Can one analyst see the opinions of hundreds of stocks?
Research organizations typically cover dozens to hundreds of stocks. With dozens of news articles pouring in for each stock every day, along with public disclosures, securities communities, and social media mentions, it is clear that there are limitations to what a person can visually process. Eventually, they can only keep up with a few stocks, and the rest are only confirmed after negative events have already affected stock prices.
The issue is not just speed. When a person tries to judge when and how quickly an issue is spreading based on intuition, they may miss important signals. To transform research and risk detection into data, the first step is to gather scattered mentions into one place.
Gather scattered mentions by stock
Hashscraper regularly collects documents mentioning specific companies, stocks, and sectors from various sources such as news portals, public disclosures, securities communities, and social media. It collects based on stock names and related keywords, then organizes the source, publication time, content, and link into structured fields for delivery. Instead of individuals browsing through each media outlet, they receive sorted mention data by stock.
The collection criteria are adjusted according to research purposes. For communities, reaction metrics like views, comments, and likes are included, while for disclosures, report types and submitters are included for future analysis.
Example of actual crawling data
{
"collected_at": "2026-07-30T09:15:00+09:00",
"source_type": "news",
"source_name": "네이버 증권 뉴스",
"ticker_keyword": "○○전자",
"sector": "반도체",
"title": "○○전자, 신규 팹 증설 계획 발표",
"published_at": "2026-07-30T08:40:00+09:00",
"author": "△△경제",
"body_snippet": "○○전자가 하반기 설비 투자를 확대한다고...",
"url": "https://example.com/news/12345",
"engagement": {
"view_count": 4210,
"comment_count": 37
}
}
Utilization scenario of a domestic asset management company
A domestic asset management company wanted to reduce instances of late recognition of negative events due to the inability of individuals to review all news and community opinions on covered stocks. Therefore, they established a flow where they regularly collect stock mentions, allowing staff to view mention volume and trend reactions on one screen and only investigate when specific keywords spike.
Crawling settings
- Collection targets: News portals, public disclosures, securities communities, social media mentions
- Collection criteria: Covered stock names, tickers, sector keywords
- Collection frequency: Regular collection four times a day (including before and after market hours)
- Collection period: Ongoing operation, storing data accumulated over the past 12 months
Analyzable items from crawling data
| Analysis Item | Utilization Method |
|---|---|
| Daily trend of stock mentions | Prioritize reviewing stocks with unusual mention spikes |
| Detection of increased negative event keywords | Early detection when specific keywords like recalls, lawsuits, regulations surge |
| Changes in sentiment trend (positive/negative events) | Identify turning points when sentiment shifts from positive to negative |
| Integration of supply chain and competitor issues | Check the impact of supplier and competitor issues outside of covered stocks |
| Distribution of mentions by source | Determine which source, news or community, first circulates issues |
Further organize mentions with AI analysis
By adding AI analysis to collected texts, each mention can be classified as positive or negative events and tagged with event types for summarization. As analysis columns are added to the original text data, staff can narrow down review targets through sorting and filtering.
AI analysis example
{
"url": "https://example.com/news/12345",
"sentiment": "positive",
"event_type": "capex_expansion",
"keywords": ["설비투자", "증설", "반도체"],
"summary": "하반기 설비 투자 확대로 생산능력 증가 기대"
}
Other scenarios for utilization
- IR departments: Regularly monitor how their company and competitors are mentioned in media and communities
- Sector research: Observe the rise and fall trends of themes based on mentions of specific industry keywords
- Risk management: Early detection of events by continuously monitoring keywords related to regulations, lawsuits, and disclosures
Summary
Instead of relying on human intuition to track opinions and issues of hundreds of stocks, collecting news, disclosures, and community mentions by stock allows for research and risk detection to be handled as data rather than feelings. By regularly examining mention volume, sentiment trends, and keyword spikes, signals that are easily missed can be confirmed at an earlier stage.
The collection itself is from a data provision perspective for research, not investment advice, and considerations regarding personal information and copyrights are made within the scope of public data and utilization purposes. As Hashscraper is responsible for operating, maintaining, and monitoring the crawler, organizations can focus on interpreting the received data.
Recommended Readings
- From crawling data to decision-making
- Outsourcing crawling process — Steps from inquiry to initial data, broken down by stages
- Crawling cost estimates, why they vary by vendor — Cost structure and hidden fees
Get Started Now
We can design together how to collect and analyze news, disclosures, and community mentions of covered stocks.
