NewsInEssence Help

Welcome to NewsInEssence!

NewsInEssence may be used to retrieve and summarize a cluster of articles from the web. NIE can start from a URL and retrieve documents that are similar, or NIE can retrieve documents that match a given set of keywords.

NewsInEssence also downloads hundreds of news articles daily and produces news clusters from them. NIE's own clustering system, CIDR (pronounced "cider"), produces clusters from over 20 sources. In addition, NIE also produces summaries of the clusters shown at news.google.com.

I. The UI

A. Nav Bar

The Navigation Bar to the left provides quick links to all parts of NewsInEssence, including summarizing and tracking an existing cluster, creating a new cluster, and viewing previously built clusters.

B. The Top Story

At the center of the top of the NewsInEssence home page, a partial summary of the "top story", is shown. Clicking either the headline or the "FULL SUMMARY" link will take you to a full summary of the story.

C. Cluster Lists

II. Viewing a Cluster

Use a pre-existing cluster:
You may create news summaries based on clusters of articles already gathered from Google by choosing the desired cluster number under Google News Clusters. You may also choose to use a cluster created by another using by selecting one of the cluster numbers under Recent User Clusters.

Select an exisiting summary:
If a summary has been previously created at any compression rate, a hyperlink to the summary will be available at the lower left on the screen. Click on the hyperlink to view the summary.

III. Summarizing a Cluster

Once you have selected or created a cluster (see the next section), a table will appear listing the source articles contained in your cluster. You can specify which articles to use and select a compression rate for your summary. Finally, you may choose whether to view the summary online or have it emailed to you.

Specify summary documents

Select article for summary:
You may choose which articles from the cluster to include in the summary. To do so, select or deselect articles from the document table by clicking on the check boxes located beside each article record.

Select article for seed:
You may also decide to use any one of the listed documents as a seed document for a new cluster. To do so, simply click on the the [Use as Seed] link provided in each article record.

Choose a compression rate and create summary

Select a compression rate to create a new summary:
If a summary at the compression level you want has not yet been created, you may create a summary at that compression rate by following these steps:

Select compression rate: Choose a compression rate from the drop down menu at the lower right on the screen.

Create summary: To view the new summary as it is created in real-time, select Live Summary (this may take several minutes). To have the summary emailed to you after it has been completed, enter your email address and select Email Summary.

IV. Creating a New Cluster

By Query:

You may specify a query by typing in key words in the query box. NIE will search the web for news articles related to your query.

By Seed:

You may also build a cluster from a seed article. NIE then uses this seed to find other articles related to the same topic. There are two ways you can specify a seed URL:
If you already have a seed URL in mind: If you know the URL for the article you wish to be the seed, type it in the seed URL box.
If you do not have a seed article in mind, you may find one by following the NIE Headlines hyperlink, and selecting an article from among the list of headlines. Simply follow the NIE link, which will take you to the NIE specify cluster section, with the selected article as the seed.

Specify cluster parameters

Once you have specified a seed article or query, you must specify the parameters for your cluster. These variables may be changed in the table provided.

Sources: Indicate the sources to use by selecting one or more of the news sources listed in the left side of the table. The following sources are currently available:

  • Chicago Suntimes
  • Globe and Mail
  • Guardian
  • International Herald Tribune
  • Newsday
  • Reuters
  • San Francisco Chronicle
  • Seattle Post Intelligencer
  • The Boston Herald

Min Article Similarity: Select the desired minimum article similarity (high, medium or low) that will be used by NIE for finding articles related to the seed.

Search Timeout: It is recommended that you use the default value (unlimited). NIE will send you an email when it is finished searching for related articles, if you provide your email address when prompted.

Max Sim Tests/Site: This is the maximum number of articles that will be compared to the seed article (or query) for similarity.