| Comparison Point | Structured Data | Semi-Structured Data | Unstructured Data |
|---|---|---|---|
| Data Model | Strict data model, all the data has the same structure | The data has some kind off structure, Ex: JSON The data contains metadata and tags which helps it to be grouped appropriately | Does not conform to any Data model, no identifiable structure, Ex: Videos and word documents, reports, surveys |
| Data Analysis and Visualization | Easy and Fast to do data analysis, but what we CAN do is limited to the available relations - Due to the elements are able to be addressed, this enables efficient processign and analysis | Requires an intermediate stage where data processing happens, this gives more possibilities but is a bit slower than structured data | Requires an advanced data processing stage, like NLP, but this gives us wide data anaysis possibilites. |
| Joining and Query Model | Very efficient to do complex joins because we have a strict data model | Difficult to join data, querying is done with a search mechanism | Joining and relations are next to impossible, very limited to ex: full text search |
| Scaling and Availablity | Usage of clustering/mirroring when distributing among servers and RAID technologies when distributing on local storage - Availbility is done by mirroring on DB servers and RAID redundant storage | Uses Distributed files system that allows sharding and replication, Ex: MongoDB | Same as Semi-structued, Examples are Hadoop and GFS |
| Examples | Emails, XML, JSON, Binary files, Zipped files, web files | Videos, Reports, Surveys, Word docs, images and memos | |