Xerox Corporation

Zerox Corporation

Marti A Hearst

Eric Saund

Chandu A Karadi

An interactive interface for visualizing results from a search of a corpus of machine-readable documents, each of which is associated with at least one category of a category hierarchy. The interactive interface includes a page object generation component, a book generation component, a page turning component, a cone tree generation component, and a cone tree rotation component. The page object generation component generates for each document of the results a link page and a content page. The link page of each document includes at least a link to categories of a category cone tree, which visually represents the category hierarchy. The content page of each document includes a portion of the document. The book generation component generates a visual representation of the pages for the documents of the results as a book. The book includes a first area for displaying a first link page of a first document of the results and a second area for displaying a first content page of the first document. The page turning component responds to a cursor control device by causing a second link page of a second document of the results to be displayed in the first area and a second content page of the second document of the results to be displayed in the second area. The cone tree generation component generates and displays the cone tree, which is perceptible as three dimensional and which represents a portion of the category hierarchy. The cone tree represents each category as a node and has a primary display position for displaying first category nodes associated with a first document. The cone tree rotation component responds to the page turning component by causing second category nodes associated with the second document to be displayed in the primary viewing position.

Interactive interface for viewing retrieval results

An interactive interface for creating a search query for a corpus of machine-readable documents, each of which is associated with at least one category of a category hierarchy. The interactive interface includes a cone tree generation component, a query specification component, a begin search component, and a query generation component. The cone tree generation component generates and displays a cone tree representing the category hierarchy. The cone tree represents each category of the category hierarchy as a node having a selection object for indicating inclusion of the category in a one of a first group, and a second group. Each selection object is responsive to a cursor control device. The query specification component generates and displays a query specification object including a first group object and a second group object. Each group object is responsive to the cursor control device and indicates members of the group. Each group includes a one of a term included within the corpus and a category of the category cone tree. The begin search component generates and displays a begin search object responsive to the cursor control device. The query generation component generates a query for searching the corpus to find documents that include at least one member of the first group and at least member of the second group.

Interactive interface for specifying searches

An unsupervised method of learning the relationships between words and unspecified topics in documents using a computer is described. The computer represents the relationships between words and unspecified topics via word clusters and association strength values, which can be used later during topical characterization of documents. The computer learns the relationships between words and unspecified topics in an iterative fashion from a set of learning documents. The computer preprocesses the training documents by generating an observed feature vector for each document of the set of training documents and by setting association strengths to initial values. The computer then determines how well the current association strength values predict the topical content of all of the learning documents by generating a cost for each document and summing the individual costs together to generate a total cost. If the total cost is excessive, the association strength values are modified and the total cost recalculated. The computer continues calculating total cost and modifying association strength values until a set of association strength values are discovered that adequately predict the topical content of the entire set of learning documents.

Method for learning to infer the topical content of documents based upon

An iterative method of determining the topical content of a document using a computer. The processing unit of the computer determines the topical content of documents presented to it in machine readable form using information stored in computer memory. That information includes word-clusters, a lexicon, and association strength values. The processing unit beings by generating an observed feature vector for the document being characterized, which indicates which of the words of the lexicon appear in the document. Afterward, the processing unit makes an initial prediction of the topical content of the document in the form of a topic belief vector. The processing unit uses the topic belief vector and the association strength values to predict which words of the lexicon should appear in the document. This prediction is represented via a predicted feature vector. The predicted feature vector is then compared to the observed feature vector to measure how well the topic belief vector models the topical content of the document. If the topic belief vector adequately model the topical content of the document, then the processing unit's task is complete. On the other hand, if the topic belief vector does not adequately model the topical content of the document, then the processing unit determines how the topic belief vector should be modified to improve the prediction of modeling of the topical content.

Growing community of inventors

Kensington, CA, United States of America

Marti A Hearst