In this section of the guide, you can find in depth information about the Butterflyzer features and how to use them.
Want to know more about how the Outline and other Views work together? How to work with Tables? Explore the Butterflyzer User Interface, including the Browser, Graph View and more.
Find out how to work with Butterflyzer more efficiently and effectively. Discover capabilities that you didn't know existed. Wondering what that Flower Icon is for? Or how you can make the Graph less busy? We answer those questions here.
What's Focus for? What happens when I collect a Web Page? What are all of the things I can do with Tweets? Find out about the actions that you can take on Butterflyzer items.
This reference section provides a complete set of definitions for "almost everything" in Butterflyzer. It also provides a good overview of the overall data structure for Butterflyzer catalogs. (For readers with a more technical bent, this includes object types used within Butterflyzer. If you'd like more information about technical details, please contact us.)
Author Author References Authors Catalog Collection Configuration Event Interest Query Issue Item Location References Locations Reference Collection Resource Search Term Selection History Item Semantic Collection Semantic Tag Semantic Type Set Tag Term Time Selection History Item Tweet Tweet References Tweets Twitter Session Twitter Sessions Web Page Web Pages Web Site Web Sites
Graphs are quite simple, really. They are a way of organizing and visualizing data that makes it possible to understand complex relationships between different things. As a simple example, a family tree is a kind of graph. And you've probably seen visualizations of social networks. But all information on the web is connected through graphs. Butterflyzer reveals those graph relationships so that you can work with them directly. Here are some basic terms to be familiar with:
A node in a graph is any object that is used to represent a particular piece of information. An example of a node would be a particular person within a family tree. In Butterflyzer, all nodes are Items.
A relationship is any connection that exists between any two nodes. (Technically, we often refer to them as "edges".) An example of a Relationship within a family tree is "mother". In Butterflyzer there are many different kinds of relationships. Here are just a few examples:
As you use the Butterflyzer graph tools, you'll become more familiar with the different kinds of relationships.
The configuration contains all of settings that determine how your catalog is setup. Because all of this information is stored in the
When selected, includes Google Web Search in search results. Please see the Terms of Service link to ensure that your usage meets those terms. Butterflyzer does not monitor your usage and you alone are responsible for ensuring your compliance.
External Sites: Google Web Search API Terms of Service.
When selected, includes Google News Search in search results. Please see the Terms of Service link for important legal information. Butterflyzer does not monitor your usage and you alone are responsible for compliance with the service provider's terms.
External Sites: Google Web Search API Terms of Service.
When selected, adds Topsy Search to web page results, finds related Tweets using Topsy, and adds Topsy ranking and influence metrics to Web Page and Tweet results. Please see the Terms of Service link for important legal information. Butterflyzer does not monitor your usage and you alone are responsible for compliance with the service provider's terms.
External Sites: Topsy Terms of Service.
When selected, includes Tweets obtained from Twitter in search results. Note that this can result in a large number of Tweets that you may or may not be interested in. When you are primarily interested in Web Pages, consider deselecting this option to avoid additional performance costs and storage requirements and avoid hitting Twitter API limit. Please see the Terms of Service link for important legal information. Butterflyzer does not monitor your usage and you alone are responsible for compliance with the service provider's terms.
External Sites: Twitter API Terms of Service.
When selected along with the Twitter Search Engine option, includes retrospective Tweets in the results. This option performs a separate search for each of that last seven days and can result in a large number of Tweets selected. Users who are only interested in current results or Web Pages should consider deselecting this option. Please see the Terms of Service link for important legal information. Butterflyzer does not monitor your usage and you alone are responsible for compliance with the service provider's terms.
External Sites: Twitter API Terms of Service.
When selected, enriches the results of any web searches with the powerful semantic tagging service provided by OpenCalais. Because this option collects a large number of tags, deselect it when you're only interested in web search results. Please see the Terms of Service link for important legal information. Butterflyzer does not monitor your usage and you alone are responsible for compliance with the service provider's terms.
External Sites: OpenCalais API Terms of Service.
Remove items that aren't of the selected Resource Type from the graph and other results. For example, if Item Type is set to Web Pages, selecting this option will remove Tweets and Authors from the results.
See also: Resource Type Menu.
Remove any items that don't have at least two connections to other items from the graph. This option is very useful for reducing clutter, especially when combined with the Filter Categories option.
Remove those items that don't have any influence recorded by the Topsy API. In order to be included, Web Pages must have at least one Tweet referring to them, and Tweets and Users must have an influence level greater than 0.
Remove any categories from the graph. Categories are collections, semantic types and other outline items that organize items. Selecting this item typically causes items and tags to no longer be organized in a hub and spoke style. This option is most useful for the Spring layout, and is quite powerful when combined with the Filter Cliques option. It doesn't work as well for Radial and Tree layouts, as it often causes items to appear on top of one another.
Remove any Retweets from the results. This option is useful for narrowing down the set of Tweets to only those Tweets that carry unique information. On the other hand, if you want to gauge overall interest in a topic you should deselect this option.
Combines the selected items according to a logical rule. An Or selection contains those members (web page, tweet, etc..) related to any of the focused items. This is a logical "OR" search." An And selection contains only those members related to every focused items. This is a logical "AND" search. A Not selection contains no members. This is a logical "NOT" search. This might be useful if you want to make changes in the Outline View without affecting the other views.
Combines selected items so that matching items are displayed only for those selected items that are related to all selected items or tags.
Ignore Twitter User followers when searching for related users. Note that if a Twitter user would be included for another reason -- for example, if a user was also following the related user -- that user will still be included in the results.
Ignore other Twitter Users following the User when searching for related users. Note that if a Twitter user would be included for another reason -- for example, if the user was also being followed by a related user -- that user will still be included in the results.
Ignores Mutual (two-way) relationships for Twitter Users when searching for related users.
Limit the number of related items. This option is useful when you have a large number of results and want to explore them while avoiding performance issues. After discovering the set of results that are most relevant, you can then deselect this option to see all related results.
Specifies how deeply relations between objects are searched. When an object is focused (for example, selected in the outline view) related items are shown in the table, graph and browser as appropriate. The depth defines how those related items are selected. For example, if Mark Twain has a 'book' relationship to Tom Sawyer, and Tom Sawyer has a 'friendship' relationship to Huckleberry Finn, then if Depth 1 were selected, only Tom Sawyer would appear in the graph. If Depth 2 were selected, both Tom Sawyer and Huckleberry Finn would appear. (One way to think about this is as degrees of separation in "Kevin Bacon Space". The idea is that you are looking for the number of connections between "friends of friends" that you would have to go through to contact the actor.)
Sets the relationship depth to 1. This option works well for quickly examining relationships between tags and items and other simple results.
Sets the relationship depth to 2. This option is often a good compromise between richness of relationships, useful results and good performance.
Sets the relationship depth to 3. This option can result in a large number of results and can result in reduced performance.
Include all relationships in the results regardless of distance from the focused items or tags. This means that any items connected to the focused items will appear in the selection. This option only makes sense to use when relations are very sparse. It can return a very large number of results it can result in poor performance.
Defines how items are placed on the graph. Different layouts are appropriate for different situations. For example, when looking at hierarchical relationships such as for a web site, a horizontal tree might be appropriate, and when looking at complex relationships between many different kinds of entities, a spring layout might be appropriate.
Graph members organize themselves so that they are closest to those other items with which they share the most edges. The graph continues to update automatically so that over time the graph structure will become more organized.
Graph members are arranged in a vertical tree format. Note that this can cause some items to appear off the screen.
The primary resource of interest. The selected type defines what kinds of resources are shown in the table view and can be used to filter out other types. When combined with the Filter Resource, graphs will show just the selected resource type.
Selects Authors (Users) as the primary interest type.
See also: Filter Other Resources.
Display information about where and when the item was collected in browser and caption views.
When selected (default), items are shown in the Graph View. Hide Items by deselecting this option. This allows you to focus on the relationships between terms and concepts embedded within Web Pages and Tweets. Hiding items and connections (lines) greatly reduces screen clutter, but the underlying layout is still at work organizing the information. This unique visualization combines the clarity of a cloud layout approach while gathering related concepts together. Hiding items is most effective when combined with the Spring Layout.
See also: Spring Layout.
When selected (default), tags are shown in the Graph View. Hide Tags by deselecting this option. This shows related tags clustered together, allowing you to focus on the relationships between Tags without showing the related resources. Hiding items and connections (lines) greatly reduces screen clutter, but the underlying layout is still at work organizing the information. This unique visualization combines the clarity of a cloud layout approach while gathering related concepts together. Hiding items is most effective when combined with the Spring Layout.
See also: Spring Layout.
When selected (default) text for Resource Items are displayed in the graph. For example, a Web Page node will display the title of the web page, and a Tweet node will display the text of the Tweet. When unselected, the nodes will be displayed as filled circles with their size defined by the scale settings.
See also: Scale Related.
When selected (default) text for Tags are displayed in the graph. For example, the tags for "Apple" and "IBM" will be shown. When unselected, the nodes will be displayed as normal text.
See also: Scale Related.
Scales items displayed in the graph according to the number of related items that are connected to them. The scale can be adjusted in the graph settings dialog.
See also: Graph Settings.
Scales items displayed in the graph according to the Topsy ranking for those items. The scale can be adjusted in the graph settings dialog.
See also: Graph Settings.
Scales items displayed in the graph based on the number of items in a node. (This option is only relevant when items are grouped together. The scale can be adjusted in the graph settings dialog.
See also: Graph Settings.
Displays any web pages in a Reader mode. The reader attempts to simplify text and removes most images resulting in faster loading times and easier scanning. It works well with the Tile Browser Mode. Note that when using the Semantic Browser, web pages are actually processed and then stored locally. Occasionally there might be issues with processing web pages using the Reader. In that case, just unselect this option while browsing those pages.
See also: Tiles Browser Mode Menu Item.
Displays web pages with Semantic Tagging. (Default.) Semantic Tags show up as colored text on a black background. You can perform any available action on Tags within the Semantic Browser. Just hover over a tag and the Context Menu will be displayed.
Shows the Table View. See the Table View documentation for more details.
See also: Table View.
Shows the Graph. See the Graph View documentation for more details.
See also: Graph View.
Shows the Graph. See the Outline documentation for more details.
See also: Outline.
Defines how multiple browser items are displayed together. Choices include showing a single item with a list selector and displaying two kinds of sliding tile layouts.
Browser pages are selected from a list.
See also: Using the List Browser Mode.
Browser pages are sliding tiles.
See also: Using the Tiles Browser Mode.
All of your collected content is stored within Catalogs. There is one Catalog for every Butterflyzer document. Each Catalog appears in a single "Tab" within the Butterflyzer interface. See the
A Resource is a single content item that is stored within a Catalog. Resources are the things that you're interested in keeping track of, like Web Pages, Tweets, and content Authors. Generally resources can be uniquely identified.
An Item is simply any unit of information that is stored within a Butterflyzer Catalog, such as a Resource or a Tag.
A Twitter message of 140 characters or less. In addition to the message itself and the related Author / Twitter User data, Butterflyzer tracks other information about Twitter content, such as when it was collected, RT type, and so on. There are a number of actions that you can take with Tweets -- see the Tweet Actions section for more details.
A collection of Tweets.
Tweets items are connected to Tweet items.
A collection of Tweets for a particular Session. Tweets removed from the session will still exist in the Catalog.
Tweet References items are connected to Tweet items.
An Author is anyone who has created content. Currently all authors are assumed to be Tweet Users, as this is a convenient way to track content authors across Web Sites and Tweets. In addition to basic information like user name, Butterflyzer tracks other detailed information twitter provides such as Icon, Location and number of Tweets. Using the Collect User Network actions, you can explore and track complex a User's complete Social Network and then visualize the network using the Graph View. If the Topsy Search Engine is enabled, the User data will include topsy Influence rankings. By the way, "Author" and "Twitter User" mean pretty much the same thing, but future versions of Butterflyzer may support other content author types.
See also: Topsy Search Engine and Author Actions. External Sites: Topsy Influence Rankings. Author items are also Resource items. Author items are connected to Event, Location References, Author References and Tweet items.
A Web Page is a Resource on the internet with a specific URL. In addition to the URL itself, Butterflyzer tracks important details about the page, including when the page was collected, what Search Engine was used, if any, and so on. Deep relationships between other Web Pages, Tweets and content authors are also tracked. If you collect links for a page, those links are stored within the Catalog as well, enabling exploration of complex relationships between content sources. If the Topsy engine is enabled, Butterflyzer will also collect Tweets and Web Pages that mention the Web Page (Trackbacks). Topsy data is also used to track how many Tweets there have been on a particular Web Page, and how many of those were from Influential users. Of course, content for a page can change over time; currently Butterflyzer does not cache results, so a page might not match perfectly with Tags that were created on earlier searches. See the Web Page Actions section for more on what you can do with pages.
See also: Web Page Actions and Topsy Search Engine. Web Site items are also Web Page items. Web Page items are also Resource and Event items. Web Page items are connected to Issue, Search Term, Author, Tweet and Selection History Item items.
Contains all content authors (users) collected within the Catalog.
Authors items are connected to Author items.
Contains Authors related to the parent item. An Author removed from References will not be removed from the Catalog unless it is deleted.
Author References items are connected to Author items.
An Issue represents an area of concern. (Issues are not currently used within Catalogs but they will be supported in future versions.)
A Term is anything that we might want to track within a Catalog.
A Search Term is any Item that can be the subject of a generic search. Not all Search Terms are Tags; Search Terms can also be made up of of a set of Tags such as in the case of Sets.
Semantic Tag, Tag and Set items are also Search Term items. Search Term items are also Item and Term items. Search Term items are connected to Event, Tweet, Selection History Item and Interest Query items.
A collection of Semantic Types, which in turn contain Semantic Tags.
A Tag with well-defined Semantics. What this means is that the Tag is classified in a particular way, generally by Semantic Type. For example the Semantic Tag "Apple" might have the Semantic Type "Company". Note that different Semantic Tags may have the same value. For example, there might also exist a Semantic Tag "Apple" with the Semantic Type "Food".
External Sites: OpenCalais Entity Disambiguation. Semantic Tag items are also Search Term items. Semantic Tag items are connected to Event, Tweet, Selection History Item, Semantic Type and Interest Query items.
A Semantic Type is a collection of Semantic Tags that have the same type -- that is they share the same characteristics. For example, the Semantic Type "Food" might contain Semantic Tags like "Apple", "Corn Flakes", and "Steak Tartare".
Tags an items for the given keyword. Tags are typically user-created, but can also be collected by automatically. For example, Butterflyzer automatically creates tags for Tweet keywords.
Combines any number of collections according to a logical rule. Drag another collection into this collection to add it to the collections considered. An Or Set contains those members (web page, tweet, etc..) related to any of the focused items. This is a logical "OR" search."An And Set contains only those members related to every focused items. This is a logical "AND" search.
A Not Set excludes those members related to any focused items. This is a logical "NOT" search.
For example, if we created an "Or" Set that contained a collection of Flightless Birds and a collection of Sea Birds, that collection would include Ostriches, Gulls and Penguins. If we placed those same collections within an "And" Set, it would contain Penguins, but not Ostriches or Gulls. If we created a "Not" search containing the Flightless Birds, and included that in an "Or" Set containing Sear Birds, the collection would include Gulls, but not Ostriches or Penguins. Sets may be combined to create quite complex searches. For example, you might create an "And" Set that contains Flightless Birds and that also contains another "Or" set containing Sea Birds and Land Birds. The results would then include Ostriches and Penguins.
Collections are used to organize Terms within a Catalog. A collection can contain Terms and other collections.
Collection items are connected to Term items.
Collections are used for organizing a Catalog. A Reference Collection is made up of Terms and other collections that are also members of a non-reference collection. When items are removed (but not deleted) from a Reference Collection, they are not removed from the Catalog as a whole.
Interest Queries collect and store measurements of web user interest from Google Insights on Terms. To modify what is collected, open the Properties Editor. Note that you must be signed into a Google service from within Butterflyzer in order to collect Google Insights data. Please see the Terms of Service link for important legal information. Butterflyzer does not monitor your usage and you alone are responsible for compliance with the service provider's terms.
Web Sites are the primary locations (top-level domains) for Web Pages. For example, if you were looking at a NY Times article Web Page at http://nytimes.com/2010/11/01/ManBitesDog.html, the Web Site would be NY Times at http://nytimes.com. Web Sites are automatically gathered for any sites when a search is performed.
Butterflyzer stores Twitter searches in the contained sessions so that you can track collection activities. Twitter Sessions are also used for streaming. Sessions are created automatically when you perform another kind of search, and you can also create them manually with Add:Twitter Session. See the Twitter Streaming documentation for more details on how to start streaming.
Twitter searches are stored as sessions. Sessions are also used to define and execute Twitter Streams. If you've created a session manually, you'll need to define the terms for it. See the Twitter Streaming documentation for more details on how to start streaming.
See also: Begin Streaming. External Sites: Twitter API Terms of Service. Twitter Session items are connected to Reference Collection, Location References, Tweet References, Author References and Twitter Sessions items.
Contains all Web Sites collected by Butterflyzer.
Web Sites items are connected to Web Site items.
A collection of Web Pages within the Catalog.
Web Pages items are connected to Event items.
An Event represents a specific piece of information that is related to a specific time. For example, this might be an actual event, a blog entry or a news story that appeared on a particular date.
Selection History Items track user (in the sense of a Butterflyzer user, i.e. "you") interactions with Catalog Items. These are used to track navigation history, but they can also appear in the Graph, allowing the visualization of the Web Browsing process itself!
A selection that also specifies a particular time period for the selection. These can be used to constrain Tweets or Web Pages to just those that occurred within a particular date range. See the Select Time Span documentation for more information.