\relax \@writefile{toc}{\contentsline {chapter}{\numberline {1}Locality-Sensitive Hashing}{1}} \@writefile{lof}{\addvspace {10\p@ }} \@writefile{lot}{\addvspace {10\p@ }} \@writefile{toc}{\contentsline {section}{\numberline {1.1}Similarity of Documents}{1}} \newlabel{sec:lsh-simdocs}{{1.1}{1}} \@writefile{toc}{\contentsline {section}{\numberline {1.2}Similarity-Preserving Summaries of Sets}{3}} \newlabel{sec:lsh-simpres}{{1.2}{3}} \@writefile{lot}{\contentsline {table}{\numberline {1.1}{\ignorespaces A characteristic matrix for 4 sets with a universe consisting of 5 elements.}}{3}} \newlabel{tbl:chr-mat}{{1.1}{3}} \@writefile{lot}{\contentsline {table}{\numberline {1.2}{\ignorespaces Characteristic matrix after the permutation of rows of Table 1.1\hbox {}.}}{3}} \newlabel{tbl:permuted}{{1.2}{3}} \newlabel{lem:js-sim}{{1.2.1}{4}} \@writefile{lot}{\contentsline {table}{\numberline {1.3}{\ignorespaces A signature matrix for 4 sets.}}{4}} \newlabel{tbl:minhash-sig}{{1.3}{4}} \citation{Rajaraman10} \@writefile{toc}{\contentsline {section}{\numberline {1.3}LSH for Minhash Signatures}{5}} \newlabel{sec:lsh-minhash}{{1.3}{5}} \@writefile{lof}{\contentsline {figure}{\numberline {1.1}{\ignorespaces A signature matrix which has been separated into four bands of three rows each.}}{6}} \@writefile{lof}{\contentsline {figure}{\numberline {1.2}{\ignorespaces The S-curve.}}{7}} \@writefile{toc}{\contentsline {section}{\numberline {1.4}Theory of Locality Sensitive Functions}{7}} \newlabel{sec:lsh-theory}{{1.4}{7}} \newlabel{def:sensitive}{{1.4.1}{7}} \@writefile{lof}{\contentsline {figure}{\numberline {1.3}{\ignorespaces A smaller distance between items corresponds to a higher probability of similarity.}}{8}} \@writefile{lot}{\contentsline {table}{\numberline {1.4}{\ignorespaces Illustration of four families obtained for different values of $p$. ${\cal F}_1$ is the AND family for $r=5$. ${\cal F}_2$ is OR family for $b=5$. ${\cal F}_3$ is the AND-OR family for $r=5$ and $b=5$. ${\cal F}_4$ is the OR-AND family for $r=5$ and $b=5$.}}{9}} \newlabel{tbl:amplification}{{1.4}{9}} \@writefile{toc}{\contentsline {section}{\numberline {1.5}LSH Families}{9}} \newlabel{sec:lsh-fams}{{1.5}{9}} \@writefile{toc}{\contentsline {subsection}{\numberline {1.5.1}Hamming Distance}{9}} \@writefile{toc}{\contentsline {subsection}{\numberline {1.5.2}Cosine Distance}{10}} \@writefile{lof}{\contentsline {figure}{\numberline {1.4}{\ignorespaces Two vectors $x$ and $y$ shown in a plane; two possible hyperplanes are also shown.}}{10}} \newlabel{fig:lsh-cosinedist}{{1.4}{10}} \@writefile{toc}{\contentsline {subsection}{\numberline {1.5.3}Euclidean Distance}{11}} \@writefile{lof}{\contentsline {figure}{\numberline {1.5}{\ignorespaces With a distance $d > 2a$, points $x$ and $z$ are in separate intervals unless $\theta $ is large enough.}}{11}} \@writefile{toc}{\contentsline {subsection}{\numberline {1.5.4}Fingerprint Matching}{12}} \@writefile{toc}{\contentsline {subsection}{\numberline {1.5.5}Image Similarity}{13}} \@writefile{toc}{\contentsline {subsubsection}{VisualRank}{13}} \@writefile{lof}{\contentsline {figure}{\numberline {1.6}{\ignorespaces Images are considered similar if their local features are hashed into the same bin in multiple hash tables. In this example, the blue and green circles are buckets where features are hashed; $B$ is similar to $C$, and $C$ is similar to $D$.}}{13}} \@writefile{toc}{\contentsline {subsubsection}{$p$-Stable Distributions}{14}} \citation{Manber94} \citation{Broder97} \@writefile{toc}{\contentsline {subsubsection}{$\chi ^2$ Distance}{15}} \@writefile{toc}{\contentsline {section}{\numberline {1.6}Bibliographic Notes}{15}} \newlabel{sec:lsh-bib}{{1.6}{15}} \citation{Indyk97,Indyk98} \citation{Indyk98} \citation{Broder97} \citation{Indyk06} \citation{Broder98} \citation{Indyk04} \citation{Indyk98} \citation{Indyk06} \citation{Kang04} \citation{Hu05} \citation{Indyk97} \citation{Gorisse12} \citation{Jing08} \citation{Kulis12} \citation{Mondal13} \citation{dePaula10} \citation{Jeon12} \citation{Rasheed12} \citation{Kala13} \citation{Wang12} \citation{Indyk04} \citation{Kang04} \citation{*} \bibstyle{plain} \bibdata{biblio} \@writefile{toc}{\contentsline {section}{\numberline {1.7}Exercises}{17}} \@writefile{lot}{\contentsline {table}{\numberline {1.5}{\ignorespaces A characteristic matrix.}}{17}} \newlabel{tbl:chr-mat-ex}{{1.5}{17}} \bibcite{Indyk06}{1} \bibcite{Broder98}{2} \bibcite{Broder97}{3} \bibcite{Charikar02}{4} \bibcite{Indyk04}{5} \bibcite{dePaula10}{6} \bibcite{Indyk97}{7} \bibcite{Gorisse12}{8} \bibcite{Hu05}{9} \bibcite{Indyk98}{10} \bibcite{Jeon12}{11} \bibcite{Jing08}{12} \bibcite{Kala13}{13} \bibcite{Kang04}{14} \bibcite{Kulis12}{15} \bibcite{Manber94}{16} \bibcite{Mondal13}{17} \bibcite{Rajaraman10}{18} \bibcite{Rasheed12}{19} \bibcite{Raval11}{20} \bibcite{Wang12}{21}