Finding networks within networks

When Gary W. Flake searches for a certain computer scientist on the World Wide Web, he gets deluged with unwanted information about someone else. The scientist’s name? Michael Jordan.

A new algorithm developed by Flake of the NEC Research Institute in Princeton, N.J., and his colleagues could lead to more precise Web-searching engines, Flake says. Such an engine could, for instance, zero in on University of California, Berkeley Professor Jordan despite all those Web references to the basketball superstar.

Most conventional search engines seek Web pages with key words that simply match the words a searcher supplies, no matter the context. In the March issue of Computer, Flake and his coworkers describe their alternative: the community-identification algorithm. It recognizes communities of Web pages by taking advantage of patterns of interconnectivity in these so-called small-world networks (SN: 8/22/98, p. 124). In such networks, members tend to have lots of links to a small nucleus of other members and fewer links to everything else.

In proposed Web applications of Flake’s algorithm, a person would call up a Web page and then activate a program that follows links. It would start on the original page and crawl through the Web to find other pages for provisional membership in a community. Then, the algorithm would discard candidates with more than half of their links outside the evolving community.

In the case of Michael Jordan, the search engine could be asked to first identify a community of computer-science Web pages. Then typing “Michael Jordan” would lead to the professor rather than the sports icon, Flake says. The same algorithm would also be adept at demarcating seedy cyberneighborhoods as off-limits to children or bored employees, he adds. Moreover, because the algorithm can handle databases such as marketing data or phone records, it can potentially identify communities of shoppers with similar tastes or maybe even terrorist cells.

Use up and down arrow keys to explore.Use right arrow key to move into the list.Use left arrow key to move back to the parent list.Use tab key to enter the current list item.Use escape to exit the menu.Use the Shift key with the Tab key to tab back to the search input.
Logo

Looks like your ad blocker is on.

×

We rely on ads to keep creating quality content for you to enjoy for free.

Please support our site by disabling your ad blocker.

Continue without supporting us

Choose your Ad Blocker

  • Adblock Plus
  • Adblock
  • Adguard
  • Ad Remover
  • Brave
  • Ghostery
  • uBlock Origin
  • uBlock
  • UltraBlock
  • Other
  1. In the extension bar, click the AdBlock Plus icon
  2. Click the large blue toggle for this website
  3. Click refresh
  1. In the extension bar, click the AdBlock icon
  2. Under "Pause on this site" click "Always"
  1. In the extension bar, click on the Adguard icon
  2. Click on the large green toggle for this website
  1. In the extension bar, click on the Ad Remover icon
  2. Click "Disable on This Website"
  1. In the extension bar, click on the orange lion icon
  2. Click the toggle on the top right, shifting from "Up" to "Down"
  1. In the extension bar, click on the Ghostery icon
  2. Click the "Anti-Tracking" shield so it says "Off"
  3. Click the "Ad-Blocking" stop sign so it says "Off"
  4. Refresh the page
  1. In the extension bar, click on the uBlock Origin icon
  2. Click on the big, blue power button
  3. Refresh the page
  1. In the extension bar, click on the uBlock icon
  2. Click on the big, blue power button
  3. Refresh the page
  1. In the extension bar, click on the UltraBlock icon
  2. Check the "Disable UltraBlock" checkbox
  3. Marque la casilla de verificación "Desactivar UltraBlock"
  1. Please disable your Ad Blocker
  2. Disable any DNS blocking tools such as AdGuardDNS or NextDNS

If the prompt is still appearing, please disable any tools or services you are using that block internet ads (e.g. DNS Servers).