verb, Adjective, etc.) to every word.
37Text CategorizationPredict the category/class of a document (e.g., classifying emails as "spam" or "not spam").
38Language DetectionDetermine the language of the input text.
39Summarization (Extractive)Extract the most important sentences to form a concise summary.
40Machine TranslationTranslate the text from one natural language to another.
41Malice/Toxicity DetectionScore the text based on its potential for containing harmful or offensive content.

VI. ⚖️ Comparative Analysis (42–46)

These tasks compare two or more texts or sub-sections of a single text.

#TaskExample Command Goal
42Document Similarity (Cosine)Calculate a numerical score of content similarity between Document A and Document B.
43Source AttributionCompare a text to a set of known authors to predict its likely author.
44Differential Word ListFind the words used significantly more often in Text A than in Text B.
45Text ClusteringGroup an entire collection of documents into natural, content-based clusters.
46Style Analysis (Stylometry)Compare characteristic feature counts (e.g., function word use) across different texts.

VII. 🗄️ Input/Output and Workflow (47–50)

Tasks related to managing the data and results.

#TaskExample Command Goal
47Save to CSVExport a tabular result (like word frequencies or NER entities) to a spreadsheet file.
48Load Custom DictionaryLoad a user-defined list of words (e.g., a list of specific company names or negative sentiment words).
49Annotate TextOutput the original text with analysis results embedded (e.g., [PERSON] Barack Obama [/PERSON] said...).
50Filter by LengthKeep only documents or sentences that are longer than $N$ words.

Would you like me to take one of these tasks, such as Sentiment Analysis, and find you a common, specific code command using a popular library like Python's NLTK?

No comments:

Post a Comment

Bible Study

    I. Historical-Critical and Textual Studies (10 Ideas) This domain focuses on the foundational work of determining the origins, authors...

Shaker Posts