Online news streaming company deicides to develop a utility$-$based intelligent agent that delivers personalized news for every reader. The set of news is organized, sequenced and collated so that the news order is specific to every client based on his$/$her emotional makeup and progressive relevance.

Assume news with set of $50$ news $($date$,$ title and news content$).$ Also, take a set of $15$ readers $($assume profile of readers $$ profession, age, gender, recent news read, location$)$

a$.$ First use hill climbing $+$ with supervised learning to identify and map the pivot news to every reader

b$.$ Collate and order news set for each reader using simulated annealing

c$.$ Improve the order using reinforcement learning with reference to social context

d$.$ Give detailed algorithm

e$.$ Draw the detailed diagram for the above agent

f$.$ Provide pseudo code

$2.$ In above example use K means clustering and cluster the news $$ map these clusters to readers. The centroid will be used to assign the pivot news. Re$-$write the above algorithm using clustering.

$3.$ Propose your own $($new and different$)$ approach to deliver better solution. Provide all details of this approach including diagram, pseudo code and mathematical model.