- Print
- DarkLight
- PDF
iO Artificial intelligence
1. Artificial Intelligence
The Artificial Intelligence feature helps the user when classifying documents, via recommendations, which make it possible to detect duplicates of a document. The feature extracts the most relevant keywords from each document diagnosed, thus indicating content similarities.
Concretely, the Artificial Intelligence module makes it possible to:
- Find the latest version of a document.
- Find a semi-duplicate that can be used as a versioning aid.
- Find and suggest a filing folder, so the user doesn't have to search for a folder for document filing.
For a document to be considered a duplicate, its percentage of similarity must be complete (up to 100%). In the case of an almost complete rate (≈ 90%), it will be considered a quasi-duplicate.
When using the diagnostic module, three sections are offered:
- The section indicating duplicates and versions;
- The quasi-similarity section, which indicates the rate of similarity with other documents;
- The classification file recommendation section.
- The module only supports text documents.
- The change in format (e.g. Word -> PDF/A) creates differences in content between versions. An exact duplicate will then become a quasi-duplicate of the document.
2. Go to the ranking recommendation window
There are two ways to access the ranking recommendation window:
- The first way is diagnosis. To perform a diagnosis, it is necessary to consult the sheet of a classified document, and then select the diagnostic option from the left menu;
- The second way is that of the User document. When a document is placed in the desktop, a move button is provided to the right of the table. Once opened, the recommendation window will indicate the best choice of ranking folder, based on the relevance of the content.
3. Diagnostic sections
3.1 Version Indication Section
When the diagnosis is performed, the recommendation window will display in the foreground the list of duplicates to the analyzed document, as well as the different versions (minor and major) of the documents.
The first column of the list displays the current version of the document, the one that is currently in use. The second column indicates the most recent version of the document. This feature is particularly useful to ensure that the version of our document is the most recent, and if not, to find it.
3.2 Quasi-similarity section
This section of the recommendation window displays documents similar to the diagnosed document. A percentage of similarity indicates the degree of content indicates. A high percentage will therefore mean a high rate of resemblance between the two documents. A small percentage will indicate a significant difference in content between documents.
It is possible to overwrite the diagnosed document with a similar document with a high rate of similarity. To proceed with the overwriting, simply click on the desired similar document. A major version will later be created. It will appear both in the duplicates and versions section, and in the version history of the document.
3.3 Classification File Recommendation Section
The file recommendation section is used to identify a relevant file for classification. The module will analyze the contents of the other files, then it will evaluate, according to the rate of similarity, the ideal record of the document.
The module will not indicate other relevant records if the document is already filed in an ideal similarity folder. In this case, the name of the folder used will appear in the recommendations view.