Cryptoracle Data Analysis Team

Date: August 25 - August 31, 2025

Data source coverage

Anomaly Alerts

All channels scraped under the TG account are 197 more than the TG channels provided by the operator.

Core indicators

Total Records Ingested：117,604,610

Total Scraped Groups：3772（Operational Effective Groups 3772）

Effective Data Rate：100%

Newly Scraped Groups：131

New field

Inspection Date: August 29, 2025

Inspection Object: Data

Random Sampling (Sample Size: 300)

Main Issues:

Accuracy: 48% (100 items)

Omission Rate: 0.5% (200 items)

1. Anti-telecommunication fraud reminder messages, some of which were identified as advertisements while others were not, should not be considered advertisements;

2. Collections of news about tokens/projects, numbered, were identified as advertisements;

3. Pure links, image URLs, and platform account addresses were identified as advertisements;

4. Excerpts with content summaries and original links were identified as advertisements;

5. Explanatory and guidance texts were identified as advertisements.

Inspection Date: August 29, 2025

Inspection Object: Data-Dictionary

Random Sampling (Sample Size: 10)

Accuracy: 0%

Main Issues:

Accuracy: 48% (100 items)

AR：No exact match, for example: “pra começar não comprar von leon”;

ME: Match error，for example：“💗💗💗🅰💗💗💗💗 beyler Dıldo anal ,boşalmalı ve yepyeni kostümlerim ,arşiv için. @Minikmodel

sure its gonna be worth 😂anyway good night its time for me 😘”

August 29th Quality Inspection - Data Source Business Quality Inspection

Check 1: Specific Tag Logic

Community Example: Selling .btc domain

Check 2: Coin Type Matching for Common Words

Requirement 1: Coin Dictionary

Requirement 2: Common Word Dictionary

Existing Issue: Some coin types in the dictionary conflict with English high-frequency words.

1. The process for handling data invalidation issues is standardized;

2. The number of DC and TG signal sources continues to increase;

3. Increase the token dictionary.

1. Structured Deployment of Content Marketing: Identify the product's market positioning and target marketing to different types of customers;

2. Optimization of Ad Recognition and Dictionary Matching;

3. Periodic Screening of Low-Quality Communities;