What's New in I2E?

Introducing I2E 5.6

In addition to the changes described below, there are the usual bug fixes and under-the-hood improvements.

Charting 3

Version 3 of Charting is included with I2E 5.6 (see figure 1) and improves the usability of the Chart output from I2E.

Figure 1 Charts and output are laid out with better proportions in I2E 5.6, so that more results can be seen on the page

Charts have been rotated and the text size increased to make them easier to read. They have been moved to the left of the page to move results higher up the page. Charts now display the first few items (rather than filling the space available) and hide rows in the charts that, after filtering, match zero documents.

Querying Changes

Prioritization of items in alternatives

I2E 5.6 has introduced a new way to prioritize multiple query terms within a search, so that results are streamlined and easier to review. Alternatives with Prioritization of children set to Within container or Same span (figure 2) will now prefer matches from the top child and only try to match child further down the list if it hits nothing before.

Figure 2 New Prioritization of children settings allow you to rank the items and find the best hits, not all the hits

This is a very powerful new feature in I2E with a lot of potential use cases. For example, you may want to tag a document as being related to Humans and only tag a document as being related to Animals if Humans are not mentioned. Or you may want to prioritize one ontology (e.g. Linguamatics Diseases) but ensure that you capture all of the missed hits by using other ontologies (e.g. MeSH, Orphanet, MedDRA).

Resources Updates

As part of our continual multi-lingual improvements within I2E, version 5.6 will now treat the dot symbol as a thousands separator in European languages and find numerical ranges in German, French, Italian, Spanish and Dutch (as well as English). These ranges can be found in the Numerics, Measurements and Duration & Frequency pattern ontologies.

The Mutations ontology has been improved to match additional constructions and more consistent identify and normalization different substitution patterns.

As part of our quarterly ontology update, sources for ChEBI, CPC, Diseases, Entrez Gene, HPO, NCI, MedDRA and RxNorm have been updated to later versions.

I2E OnDemand: Index Updates

There is also a new Document Identifier shadow region, which is included in all of the I2E OnDemand indexes. Queries that extract the identifier from your documents can run against any OnDemand index: simply replace the existing source-specific region (e.g. PMID) with the new Document Identifier shadow region.

Better Management of Large Indexes and Restrictions

Index Restrictions were introduced as an API-only feature in I2E 5.4.1. In I2E 5.6, they have been made accessible via the I2E Administration view. This, along with the introduction of the Document Identifier shadow region, greatly simplifies their creation and use.

Index Restrictions can be created, read and deleted via the I2E Administration view and both types -- include and exclude – are supported. Earlier versions of index restrictions could have lists of both documents to include and exclude, but this functionality has been removed and those older restrictions cannot be displayed.

These combinations of indexes and their restrictions are defined in virtual indexes written into files that live on the I2E server (as distinct from virtual indexes created within Index Trees). These file-based virtual indexes have traditionally been used within I2E to support command-line index management, but in I2E 5.6 they have been added to the I2E Administration view.

Virtual Indexes can be created in the I2E Administration view and, once displayed, Indexes and Published Indexes can be dragged into them. Index Restrictions can similarly be dragged in from their location in the tree.