BigQuery for unstructured data

1
votes

I am contemplating the idea of using BiqQuery for data (unstructured) analysis.

I am aware that the ability to run ad-hoc queries over terabytes of data forms one of the biggest strengths of BigQuery.

How do I use this potential to handle unstructured data?

As per my understanding of BigQuery, it operates over data represented in form of relations, and that is the structure to follow when you feed data to BigQuery.

Is there any way BigQuery be made to operate over unstructured data, say for example data contained in documents? (Without of-course processing the documents first and then feeding the output to BigQuery.)

1
bigdatagoogle-bigquerydata-analysis
You could load the document, and from there create structured data this is discussed here: stackoverflow.com/questions/18899159/… - Pentium10

1 Answers

1
votes

BigQuery works with SQL (Structured Query Language) over tables stored in columnar format - so everything is pretty structured.

Still, you could import documents into BigQuery in a one string column table that can store up to 2MB per line - then you could apply the power of BigQuery to that text - as long as you can express your analysis using SQL.

Coming soon: The ability to write Javascript inside your SQL queries.