Implement solution for deduplication of random results

When requesting random results, we should make best efforts to deduplicate results for the current user. To action this, a couple of changes will need to be made.

1. A new cache region will need to be created that will contain a list of seen entries for a given user. This region should have time to live for entries, as we only want items to live for a shorter time to reduce impact on cache.
2. Logic will need to be added to retrieve a unique key for each user based on available information. Initial thought is to use OAuth token, but this has the implication that every user will be authenticated for API access.
3. Cache will need to be passed to the Query layer so that an explicit list of seen IDs can be included in the request for entries to exclude.
4. On successful query, post-processing will need to be done to record the newly seen entries into the cache entry for the current user.