Transcript Application of Data Mining in TV and Films

Application of Data
Mining in TV and Films
Daniel Johnston and Nabeel Hanif
Aim
●
●
To look at the use of data mining within the
Television and Film industry.
To examine how DM is able to improve the
Tv/Film industry for both viewers and companies
Who uses Data Mining?
A Variety of companies linked to the film industry:
●
IMDB – vast database available on movies
●
Amazon – use their user DB to recommend films
●
Netflix – commissioned a show called house of
cards.
Using algorithms to predict
success
Using algorithms to predict
success
●
●
●
Netflix relied on Data Mining when
commissioning 2 series' of the show 'House of
Cards' at a cost of $100m (10% of revenue),
without even viewing a pilot episode.
By parsing subscriber data, Netflix were able to
see how popular the show would likely be.
This data indicated that people who enjoyed the
original BBC miniseries of House of Cards, also
liked movies starring Kevin Spacey and David
Fincher.
Did it work?
●
●
●
●
Yes.
Compared to a $1.9m loss in 2012, Netflix posted
a $31.8m profit this year.
Gained 2 million new US subscribers.
Netflix does not release viewing figures for
original content, but acknowledged House of
Cards had provided a 'Halo effect' on their entire
service.
How else is Data Mining used?
●
●
●
●
In the analysis of scripts to select “successful”
proposals.
Predicting success via analysis of web searches
relating to the film.
Calculating whether cinema attendance has an
effect on DVD sales.
Preventing spoilers.
Spoiler prevention
●
●
A computer science PhD student and his
supervisor created a data mining algorithm that
identified spoilers by using linguistic cues.
By linking combinations of words relevant to a
films plot, he was able to avoid spoilers in film
reviews that may not have been tagged by the
website IMDB.
Conclusion
●
●
●
Data mining benefits customers by improving the
service provided to them, in the form of better
Film and TV.
Benefits companies in the form of greater profits if
implemented correctly.
Although data mining has led to improvements,
the entertainment industry is still very much a
creative business – meaning data analysis can only
go so far.
Any
Questions?
Thanks for
listening!
References
●
https://www.student.gsu.edu/~mkaul1/DMMidtermPaper.htm
●
http://filmmakermagazine.com/64544-house-of-cards-the-netflix-algorithms-and-me/
●
http://www.kdd.org/sites/default/files/issues/14-2-2012-12/V14-02-05-Amatriain.pdf
●
http://www.nytimes.com/2013/02/25/business/media/for-house-of-cards-using-big-data-toguarantee-its-popularity.html?_r=0
●
http://www.fastcodesign.com/1671893/the-secret-sauce-behind-netflixs-hit-house-of-cardsbig-data