Useful meeting with Duncan Brown from CSD this morning, in which we invited his comments on our draft Project Plan. Some outcomes and suggestions as follows:
- Idea for dealing with Google Custom Search API’s current quota of 100 free requests per day
- Consider pre-population of database by using a service running in the background; likely to be quicker than relying on live searching across domains. The search engine will then first query the database before deciding whether to launch a new request to Google CSE
- Design UI to include a mechanism to allow users to mark individual results as “no longer available/valid“
- Design UI to include a mechanism to allow users to mark duplicate/similar resources