I’ve written a little bot that listens out for people talking about Sydenham, and retweets the most relevant tweets:

If anyone has ideas for this bot, I’m interested to hear.

Thanks

TODO

Filter out low-quality tweets

Throttle tweets from over-tweeters

Follow human-curated list of “tastemaker” tweeters to analyse which accounts / content they like/retweet

Identify high quality tweeters via like_rate/follower_count and RT_rate/follower_count

Filter out mentions of SE26 that aren’t about the postcode

Use following/followers of user to see if they’re in the clique of relevant local users

Use DBPedia (Wikipedia interface for bots) to identify keywords related to Sydenham

Filter repetitive social media spamming

Manually curated - mute and retract RT

Identify spam tweets by proportion of hashtags

Filter tweets with text matching earlier tweets (from any account)

Bayesian filter?

Reduce repetition of the same news

Natural language processing to extract semantics and keywords, then compare to previous tweets. See https://code.google.com/archive/p/ws4j/

Filter Obcenities

Word blacklist?

Bayesian filter?

Filter non-English / non-local tweets

Language filter

Bayesian filter?

Prevent gaming

Human curation (blocking bad actors)

Whitelist accounts by analysing quality metrics (e.g. average like count per tweet per follower count)? New / spam accounts will have low scores for these metrics.

Not sure about filtering non-English IF you have all the other filters working well - there could be a Non-native English speaker looking for support / social groups / meet-ups with people from same background etc?

Just a thought and doubtful there would be loads of non-English tweets about Sydenham/SE26 anyways?

Thanks for the feedback Jon. Perhaps there is scope for more bots that each act for a different languages and cater for different communities within Sydenham.