1 / 6

Innovations in Computational Linguistics and Peer-to-Peer Information Retrieval

This document summarizes key insights from presentations by Sven Abels, David Parry, Katarzyna Wegrzyn-Wolska, and Wai Gen Yee, exploring various topics in computational linguistics. Highlights include the splitting of compounds, the use of attribution for author identification, the dynamics of web page lifetimes, and the challenges of dynamic document archiving. It also delves into peer-to-peer information retrieval, addressing issues like reputation management and spam identification. These discussions pave the way for future improvements in language processing and information retrieval methodologies.

Télécharger la présentation

Innovations in Computational Linguistics and Peer-to-Peer Information Retrieval

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. ?Miscellaneous Group¿What we learned in Compiegne Sven Abels, David Parry, Katarzyna Wegrzyn-Wolska, Wai Gen Yee

  2. Contents • Abels: Splitting compounds • Parry: Attribution • Wegrzyn-Wolska: Web page lifetimes. • Yee: P2P information retrieval.

  3. jWordSplitter • Usage of "Bloom Filters"? • to test whether or not an element is a member of a set strong space advantage compared to hash tables • Connect words instead of splitting them? • Needed e.g. in China (Google vs. Baidu) • Reduction of dictionary to atomic words? • To further improve size of dictionary and checking time • Consideration of further language specific rules • For words that might need a grammatical change after the decomposition • Next: Evaluation of improvement • By using two projects as introduced in the presentation

  4. Attribution • K-distance has some similarity to n-grams, but the compression algorithms give more flexibility. • Location of centroids for clustering can be simplified—this may make clustering via this approach more practical. • The work in computational linguistics for author identification is related. • Use of the compression dictionary directly, may allow comparison between dictionaries rather than the “black box” approach.

  5. Lifetimes of Web Pages • The lifespan, accessibility and archiving of dynamic documents • Why is this problematic? • Interesting comments and questions: • measuring the lifespan of dynamic documents and its interest for the Search Engines. • definition of the lifespan, where the page can be consider as a new one.

  6. Peer-to-Peer IR • Reputations of peers. • Identify spoofers, spam. • Expand the model: • Development of P2P Googles.

More Related