[back] feat: include entity contexts in the public dataset?

[GresilleSiffle]

Context

The entity contexts are designed to inform the users, to make them aware of specific events before comparing videos.

Because they can influence the user's judgements, and impact the individual and collective scores, should we include them in the public dataset? It may have value for the researchers.

See the initial discussion here: #1873 (comment) (if we add the contexts to the public dataset, we can then update the ComparisonSerializer as discussed in the previous comment)

[GresilleSiffle]

I think it's a good idea to include the entity contexts in the public dataset. Without them, the individual score cannot be fully "explained" by the data already present in the dataset, or the video metadata / scripts.

If we include them, we could also add a column publication date so that researchers can make a distinction between comparisons made before a context was published, and comparisons made after.