Skip to content

Latest commit

 

History

History
60 lines (32 loc) · 2.37 KB

notes.md

File metadata and controls

60 lines (32 loc) · 2.37 KB

Covered in meeting on 10/5

What we have

Search quality types - Addresses, dates, names, good searches, etc. - This is where by user data would be helpful

Topic search tool - To be able to look at searches by topics - Can do more analytics around searches by topic

General field explanations - Simple summaries from some of these fields

Potential questions we can explore

Are people finding what they are looking for? - how can we identify successful vs unsuccessful search.

Where do people search from? - different searches per search area

Maybe people go from categories find what they are looking for more often.

Can we figure out what things people search for often, but is not available?

To ask Jason

Can get some sort of time stamp or session data?

What other specific questions would you like to see us explore?

Jason

  • A set of compelling summary graphics. "Insights that are blog worthy" "We kind of knew this before but not to what extent." "Something simple that sums up the analysis or work"

  • Bao: Most known people searched for

  • From operational perspective: interested in categories and summaries

  • Might not be able to say with certainty but people t

  • implication of all geo data is available as a shapefile, but they don't have shapefile in the name anymore (so they will not get what they want)

  • Basically layout anything

  • (optional) Findings to share with Socrata about what makes the portal less useful (like not having spell checking)

Specific Questions

  • What are people interested in finding in the portal

  • Unique interesting and surprising things people are searching for on the portal

  • Could people find what they are looking for on the portal.

    • What % of search terms actually produce results (specifically ones that match intent of the search) i.e. say we didn't have crime data but have something that reference crime Failure rate: did the person get where they want to go or not Are there any terms that are searched a lot but don't have data related to it.
  • Based on searching are people looking for data sets we do have but they don't find

    • Failure by their search terms not matching our descriptions.
    • Failure by us having the data but not making it public yet.

TODO

  • Do a little research in advance that we structure them the right way, don't add too many fields we don't have to have