With Information Management (IM), users can manage their research findings to reduce complexity and organize information. Derived from ontologies concept, we provide knowledge graphs for knowledge representations with nodes to represent concepts and edges to represent relations between the concepts. The management enables users to edit, delete, or add new concepts and relations so that users can create their own concepts and environments.

One of the novelties in our information management is that we implement compound graphs for adding another abstraction layer to model concepts with generalized hierarchies, in order to reduce the cognitive load of the user. So the user could get better and faster understanding of the concepts.

autolinks prototype can be tested here: http://ltdemos.informatik.uni-hamburg.de:8093

Note that the server is not secure and we do not take responsibility for your data, this server is for research and demo purposes only!

This prototype is still in demo / experimental version, so there are still limitations and restrictions in the application. Especially the graph results generated by the services, we are still working to develop the better semantic services.

User Manual

• Create a user and log in to the Autolinks server.
• Building a knowledge graph:
  • Click and hold, then select add icon to create a new node.
  • click a node to show a qTip, click link icon to draw a relational edge to other nodes.
  • click a node to show a qTip, click arrows icon to move to a parent node or merge to other single nodes.
  • 
  • click a node / an edge to show a qTip, click edit icon to open node details, to open editable metadata and sources.
  • To change nodes name, open metadata and add "label" as a key and the desired name as value, then click update.
  • 
  • Note: At the moment, only nodes created inside compound graphs provided by service that can be retrieved from query exploration. Nodes which are created outside service compound will vanish after reloading.
  • Note: For now, creating edges is only possible in one level of hierarchies. If the target node is in the different level of hierarchies, it will be duplicated to the same level of the source node.
  • Merging scenarios:
    • 1. single node / compound without edges to target compound -> move to target compound.
    • 2. single node without relation to single node without relation -> create new compound.
    • 3. single node with relation to single node with relation -> create new compound and duplicate nodes.
• Query Exploration:
  • Enable active semantic services in the main navigation.
  • Enable annotation search and case insensitive (OPTIONAL).
  • Type a query or queries in the search bar. Enter or click the search icon.
  • 
  • Knowledge graphs related to the queries are generated.

• Document Exploration:
  • Click the circle navigation in the lower-right side, and select upload icon.
  • Select a text document, select overwrite (OPTIONAL), then upload. Note: At the moment, we limit a file extention to (.TXT) only with 10 kB of size (1 - 10 paragraphs).
  • Since we use MER (Medical Entity Recognition), so you can try to find an article from the bio / medical domain, probably from wikipedia, and put 1 - 10 sentences from that article to your .txt file and you can see the entity recognition for the annotations in the text works better.
  • The document then will be uploaded, analyzed, and interpreted.
  • A new document will be shown in the document lists, the annotation types of the document will be extracted, and the document lense will be enabled.
  • 
  • Choose annotation types that you want to show annotations in the document lense.
• Annotation Exploration and New Annotation:
  • Once we have annotations in the document lense, we can start to explore annotation resource based on offsets.
  • Click one of annotation, and annotation resources node related to the offsets will be shown inside the annotation graph.
  • The annotation resource node will highlight the annotation offsets in the document lense if we hover it.
  • Hold Alt key to enable new annotation mode, and we can select texts to be a new annotation with a new / existing types, then release Alt Key to show add annotation modal. (Autolinks supports discontinuous annotation where we can annotate different offsets in the document lense)
• Provenances (Traceable sources):
  • Autolinks supports provenance where it acts like an external brain for users for the reason that users can remember what they have read and go back to it. It also enables the users to associate what they have created or edited with the original source of information.
  • To get provenance in a node, click a node to show qtip, and click edit icon to show node details, then click sources.