Ontology-driven Search Engine
Use the JACKALFISH® tool to perform targeted searches of disparate data sources using keywords, extracted phrases or paragraphs from target documents, common concepts, and/or ontology relationships. The tool's ontology-assisted text mining and natural language processing methods provide knowledge extraction from structured and unstructured text sources.
The JACKALFISH® tool is a web-based application that allows you to perform targeted semantic searches of disparate data sources using keywords, extracted phrases or paragraphs from target documents, common concepts, or defined ontology relationships. The JACKALFISH® tool applies ontology-assisted text mining and natural language processing methods in extracting knowledge from unstructured text sources.
In addition, the JACKALFISH® tool, using learning and self-adaptation mechanisms, reasons about the objects in the target domain and about relations among the search items, returning only those data artifacts or documents that are relevant to the search, including knowledge derived from predefined ontologies.
Users can enter a variety of search inputs:
- Keywords: Users can directly input the individual terms they’re looking for. Users also have the option of assigning a weight to each term, and to indicate whether the term “must,” “should,” or “must not” occur in the target data. Users can also specify exact phrase matching.
- Example Text: Users can also provide example text that discusses, in natural language, the concepts they’re searching for. This text may be as short as a single sentence or as long as an entire document or an entire directory of documents. The example text is analyzed by the query builder, which generates a weighted list of important tokens to search for.
- Ontologies: Users can enhance the search process through the use of ontologies, which identify terms of particular interest within a domain. Ontology models often provide information that associates context with a specific search and can be used to disambiguate terms and provide background knowledge that might help in interpreting content.
- Acronym Lists and Glossaries: Users can supply a set of acronym lists and/or glossaries to augment the search. Both these inputs provide additional information about specific important terms that may appear in the search inputs or in the target data. The query builder uses this information to augment the weights and content of the term list it generates.
Collaborative Information Analysis
Because information analysts seldom work in isolation, a shared understanding of analysts’ goals and the subsequent sharing of knowledge and effort can significantly improve analytic outcomes. The JACKALFISH® tool supports key collaboration mechanisms such as defining, sharing, and modifying a search tasks.
- Creating and sharing a search task: Users can create, view, and modify search tasks. Each task can be given a name, description, and an optional link to a SharePoint® discussion forum. The task can be shared with other users by defining it as either public (all users have access) or private (only specific users have access).
- Populating a search task with searches and results: The search results list display now includes a drop-down menu that allows users to assign a specific search (i.e., query) to one of their search tasks. Thumbs up and thumbs down buttons were also added next to each search result, allowing users to indicate which results are a good match to the query.
- Finding related search tasks: A “Find Similar Tasks” button now appears in the result list display that executes a search across the set of existing search tasks, returning those that closely match the user’s query. Pressing this button opens the search task result window, which displays details of the top matching search task, including results that have been “thumbed up” by other users for that task. When a SharePoint® installation is available, the name of the user who “thumbed up” the result is a hyperlink to that user’s SharePoint® profile page.
- Visualizing clusters of search tasks: The tool automatically groups all users’ search tasks into clusters of similar topics. A visualization of the relationships between these clusters and their most important terms is displayed in a graph visualization.
- Documenting and sharing search results: Users can copy-and-paste search results to support easy documentation and sharing of search results with source citations. A copy-to-clipboard button was added to provide a quick reference to the document or paragraph on the clipboard. If the search result being shown is a document match, a small clip of the first text in the document is copied to the clipboard. If the search result is a paragraph match, the matching paragraph is copied to the clipboard. In either case, the text that is copied will include a reference to the source document from which it came.