Specification for the use of blog content structure and classification labels (V1.0)
This article records the actual use of the current blog in the content structure (category/tags/series), and based on the current situation, we have sorted out a set of content management specifications that can be executed for a long time, which are used for subsequent article releases and structural optimization.
The current state of content structure
At present, the blog has formed a large-scale content system:
- Chinese Articles: 1188
- English articles: 1178
Tags and classifications are as follows:
- Chinese label: 8617
- English tags: 8617
- Category Quantity: 740 (Chinese and English consistent structure)
- Number of series: 11 (Chinese and English are the same)
1. Current structural features
The overall structure adopts a three-layer system:
- Category: Multi-level structure (up to 5 layers)
- Tags: High Density Keyword System
- Series (Series): A small number of theme aggregations
2. Current usage
In the actual writing process, there are the following usage habits:
- The classification system adopts multi-level path structure
- One article may choose multiple categories (up to about 4 to 5 levels)
- The number of labels is large (about 5~15/article)
- The Chinese and English label system is completely copied, only language switching
- The labeling system is not semantically localized in the English site
2. The current core problems
With the expansion of content, the current structural problems have gradually emerged, mainly focusing on the following aspects:
1. The level of the classification system is too deep, resulting in vague attribution
The current classification structure can reach up to 5 layers, for example:
Web Application Development → Web Framework → Laravel → Laravel Theme
In actual use, an article often selects multiple hierarchical classifications at the same time.
Problem performance:
- Category is more like ‘path selection’ than ‘attribution judgment’
- An article corresponds to multiple categories, resulting in structural repetition
- SEO weights are scattered to multiple hierarchical pages
2. The labeling system is too large and semantics are mixed
The current number of tags has reached 8617, and there are obvious problems:
- Label size is close to ‘keyword database’, not an auxiliary system
- Label semantic hierarchy confusion (technical/category/scene mix)
- Too many labels, reduce efficiency
- English tags are not semantically localized
3. The Chinese and English labeling system is not localized
The current English label is basically copied directly from the Chinese label, and only the language is replaced.
causing the problem:
- English labels are not in line with natural search habits
- SEO keywords do not match local user expression
- Label page quality decline
- Search engines find content structures unnatural
4. The content attribution selection cost is too high
When publishing articles, the following situations often occur:
- Classification is difficult to determine (multiple levels are ‘reasonable’)
- The number of labels is difficult to control
- Category and label boundary blur
5. Structural signal overload (SEO level problem)
There is obvious signal redundancy in the current structure:
- Category Repeat Nesting
- Labels overlap in large numbers
- An article undertakes too much structural attribution
Result:
- SEO weight dispersion
- The page theme is not focused enough
- Search engines have difficulty identifying core topics
The principle of structural adjustment (core design idea)
Based on the above questions, the content structure is redefined:
1. Category: Only responsible for ‘unique attribution’
Only one question is answered:
Which technical field is this article most specific to?
- Only 1 category can be selected for each article
- Only select the lowest-level classification
- No more selection of superiors
2. Tag: only responsible for ‘keyword index’
Labels are no longer used to express structures, but for search optimization:
- technical keywords
- Product / Framework
- Scenario / Question Type
3. Series: Responsible for ‘Theme Aggregation’
The series is used to build the main line of content:
- Cross-article topic concatenation
- Long-term content output
- SEO topic weight concentration
4. Content release specifications (after adjustment)
1. Classification and usage rules
- Only 1 category is allowed per article
- Only select the lowest level category
- Select multiple hierarchical classifications
2. Label usage rules
- Each article: 3 to 8 labels (up to 10)
- Labels must be keywords, not structural hierarchies
- Prohibit the use of type labels
3. English Labeling Rules (New)
The English label must meet:
- Search for expressions in natural English
- No Chinese literal translation
- Meet Google User Search Habits
Example:
| Chinese Tags | Wrong English | Recommended English |
|---|---|---|
| nginx log | nginx-log | nginx logging |
| performance optimization | performance-optimization | performance tuning |
| high concurrency | high-concurrency | High Concurrency |
5. Article Publishing Standard Process (SOP)
Each article is executed in the following order:
Step 1: Determine whether it belongs to the series
- There are corresponding series → priority to be included
- None → Consider creating a new one (prudential extension)
Step 2: Select Unique Category
- Only choose the lowest-level classification
- Do not choose superior classification
- Only 1 category is allowed
Step 3: Select Labels (Keywords)
Selection basis:
- What will users search for?
- What is the technical core?
- What is the scene problem?
Final output:
- 3 to 8 labels
- no more than 10
Step 4: English Label Check
- Is it natural to express
- Is it in line with search habits
- whether to avoid literal translation
6. Long-term optimization direction
Future Structural Optimization Goals:
- The classification system remains stable (no expansion)
- Gradually streamline and semanticization of tag systems
- The series system becomes the core content structure
- English-language site to establish an independent SEO semantic system
7. Summary
Summarize the current structural adjustment direction in one sentence:
From ‘multi-layer classification + high-density label’, to ‘single classification + keyword label + series dominant’.