webLyzard technology

Topic Management and Email Alerts

admin — Sat, 17 Mar 2012 18:11:52 +0000

Registered users can add and modify topics through the topic editor, which also provides the option to set customized email alerts. The ‘settings’ symbol to activate the editor is available on mouse-over in the topic management section.

Topic Editor. Each topic is represented as a list of regular expressions (= term list with optional wildcards for matching character strings). For computing the charts and ranking search results, a document is considered relevant to the topic if it contains at least one of the stated terms. The topic label (= the name to be displayed in the topic management section) itself is not considered in the matching process.
Email Alert. For each topic, users can specify a frequency value that will trigger an alert and choose specific weekdays to receive the notifications. A sentiment filter allows to restrict the alert to positive or negative mentions. Small dots in the topic markers indicate that an active alert exists.

Sentiment Detection

admin — Fri, 06 Jan 2012 11:09:23 +0000

Measures of bias in news and social media coverage are essential when investigating trends and differing perceptions of various interest groups. A significant portion of news and social media coverage contains opinions with clear economic relevance – customer and travel reviews, for example, or the articles of well-known and respected bloggers who influence purchase decisions. Analyzing and acting upon user-generated content is becoming imperative for decision makers who aim to engage large user communities.

The ever increasing amount of articles and the limits of human cognition require automated approaches to analyzing the sentiment expressed in user-generated content. As part of opinion mining, sentiment detection identifies and aggregates polar opinions – i.e., positive or negative statements about facts. For achieving accurate results, one needs to deal with the inherent ambiguities of human languages. webLyzard’s method to determine sentiment automatically has continually been optimized since 2003, directing particular attention to the context of opinionated terms when resolving such ambiguities.

webLyzard not only uses sentiment information to enrich visualizations such as tag clouds, geographic maps and information landscapes, but also offers high-performance data services for tagging third-party content.

Selected Publications

Gindl, S., Weichselbraun, A. and Scharl, A. (2010). “Cross-Domain Contextualization of Sentiment Lexicons”, 19th European Conference on Artificial Intelligence (ECAI-2010). H. Coelho et al. Lisbon, Portugal: IOS Press: 771-776.
Scharl, A. and Weichselbraun, A. (2008): “An Automated Approach to Investigating the Online Media Coverage of US Presidential Elections”, Journal of Information Technology & Politics, 5(1): 121-132.
Scharl, A., Pollach, I. and Bauer, C. (2003). “Determining the Semantic Orientation of Web-based Corpora”, Intelligent Data Engineering and Automated Learning, 4th International Conference (IDEAL-2003), Hong Kong (Lecture Notes in Computer Science, vol. 2690). Ed. J. Liu et al. Berlin: Springer. 840-849.

The ever increasing amount of articles and the limits of human cognition require automate approaches to analyzing the sentiment expressed in user-generated content. As part of opinion mining, sentiment detection identifies and aggregates polar opinions – i.e., positive or negative statements about facts. For achieving accurate results, one needs to deal with the inherent ambiguities of human languages. webLyzard’s method to determine sentiment automatically has continually been optimized since 2003, directing particular attention to the context of opinionated terms when resolving such ambiguities.

webLyzard not only embeds sentiment information (green = positive; red = negative) in various visualizations such as tag clouds, geographic maps and information landscapes, but also offers high-performance data services to tag third-party content.

Selected Publications

Weichselbraun, A., Gindl, S. and Scharl, A. (2011). Using Games with a Purpose and Bootstrapping to Create Domain-Specific Sentiment Lexicons. 20th ACM Conference on Information and Knowledge Management (CIKM-2011). Glasgow, UK: ACM : 1053-1060.

Gindl, S., Weichselbraun, A. and Scharl, A. (2010). “Cross-Domain Contextualization of Sentiment Lexicons”, 19th European Conference on Artificial Intelligence (ECAI-2010). H. Coelho et al. Lisbon, Portugal: IOS Press: 771-776.

Scharl, A. and Weichselbraun, A. (2008): “An Automated Approach to Investigating the Online Media Coverage of US Presidential Elections”, Journal of Information Technology & Politics, 5(1): 121-132.

Scharl, A., Pollach, I. and Bauer, C. (2003). “Determining the Semantic Orientation of Web-based Corpora”, Intelligent Data Engineering and Automated Learning, 4th Inter-national Conference, IDEAL-2003, Hong Kong (Lecture Notes in Computer Science, vol. 2690). Ed. J. Liu et al. Berlin: Springer. 840-849.

Trend Chart

admin — Mon, 06 Dec 2010 08:00:36 +0000

Showing the rise and decay of topics over the last two months (= default setting, which can be changed using the date selector), the trend chart provides the following time series:

Frequency (Total, Positive and Negative). The displayed frequency values represent the number of occurrences in the last seven days. Selecting positive or negative coverage through the global sentiment filter affects the data displayed in the chart. Once activated, the footer of the portal also shows the current filter status.
Sentiment (Mean). This time series shows the average sentiment towards the selected topic for the selected source and interval.
Disagreement (Standard Deviation of Sentiment): The standard deviation of the sentiment distribution reflects how contested a particular topic is (the term ‘oil spill’, for example, tends to have a low standard deviation since everyone agrees on its negative connotation).

Display Features. The vertical axis is rescaled automatically, a feature that is particularly useful if one keyword dominates the coverage and therefore obscures the distributions of the other keywords. Hovering above a data point displays the associated keywords and daily statistics (frequency, mean and standard deviation of sentiment). This mouse-over effect identifies topical trends and shows their impact on the individual peaks in the chart. Clicking on a data point triggers a search for this particular topic cluster in the preceding week.

Simple Search

admin — Sun, 05 Dec 2010 22:07:44 +0000

The search box is located in the top bar of the portal. The system supports the usual wildcard characters – while the asterisk (*) represents any number of unknown characters, the question mark (?) represents exactly one character. The syntax of additional options (source, location, date, etc.) is described in the advanced search section.

Display of Search Results

Once a user has entered a search query, the system lists all sentences containing the search term and groups them by document. The column headers show the publication date, the sentiment of the document, the currently selected and total number of matches, and the average sentiment of the displayed search results. Users can sort the results by their date of publication as well as the sentiment on both the document level and the sentence level (the color of the term reflects sentence-level sentiment).

Export Function

Clicking on their publication date selects and deselects the documents to be exported in either HTML or PDF format via the corresponding buttons in the upper right corner of the search results window. The average sentiment in the header is updated accordingly.

Impact on Other Views

A full-text query triggers not only the list of search results, but also an update of several other views including term associations, similar topics, keyword graph, various map visualizations as well as the content view which displays the summary of the highest-ranking document, including its source and target geography.

Advanced Search

admin — Sun, 05 Dec 2010 22:06:17 +0000

Advanced query options extend the capabilities of the simple search. They enable users to filter the search results by specifying restrictions based on document metadata (title, source, date, geographic location, etc.):

The title field of the documents can be queried using the title: operator; searching for “carbon dioxide title:ocean” will only return documents which contain “ocean” in their title.
The site: operator limits to search to a given domain; searching for “carbon dioxide site:cnn.com” returns documents on “carbon dioxide” mirrored from the CNN Web site.
Documents in a certain time interval can be searched by using the from: and to: operators; searching for “carbon dioxide from:2010-01-01 to:2010-01-31” will only return documents published in January 2010.
Source and target geography of the documents can be considered with the source: and target: operators; searching for “carbon dioxide source:US target:Kyoto” will return documents about carbon dioxide published in the US and referencing Kyoto as the main location in the text.

Ontology Graph

admin — Sat, 04 Dec 2010 10:39:31 +0000

The ontology graph displays a clickable domain model that matches documents and concepts to help users determine their current location in the information space. The ontology graph depicts hierarchical relations as arrows. The currently active document is highlighted by a yellow asterisk, and the letters [1-5] represent the classification of the five highest-ranking documents in the similar locations view. Clicking on a concept activates the highest-ranking document for this particular concept.

Information Landscape

admin — Sat, 04 Dec 2010 10:28:46 +0000

Information landscapes allow clustering and visualizing massive amounts of textual data. They implement the concept of ‘location’ in an innovative way that transcends the traditional geographic interpretation. The Information Landscape resembles a geographic map at first sight. Instead of geographic proximity, however, it represents semantic similarity between documents. At the time of map generation, its topography is determined by the content of the knowledge base. The peaks of the virtual landscape indicate abundant coverage on a particular topic, whereas valleys represent sparsely populated parts of the information space.

Interface Elements

Captions. The keywords for each peak are calculated automatically, based on the content of surrounding documents.
Document Markers. The markers [1-5] show the position of the documents from the Nearby Locations window.
Document Selection. Hovering above any location in the map shows a preview of the closest matching document in the active document window, clicking selects this document (each of the small gray dots, which become visible after zooming in, represents one document).

Geographic Map

admin — Sat, 04 Dec 2010 10:28:43 +0000

The geographic map shows the locations of documents based on analyzing their textual content (a process typically referred to as “geo-tagging”). The Active Document is highlighted by a yellow asterisk, and the letters [a-e] represent the five highest-ranking documents in the Similar Topics view. If users have an interest in a specific location rather than a topic, they can click anywhere on the map to activate the closest document (hovering above the map previews the document, but does not activate it).

After entering a search term, the set of results is visualized in the geographic map. Circular markers show the target geography of the found articles. The diameter of the marker represents the number of matching documents for a given location, its color the average sentiment of these documents. Using a color range from yellow to either green or red, trajectories link the source and target geography of an article (source geography = location of the publisher; target geography = main location referenced in the document). The little ‘+’ symbol on the right side of the window allows users choosing alternative base maps (e.g. NASA Blue Marble, Political Borders, Google Terrain, etc.), as well as deactivating the circular markers or trajectories.