Content Research Engine

Trigger Words

research this, gather info, what are people saying, competitive analysis, market research, sentiment monitoring, tech evaluation, user feedback, reputation analysis, trending topics, industry news, content aggregation

Capabilities

Not simple "search and copy-paste" — an end-to-end research pipeline.

Research Pipeline (5 Steps)

Step 1: Multi-Source Search

Based on research topic, automatically search:

Community Forums — Tech discussions, experience sharing, help requests
Q&A Platforms — Professional answers, best-practice voting
Tech Blogs — In-depth articles, tutorials, case studies
News Sites — Industry updates, product launches, policy changes
Review Sites — Product comparisons, expert ratings, user reviews
Academic Sources — Papers, whitepapers, research reports

Search Strategy:

Topic keywords + synonyms + related terms searched simultaneously
Time range filtering (past week/month/quarter/year)
Language filtering (primary language + English for authoritative sources)
Source credibility weighting (official > verified expert > regular user > anonymous)

Step 2: Intelligent Collection

Process search results through:

Relevance Score — Match with research goals (0-100)
Freshness Score — Time-decay value of publication date
Authority Score — Credibility level of source
Quality Filter — Remove spam/ads/low-quality content
Deduplication — Aggregate same-content items (multi-source reports on same event)

Step 3: Deep Analysis

Multi-dimensional analysis of collected content:

Sentiment Analysis

Positive/Negative/Neutral classification
Sentiment intensity score (-1 to +1)
Sentiment keyword extraction
Sentiment trend over time

Opinion Distribution

Major opinion clusters (what different voices exist)
Support ratio for each viewpoint
Opposition/complementarity relationships between views
Opinion leader identification (whose views are most cited)

Trend Analysis

Discussion volume timeline
Key inflection points and trigger events
Early signals of emerging topics
Comparison with historical same-period data

Step 4: Structured Report

Output standardized research report:

# [Topic] Research Report

## Executive Summary
One-paragraph summary of key findings.

## 1. Data Overview
- Sources collected: N
- Valid content items: M
- Time span: X days
- Platforms covered: [list]

## 2. Key Findings
### Finding 1: [Title]
- Data support: ...
- Representative quotes: ...
- Confidence: High/Medium/Low

### Finding 2: [Title]
...

## 3. Opinion Map
- Viewpoint A (X%): ...
- Viewpoint B (Y%): ...
- Viewpoint C (Z%): ...

## 4. Sentiment Analysis
- Overall sentiment: [Positive/Negative/Neutral]
- Sentiment distribution data
- Key sentiment turning points

## 5. Trend Insights
- Recent trends: ...
- Early signals: ...

## 6. Recommendations
Specific action items based on findings.

## Appendix
- Source inventory
- Methodology notes
- Limitations disclaimer

Step 5: Continuous Monitoring (Optional)

For topics needing long-term tracking:

Set monitoring frequency (daily/weekly)
Change threshold alerts (volume spike / sentiment reversal)
Regular summary push reports

Usage Modes

Mode A: One-time Deep Research

"Research how XXX is received in the market" → Full 5-step pipeline → One-time report

Mode B: Quick Intelligence

"What's happening with XXX lately" → Steps 1-3 condensed → Key points within 15 minutes

Mode C: Continuous Monitoring

"Keep track of developments in XXX domain" → Set monitoring rules → Periodic push + anomaly alerts

Quality Assurance

All citations sourced and dated
Distinguish factual statements from interpretive opinions
Clearly mark data limitations
Never fabricate content (if nothing found, say so)

个人创作者计费通道上线后直接升级计费