Latent semantic indexing 290
External links
- Michael Berry’s site (http:/ / http://www. cs. utk. edu/ ~lsi/ )
- Gensim (http:/ / radimrehurek. com/ gensim) contains a Python+NumPy implementation of LSI for matrices
larger than the available RAM. - Text to Matrix Generator (TMG) (http:/ / scgroup. hpclab. ceid. upatras. gr/ scgroup/ Projects/ TMG/ ) MATLAB
toolbox that can be used for various tasks in text mining (TM) specifically i) indexing, ii) retrieval, iii)
dimensionality reduction, iv) clustering, v) classification. Most of TMG is written in MATLAB and parts in Perl.
It contains implementations of LSI, clustered LSI, NMF and other methods.
Further reading
Berry, M. W., Browne M., Understanding Search Engines: Mathematical Modeling and Text Retrieval, Philadelphia,
Society for Industrial and Applied Mathematics, (2005).
Berry, M. W., (Editor), Survey of Text Mining: Clustering, Classification, and Retrieval, New York, Springer,
(2004).
Landauer, T., et al., Handbook of Latent Semantic Analysis, Lawrence Erlbaum Associates, 2007.
Manning, C. D., Schutze H., Foundations of Statistical Natural Language Processing, Cambridge, MA, The MIT
Press, (1999).Semantic targeting
Semantic targeting is a technique enabling the delivery of targeted advertising for advertisements appearing on
websites and is used by online publishers and advertisers to increase the effectiveness of their campaigns. The
selection of advertisements are served by automated systems based on the content displayed to the user.Origins
Semantic Targeting has originated from the developments arising from Semantic Web. The Semantic Web enables
the representation of concepts expressed in human language to data in such a way that facilitates automatic
processing, where software can programmatically understand and reason how different elements of data are related.
The concept of semantic targeting utilises this capability to identify these concepts and the contexts in which they
occur, enabling marketers to deliver highly targeted and specific ad campaigns to webpages.The evolution of online advertising
The targeting of advertising to specific micro segments is a fundamental requirement for an effective ad campaign.
The two methods of targeting of recent times have been behavioral targeting and contextual targeting. It is now
generally accepted that these forms have pitfalls for both advertiser and consumer.
Behavioral targeting aggregates data based upon a user's viewing of pages from a website. Generally this is
facilitated by the placing of a cookie upon the user's PC. The cookie then reports the user's viewing behavior
allowing for the identification of patterns of viewing behavior. However, great concern is expressed about the
treatment of the user's right to privacy amongst consumer groups and legislators.[1][2]
Contextual advertising scans the content of webpages, seeking to identify keywords, against which advertisers have
bid to have their ad linked. If a match is made the ad is placed alongside the content, through an automated process.
However, such systems are unable to identify the context of the entire page and therefore, a placement could be
made against content that is inappropriate, derogatory or insensitive to the subject.[3][4][5][6] They are also unable to