24 - Conventional Ranking Models Content relevance Boolean...

Info icon This preview shows pages 1–7. Sign up to view the full content.

View Full Document Right Arrow Icon
Conventional Ranking Models Content relevance Boolean, vector space, probabilistic, language model, ... Page importance Link analysis: PageRank, HITS, ... Query log mining, clickthroughs, ... Machine learning for IR ranking? We’ve looked at methods for classifying documents using supervised machine learning classifiers Naive Bayes, Rocchio, kNN, SVMs, ... Surely we can also use machine learning to rank the documents displayed in search results? Sounds like a good idea => “machine-learned relevance” or “learning to rank”
Image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Learning to rank algorithms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
Image of page 2
Simple example: Using classification for ad hoc IR Collect a training corpus of (q,d,r) triples Relevance r is binary Document is represented by a feature vector x =( α , ω ) : α is cosine similarity; ω is minimum query window size Query term proximity is a very important new weighting factor
Image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Simple example: Using classification for ad hoc IR A linear score function is then: Score(d,q) = Score( α , ω ) = a α + b ω + c And the linear classifier is: Decide relevant if Score(q,d) > θ ... this is exactly like text classification Simple example: Using classification for ad hoc IR ! " # $ % !&!% !&!"% !"#$%&#!"(& ! )&(* +(",$*$-. " R R R R R R R R R R R N N N N N N N N N N '()*+*,- +/012)(
Image of page 4
Extending the model We can generalize this to classifier functions over more features We can use methods we have seen previously for learning the linear classifier weights Machine learning for IR ranking This “good idea” has been actively researched and actively deployed at major web search engines in the last 5 years Why didn’t it happen earlier? Modern supervised ML has been around for about 15 years Naive Bayes has been around for about 45 years!
Image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Machine learning for IR ranking There’s some truth to the fact that the IR community wasn’t very connected to the ML community But there were a whole bunch of precursors: Wong, S.K. et al. 1988. Linear structure in information retrieval. SIGIR 1988.
Image of page 6
Image of page 7
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern