Meta: Website design note

• Trying to develop a graph database-modelled website (for research-assistive purposes)
  • ...without outmoded metaphors, see: Skeuomorphic computing environment (discussion)
• model flow of traffic through concepts on a Wiki to say something semantic? not controversial that pageview = attention (metrics get a bad rap)
  • pageviews tied to user give a path of interest
  • Not trying to market or sell: find imaginative patterns
• model attentional processes: capture reader's imaginative approach rather than that of the writer (see: verticals - tech note and interpretation as "imaginative verticals")
• capturing learning process isn't exploitative/surveilled, afforded differential privacy
  • very small pages such that the (dominant) path of users exerts a structuring effect
    • pages linked: semantically, temporally (git wiki), imaginatively (user vectors)
    • pages sideloaded 1-link deep: external resources too?
    • modelled on Pascal's section divisions in the Pensées ^{TODO:terminology}
      • see: description of liasses ('bundles') in Permutation graphs and copermutations
  • classify paths rather than aggregate (i.e. using dominant reader path), rearranging the presented path according to some measure of similarity to known path classifications as reader progresses?
    • similar to recommender systems: key is how you instantiate a new user (naïve locus)
    • assumes fine level of control: seems unfeasible: unless procedure itself is on a graph
      • i.e. web page connections are structured by aforementioned dominant paths through concept graph relationships and overlaid permutation graph variety
• what would a graph cut signify in this context? ^TODO:clarify
  • "low level computer vision problems (early vision)" (via) ^TODO:explore
    • see: Graph cut approximation

---

If you simply do a brain dump of your process behind the project, you’ll do a pretty good job of satisfying one user story:

Other developers who are googling around because they’re trying something similar will get a head start on solving the problems we’ve solved.

Write what you know, right? But whoever asked you to write the post probably has other user stories in mind too.

🔰 Anna Schneider, The engineering blog meta-post