Chap. 10  M ining  Object,  Spatial, Multimedia, Text, and Web  Data Data Mining
Mining Complex Types of Data <ul><li>Mining spatial data </li></ul><ul><li>Mining image data </li></ul><ul><li>Mining text...
Mining Spatial Databases <ul><li>Spatial database </li></ul><ul><ul><li>Space related data: maps, VLSI layouts, … </li></u...
Example: BC Weather Pattern Analysis <ul><li>Input </li></ul><ul><ul><li>A map with about 3,000 weather probes scattered i...
Spatial Merge <ul><li>Precomputing: too much storage space </li></ul><ul><li>On-line merge: very expensive </li></ul>
Spatial Association Analysis   <ul><li>Spatial association rule: A      B  [ s %,  c %] </li></ul><ul><ul><li>A and B are...
<ul><li>Spatial classification </li></ul><ul><ul><li>Analyze spatial objects to derive classification schemes, such as dec...
<ul><li>Constraints- based clustering </li></ul><ul><ul><li>Selection of relevant objects before clustering </li></ul></ul...
Mining Image Data - Retrieval <ul><li>Description-based retrieval systems </li></ul><ul><ul><li>Retrieval based on image d...
Mining Image Data - Retrieval C ombining searches Search for “blue sky” (top layout grid is blue) Search for “airplane in ...
Classification of Image Data <ul><li>Classification  </li></ul><ul><ul><li>Decision tree </li></ul></ul><ul><ul><ul><li>Ba...
Classification of Image Data
Mining Text Databases <ul><li>Text databases (document databases)  </li></ul><ul><ul><li>Large collections of documents fr...
Basic Measures for IR <ul><li>Precision :  the percentage of retrieved documents that are in fact relevant to the query (i...
Keyword-Based Retrieval <ul><li>A document is represented by a set of keywords </li></ul><ul><ul><li>Retrieval by keyword ...
Similarity-Based Retrieval <ul><li>A document is represented as a  keyword vector </li></ul><ul><ul><li>Retrieval by simil...
TF-IDF Weighting <ul><li>TF (Term Frequency) </li></ul><ul><ul><li>TF= f( t,d )  : h ow many times term  t   appears in do...
Latent Semantic Indexing <ul><li>Reduce the dimension of keyword matrix </li></ul><ul><ul><li>To resolve the synonym probl...
SVD <ul><li>Singular Value Decomposition  </li></ul><ul><ul><li>Decompose the matrix  A mn   </li></ul></ul><ul><ul><li>A ...
SVD
Automatic Document Classification <ul><li>Motivation </li></ul><ul><ul><li>Automatic classification for the tremendous num...
Mining the World-Wide Web <ul><li>WWW provides rich sources for data mining </li></ul><ul><ul><li>Contents  information </...
Web Search Engines <ul><li>Index-based </li></ul><ul><ul><li>Search the Web, collect Web pages, index Web pages, and build...
Web Contents Mining - Classification <ul><li>Web page/site classification </li></ul><ul><ul><li>Assign a  class label to e...
Web Structure Mining <ul><li>Finding authoritative Web pages </li></ul><ul><ul><li>Retrieving pages that are not only rele...
HITS (Hyperlink-Induced Topic Search) <ul><li>Method </li></ul><ul><ul><li>Use an index-based search engine to form the  r...
Web Usage Mining <ul><li>Mining  Web log  records  </li></ul><ul><ul><li>Discover  user access patterns   </li></ul></ul><...
Web Usage Mining <ul><li>Applications </li></ul><ul><ul><li>Target potential customers for electronic commerce </li></ul><...
References <ul><li>H. Miller and J. Han (eds.), Geographic Data Mining and Knowledge Discovery, Taylor and Francis, 2001. ...
References <ul><li>G. Miller, R. Beckwith, C. FellBaum, D. Gross, K. Miller, and R. Tengi.  Five papers on WordNet.  Princ...
Upcoming SlideShare
Loading in...5
×

Chap. 10 Mining Object, Spatial, Multimedia, Text, and Web Data

5,745

Published on

0 Comments
3 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total Views
5,745
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
389
Comments
0
Likes
3
Embeds 0
No embeds

No notes for slide

Chap. 10 Mining Object, Spatial, Multimedia, Text, and Web Data

  1. 1. Chap. 10 M ining Object, Spatial, Multimedia, Text, and Web Data Data Mining
  2. 2. Mining Complex Types of Data <ul><li>Mining spatial data </li></ul><ul><li>Mining image data </li></ul><ul><li>Mining text data </li></ul><ul><li>Mining the Web </li></ul>
  3. 3. Mining Spatial Databases <ul><li>Spatial database </li></ul><ul><ul><li>Space related data: maps, VLSI layouts, … </li></ul></ul><ul><ul><li>Topological, distance information organized by spatial indexing structures </li></ul></ul><ul><li>Spatial data warehousing </li></ul><ul><ul><li>Issue: different representations & structures </li></ul></ul><ul><ul><li>Dimensions </li></ul></ul><ul><ul><ul><li>Nonspatial: 25-30 degree  hot </li></ul></ul></ul><ul><ul><ul><li>Spatial-to-nonspatial: “New York”  “western provinces” </li></ul></ul></ul><ul><ul><ul><li>Spatial-to-spatial: equi. temp region  0-5 degree region </li></ul></ul></ul><ul><ul><li>Measures </li></ul></ul><ul><ul><ul><li>numerical </li></ul></ul></ul><ul><ul><ul><li>Spatial: collection of spatial pointers (0-5 degree region) </li></ul></ul></ul>
  4. 4. Example: BC Weather Pattern Analysis <ul><li>Input </li></ul><ul><ul><li>A map with about 3,000 weather probes scattered in B.C. </li></ul></ul><ul><ul><li>Daily data for temperature, wind velocity, etc. </li></ul></ul><ul><ul><li>Concept hierarchies for all attributes </li></ul></ul><ul><li>Output </li></ul><ul><ul><li>A map that reveals patterns: merged (similar) regions </li></ul></ul><ul><li>Goals </li></ul><ul><ul><li>Interactive analysis (drill-down, slice, dice, pivot, roll-up) </li></ul></ul><ul><ul><li>Fast response time, Minimizing storage space used </li></ul></ul><ul><li>Challenge </li></ul><ul><ul><li>A merged region may contain hundreds of “primitive” regions (polygons) </li></ul></ul>
  5. 5. Spatial Merge <ul><li>Precomputing: too much storage space </li></ul><ul><li>On-line merge: very expensive </li></ul>
  6. 6. Spatial Association Analysis <ul><li>Spatial association rule: A  B [ s %, c %] </li></ul><ul><ul><li>A and B are sets of spatial or nonspatial predicates </li></ul></ul><ul><ul><ul><li>Topological relations: intersects, overlaps, disjoint , etc. </li></ul></ul></ul><ul><ul><ul><li>Spatial orientations: left_of, west_of, under , etc. </li></ul></ul></ul><ul><ul><ul><li>Distance information: close_to, within_distance , etc. </li></ul></ul></ul><ul><ul><li>Example </li></ul></ul><ul><ul><ul><li>is_a(x, “school”) ^ close_to(x, “sports_center”) </li></ul></ul></ul><ul><ul><ul><li> close_to(x, “park”) [7%, 85%] </li></ul></ul></ul><ul><li>Progressive Refinement </li></ul><ul><ul><li>First search for rough relationship (e.g. g_close_to for close_to, touch, intersect ) using rough evaluation (e.g. MBR) </li></ul></ul><ul><ul><li>Then apply only to those objects which have passed the rough test </li></ul></ul>
  7. 7. <ul><li>Spatial classification </li></ul><ul><ul><li>Analyze spatial objects to derive classification schemes, such as decision trees in relevance to spatial properties </li></ul></ul><ul><ul><li>Example </li></ul></ul><ul><ul><ul><li>Classify regions into rich vs. poor </li></ul></ul></ul><ul><ul><ul><li>Properties: containing university, containing highway, near ocean, etc. </li></ul></ul></ul>Spatial Classification
  8. 8. <ul><li>Constraints- based clustering </li></ul><ul><ul><li>Selection of relevant objects before clustering </li></ul></ul><ul><ul><li>Parameters as constraints </li></ul></ul><ul><ul><ul><li>K-means, density-based: radius, min points </li></ul></ul></ul><ul><ul><li>Clustering with obstructed distance </li></ul></ul>Spatial Cluster Analysis Spatial data with obstacles Clustering without taking obstacles into consideration Mountain River Bridge C1 C2 C3 C4
  9. 9. Mining Image Data - Retrieval <ul><li>Description-based retrieval systems </li></ul><ul><ul><li>Retrieval based on image descriptions, such as keywords, captions, size, etc. </li></ul></ul><ul><ul><li>Labor-intensive, poor quality </li></ul></ul><ul><li>Content-based retrieval systems </li></ul><ul><ul><li>Retrieval based on the image content( features ), such as color histogram, texture, shape, and wavelet transforms </li></ul></ul><ul><ul><li>Sample-based queries </li></ul></ul><ul><ul><ul><li>Find all of the images that are similar to the features of given image </li></ul></ul></ul><ul><ul><li>Feature specification queries </li></ul></ul><ul><ul><ul><li>Specify or sketch image features like color, texture, or shape, which are translated into a feature vector </li></ul></ul></ul>
  10. 10. Mining Image Data - Retrieval C ombining searches Search for “blue sky” (top layout grid is blue) Search for “airplane in blue sky” (top layout grid is blue and keyword = “airplane”)
  11. 11. Classification of Image Data <ul><li>Classification </li></ul><ul><ul><li>Decision tree </li></ul></ul><ul><ul><ul><li>Based on descriptive features </li></ul></ul></ul><ul><ul><ul><li>Based on content features </li></ul></ul></ul><ul><li>Feature extraction </li></ul><ul><ul><li>Extract features for classification from raw image </li></ul></ul><ul><ul><li>Various image analysis techniques are required </li></ul></ul><ul><ul><ul><li>Data transformation, edge detection, etc. </li></ul></ul></ul><ul><li>Example </li></ul><ul><ul><li>Classify sky images to recognize galaxies, stars, etc. </li></ul></ul><ul><ul><li>By using properties obtained from image analysis </li></ul></ul>
  12. 12. Classification of Image Data
  13. 13. Mining Text Databases <ul><li>Text databases (document databases) </li></ul><ul><ul><li>Large collections of documents from various sources </li></ul></ul><ul><ul><ul><li>News articles, research papers, books, e-mail messages, and Web pages </li></ul></ul></ul><ul><ul><li>Data stored is usually semi-structured </li></ul></ul><ul><ul><li>Traditional information retrieval techniques become inadequate for the increasingly vast amounts of text data </li></ul></ul><ul><li>Information retrieval </li></ul><ul><ul><li>Information is organized into documents </li></ul></ul><ul><ul><li>Information retrieval problem </li></ul></ul><ul><ul><ul><li>Locating relevant documents based on user input, such as keywords or example documents </li></ul></ul></ul>
  14. 14. Basic Measures for IR <ul><li>Precision : the percentage of retrieved documents that are in fact relevant to the query (i.e., “correct” responses) </li></ul><ul><li>Recall : the percentage of documents that are relevant to the query and were, in fact, retrieved </li></ul>
  15. 15. Keyword-Based Retrieval <ul><li>A document is represented by a set of keywords </li></ul><ul><ul><li>Retrieval by keyword matching </li></ul></ul><ul><li>Queries may use expressions of keywords </li></ul><ul><ul><li>(Car and accessory), (C++ or Java) </li></ul></ul><ul><li>Major difficulties </li></ul><ul><ul><li>Synonymy : same meaning but different word </li></ul></ul><ul><ul><ul><li>Ex> Q: “software”  Doc: about programming, do not have the keyword </li></ul></ul></ul><ul><ul><li>Polysemy : same word but different meaning </li></ul></ul><ul><ul><ul><li>Ex> Q: “mining”  Doc: about gold mining, have the keyword </li></ul></ul></ul>
  16. 16. Similarity-Based Retrieval <ul><li>A document is represented as a keyword vector </li></ul><ul><ul><li>Retrieval by similarity computing </li></ul></ul><ul><li>Basic techniques </li></ul><ul><ul><li>Stop list – set of words that are frequent but irrelevant </li></ul></ul><ul><ul><ul><li>Ex> a, the, of, for, with , … </li></ul></ul></ul><ul><ul><li>Stemming – use a common word stem </li></ul></ul><ul><ul><ul><li>Ex> drug , drugs, drugged  drug </li></ul></ul></ul><ul><ul><li>Weighting – count frequency </li></ul></ul><ul><ul><ul><li>Term frequency, inverse document frequency, … </li></ul></ul></ul><ul><li>Similarity metrics </li></ul><ul><ul><li>Measure the closeness of a document to a query </li></ul></ul><ul><ul><li>Cosine similarity : </li></ul></ul>
  17. 17. TF-IDF Weighting <ul><li>TF (Term Frequency) </li></ul><ul><ul><li>TF= f( t,d ) : h ow many times term t appears in doc d </li></ul></ul><ul><ul><li>More frequent  more relevant to topic </li></ul></ul><ul><ul><li>Normalization: </li></ul></ul><ul><ul><ul><li>Document length varies : relative frequency preferred </li></ul></ul></ul><ul><li>IDF (Inverse Document Frequency) </li></ul><ul><ul><li>IDF = 1 + log ( n / k ) : in h ow many documents term t appears </li></ul></ul><ul><ul><ul><li>n : tot al number of docs </li></ul></ul></ul><ul><ul><ul><li>k : # docs with term t appearing (the document frequency) </li></ul></ul></ul><ul><ul><li>Less frequent among documents  more discriminative </li></ul></ul><ul><li>TF-IDF weighting </li></ul><ul><ul><li>weight(t, d) = TF(t, d) * IDF(t) </li></ul></ul>
  18. 18. Latent Semantic Indexing <ul><li>Reduce the dimension of keyword matrix </li></ul><ul><ul><li>To resolve the synonym problem and the size problem </li></ul></ul><ul><ul><li>Use a singular value decomposition (SVD) techniques </li></ul></ul><ul><li>Example </li></ul>
  19. 19. SVD <ul><li>Singular Value Decomposition </li></ul><ul><ul><li>Decompose the matrix A mn </li></ul></ul><ul><ul><li>A mn = U mm S mn (V nn ) T </li></ul></ul><ul><li>Reduce dimension </li></ul><ul><ul><li>Select largest k singular values </li></ul></ul><ul><ul><li>A’ mn = U mk S kk (V nk ) T </li></ul></ul><ul><ul><li>Projection of A into k dimension </li></ul></ul><ul><ul><li>A’ mn V nk = U mk S kk </li></ul></ul><ul><ul><li>Computing similarity </li></ul></ul><ul><ul><li>AA T = USV T (USV T ) T </li></ul></ul><ul><ul><li> = USV T VS T U T </li></ul></ul><ul><ul><li> = (US)(US) T </li></ul></ul>
  20. 20. SVD
  21. 21. Automatic Document Classification <ul><li>Motivation </li></ul><ul><ul><li>Automatic classification for the tremendous number of on-line text documents (Web pages, e-mails, etc.) </li></ul></ul><ul><li>A classification problem </li></ul><ul><ul><li>Training set: Human experts generate a training data set </li></ul></ul><ul><ul><li>Classification(learning): The system discovers the classification rules </li></ul></ul><ul><li>Methods </li></ul><ul><ul><li>Extract keywords and weights from documents </li></ul></ul><ul><ul><ul><li>Documents are represented as (keyword, weight) pairs </li></ul></ul></ul><ul><ul><li>Classify training documents into classes </li></ul></ul><ul><ul><li>Apply classification algorithm </li></ul></ul><ul><ul><ul><li>Decision tree, Bayesian, neural network, etc. </li></ul></ul></ul>
  22. 22. Mining the World-Wide Web <ul><li>WWW provides rich sources for data mining </li></ul><ul><ul><li>Contents information </li></ul></ul><ul><ul><li>Hyperlink information </li></ul></ul><ul><ul><li>Usage information </li></ul></ul><ul><li>Challenges </li></ul><ul><ul><li>Too huge for effective data warehousing and data mining </li></ul></ul><ul><ul><li>Too complex and heterogeneous </li></ul></ul><ul><ul><li>Growing and changing very rapidly </li></ul></ul>
  23. 23. Web Search Engines <ul><li>Index-based </li></ul><ul><ul><li>Search the Web, collect Web pages, index Web pages, and build and store huge keyword-based indices </li></ul></ul><ul><ul><li>Locate sets of Web pages containing certain keywords </li></ul></ul><ul><li>Deficiencies </li></ul><ul><ul><li>A topic of any breadth may easily contain hundreds of thousands of documents </li></ul></ul><ul><ul><li>Many documents that are highly relevant to a topic may not contain keywords defining them (synonymy, polysemy) </li></ul></ul>
  24. 24. Web Contents Mining - Classification <ul><li>Web page/site classification </li></ul><ul><ul><li>Assign a class label to each web page from a set of predefined topic categories </li></ul></ul><ul><ul><li>Based on a set of examples of preclassified documents </li></ul></ul><ul><li>Example </li></ul><ul><ul><li>Use Yahoo!'s taxonomy and its associated documents as training and test sets </li></ul></ul><ul><ul><li>Derive a Web document classification model </li></ul></ul><ul><ul><li>Use the model to classify new Web documents by assigning categories from the same taxonomy </li></ul></ul><ul><li>Methods </li></ul><ul><ul><li>Keyword-based classification, use of hyperlink information, statistical models, … </li></ul></ul>
  25. 25. Web Structure Mining <ul><li>Finding authoritative Web pages </li></ul><ul><ul><li>Retrieving pages that are not only relevant, but also of high quality, or authoritative on the topic </li></ul></ul><ul><li>Hyperlinks can infer the notion of authority </li></ul><ul><ul><li>A hyperlink pointing to another Web page, this can be considered as the author's endorsement of the other page </li></ul></ul><ul><li>Problems </li></ul><ul><ul><li>Not every hyperlink represents an endorsement </li></ul></ul><ul><ul><li>One authority will seldom point to its rival authority </li></ul></ul><ul><ul><li>Authoritative pages are seldom particularly descriptive </li></ul></ul><ul><li>Hub </li></ul><ul><ul><li>Set of Web pages that provides collections of links to authorities </li></ul></ul>
  26. 26. HITS (Hyperlink-Induced Topic Search) <ul><li>Method </li></ul><ul><ul><li>Use an index-based search engine to form the root set </li></ul></ul><ul><ul><li>Expand the root set into a base set </li></ul></ul><ul><ul><ul><li>Include all of the pages that the root-set pages link to, and all of the pages that link to a page in the root set </li></ul></ul></ul><ul><ul><li>Apply weight-propagation </li></ul></ul><ul><ul><ul><li>Determines numerical estimates of hub and authority weights </li></ul></ul></ul><ul><ul><li>Output a list of the pages </li></ul></ul><ul><ul><ul><li>Large hub weights, large authority weights for the given search topic </li></ul></ul></ul><ul><li>Systems based on the HITS algorithm </li></ul><ul><ul><li>Clever, Google </li></ul></ul><ul><ul><ul><li>Achieve better quality search results than AltaVista, Yahoo! </li></ul></ul></ul>
  27. 27. Web Usage Mining <ul><li>Mining Web log records </li></ul><ul><ul><li>Discover user access patterns </li></ul></ul><ul><ul><li>Typical Web log entry - URL requested, the IP address from which the request originated, timestamp, etc. </li></ul></ul><ul><li>OLAP on the Weblog database </li></ul><ul><ul><li>Find the top N users, top N accessed Web pages, most frequently accessed time periods, etc. </li></ul></ul><ul><li>Data mining on Weblog records </li></ul><ul><ul><li>Find association patterns, sequential patterns, and trends of Web accessing </li></ul></ul>
  28. 28. Web Usage Mining <ul><li>Applications </li></ul><ul><ul><li>Target potential customers for electronic commerce </li></ul></ul><ul><ul><li>Identify potential prime advertisement locations </li></ul></ul><ul><ul><li>Enhance the quality and delivery of Internet information services to the end user </li></ul></ul><ul><ul><li>Improve Web server system performance </li></ul></ul><ul><ul><ul><li>Web caching, Web page prefetching, and Web page swapping </li></ul></ul></ul>
  29. 29. References <ul><li>H. Miller and J. Han (eds.), Geographic Data Mining and Knowledge Discovery, Taylor and Francis, 2001. </li></ul><ul><li>Ester M., Frommelt A., Kriegel H.-P., Sander J.: Spatial Data Mining: Database Primitives, Algorithms and Efficient DBMS Support , Data Mining and Knowledge Discovery, 4: 193-216, 2000. </li></ul><ul><li>J. Han, M. Kamber, and A. K. H. Tung, &quot;Spatial Clustering Methods in Data Mining: A Survey&quot;, in H. Miller and J. Han (eds.), Geographic Data Mining and Knowledge Discovery, Taylor and Francis, 2000. </li></ul><ul><li>Y. Bedard, T. Merrett, and J. Han, &quot;Fundamentals of Geospatial Data Warehousing for Geographic Knowledge Discovery&quot;, in H. Miller and J. Han (eds.), Geographic Data Mining and Knowledge Discovery, Taylor and Francis, 2000 </li></ul><ul><li>K. Koperski and J. Han. Discovery of spatial association rules in geographic information databases. SSD'95. </li></ul><ul><li>Shashi Shekhar and Sanjay Chawla, Spatial Databases: A Tour , Prentice Hall, 2003 (ISBN 013-017480-7). Chapter 7. : Introduction to Spatial Data Mining </li></ul><ul><li>X. Li, J. Han, and S. Kim, Motion-Alert: Automatic Anomaly Detection in Massive Moving Objects”, IEEE Int. Conf. on Intelligence and Security Informatics (ISI'06). </li></ul><ul><li>Fabrizio Sebastiani, “Machine Learning in Automated Text Categorization”, ACM Computing Surveys, Vol. 34, No.1, March 2002 </li></ul><ul><li>Soumen Chakrabarti, “Data mining for hypertext: A tutorial survey”, ACM SIGKDD Explorations, 2000. </li></ul><ul><li>Cleverdon, “Optimizing convenient online access to bibliographic databases”, Information Survey, Use4, 1, 37-47, 1984 </li></ul><ul><li>Yiming Yang, “An evaluation of statistical approaches to text categorization”, Journal of Information Retrieval, 1:67-88, 1999. </li></ul><ul><li>Yiming Yang and Xin Liu “A re-examination of text categorization methods”. Proceedings of ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR'99, pp 42--49), 1999. </li></ul><ul><li>S. Chakrabarti, “Mining the Web: Statistical Analysis of Hypertext and Semi-Structured Data” , Morgan Kaufmann, 2002. </li></ul>
  30. 30. References <ul><li>G. Miller, R. Beckwith, C. FellBaum, D. Gross, K. Miller, and R. Tengi. Five papers on WordNet. Princeton University, August 1993. </li></ul><ul><li>M. Hearst, Untangling Text Data Mining , ACL’99, invited paper. R. Sproat, Introduction to Computational Linguistics , LING 306, UIUC, Fall 2003. </li></ul><ul><li>A Road Map to Text Mining and Web Mining, University of Texas resource page. http://www.cs.utexas.edu/users/pebronia/text-mining/ </li></ul><ul><li>Deng Cai, Shipeng Yu, Ji-Rong Wen and Wei-Ying Ma, “Extracting Content Structure for Web Pages based on Visual Representation”, The Fifth Asia Pacific Web Conference, 2003. </li></ul><ul><li>Deng Cai, Shipeng Yu, Ji-Rong Wen and Wei-Ying Ma, “VIPS: a Vision-based Page Segmentation Algorithm”, Microsoft Technical Report (MSR-TR-2003-79), 2003. </li></ul><ul><li>Shipeng Yu, Deng Cai, Ji-Rong Wen and Wei-Ying Ma, “Improving Pseudo-Relevance Feedback in Web Information Retrieval Using Web Page Segmentation”, 12th International World Wide Web Conference (WWW2003), May 2003. </li></ul><ul><li>Ruihua Song, Haifeng Liu, Ji-Rong Wen and Wei-Ying Ma, “Learning Block Importance Models for Web Pages”, 13th International World Wide Web Conference (WWW2004), May 2004. </li></ul><ul><li>Deng Cai, Shipeng Yu, Ji-Rong Wen and Wei-Ying Ma, “Block-based Web Search”, SIGIR 2004, July 2004 . </li></ul><ul><li>Deng Cai, Xiaofei He, Ji-Rong Wen and Wei-Ying Ma, “Block-Level Link Analysis”, SIGIR 2004, July 2004 . </li></ul><ul><li>Deng Cai, Xiaofei He, Wei-Ying Ma, Ji-Rong Wen and Hong-Jiang Zhang, “Organizing WWW Images Based on The Analysis of Page Layout and Web Link Structure”, The IEEE International Conference on Multimedia and EXPO (ICME'2004) , June 2004 </li></ul><ul><li>Deng Cai, Xiaofei He, Zhiwei Li, Wei-Ying Ma and Ji-Rong Wen, “Hierarchical Clustering of WWW Image Search Results Using Visual, Textual and Link Analysis”,12th ACM International Conference on Multimedia, Oct. 2004 . </li></ul>
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×