You always see GEODI with the motto of Search-Discover-Classify-Mask. Data discovery generally comes into play in relation to regulations such as GDPR or PCI/DSS. GEODI has much more than the data discovery required for regulations. Here, we will delve a little deeper into the capabilities of discovery.
Discovering relationships between people, dates, and places is critical information in many professions and businesses. Who appears together in which documents, who is present in the same photograph, which place names are associated with a person, who are the individuals related to a specific word, place, or date are just a few analyses that can be easily performed with GEODI tools.
It will benefit lawyers, press consultants, independent auditors, inspectors, experts, managers, or anyone involved in digital forensics. The tools are easy to use, and the results are effective.
We've mentioned names of people, places, or dates, but what's written applies to everything discovered by GEODI data discovery tools.
What do we mean by Relationship?
GEODI discovery tools recognize many entities in the data such as names, place names, addresses, ID numbers, credit card numbers, company names, or various entities that users can define. These recognized entities are summaries of the content. We call them 'words' in GEODI.
Person-Person may be in the same document. This is a relationship.
If person names appear in very different parts of the same document, it may not be a relationship, but if they are close, we can call it a relationship.
If the same name appears in different documents, the documents can be related.
!!!RESİM!!!
Which Data?
GEODI covers over 200 data sources, leaving no digital data behind. Documents scanned with OCR are also processed. Databases, Word, Excel, Open Office files, emails, UYAP files, videos, and images are all included.
So your invoices, reports, proposal documents, project files, specifications, contracts, and more can all be processed by GEODI without any hassle. It's always possible to add new sources not listed in the list.
GEODI Data Sources - geodi-en - Confluence (atlassian.net)
Proximity of Two Names
When we write Name1 Name2, it gives us the content where both names appear. However, these names don't need to be close together. Name1 could be on the first page, and Name2 could be far away, on the 50th page. In this case, both doc1 and doc2 will be returned.
GEODI solves this issue with a special search operator (proximity). See: GEODI Query Rules - GEODI - SUPPORT (decesoftware.com)
When you write Name1() Name2(), it means the names should be within a maximum of 10 words apart. Only doc1 will match your query. If 10 words are not suitable, you can specify the proximity like Name1(5) Name2(5) or Name1(30) Name2(30).
You should determine the distance value according to your intuition. The 10-word distance is chosen based on the average paragraph size and is the default value.
Proximity of a Name and a Place Name
The rule is exactly the same.
Simply write Name1() CityName1().
Alternative spellings like Name1() [CityName]*() or Name1() [CityName]~() are also possible. * covers variations like "from [CityName]" or "in [CityName]", while ~ covers misspellings. You can find more details on the GEODI Queries - GEODI - SUPPORT GEODI Query Rules - GEODI - SUPPORT (decesoftware.com) page.
Adding Additional Criteria
In GEODI queries, you can apply multiple rules simultaneously.
For example, if you write name1() name(2) economy, you'll find content where two names are close and contain the word "economy". You can add as many criteria as you like.
For example, name1() name(2) economy January-March 2023 will bring content containing "economy" and dates between January and March 2023. As you know, thanks to semantic features, GEODI will find the correct format for dates regardless of how they are written (different languages, formats, alphabets, etc.). You can find details on the GEODI Queries - GEODI - GEODI Query Rules - GEODI - SUPPORT (decesoftware.com) page.
Different Views of the Query
So far, we have used the Contents tab. We queried the proximity of two names or a name and a place name. Queries contain much more information than just listing content. For this purpose, GEODI also offers different view options. Words, Calendar, Network Graph, or Map provide us with much more information.
Words Page
Let's consider the query Name1() Name(2). The Words page lists names, dates, and other information close to these two names.
If you only want to see close dates, select dates in the facet area. If you want both dates and places, you can select both. If there is no selection, all discovered words will be displayed.
The Words page also allows you to query the proximity to a single entity. So, you can simply write Name1(). In this case, you will get dates close to the name you provided. This query also affects the Map, Calendar, and Network Graph. Contents are not affected, so the result of Name1 or Name1() in Contents is the same.
Calendar Page
The Calendar page is automatically generated from the date information discovered in the texts. When any query is made, the dates from the contents that match the result are displayed.
If you search for Economy you will get all the dates from documents containing the word "Economy." However, if you search for Economy(), the situation changes slightly; this time, only dates close to the word "Ekonomi" will remain. The dates listed on the Words page now appear as a calendar. The frequency of dates over time is also additional information.
Takvim üzerinden istediğiniz bir tarihe tıklayın ve aradığınız belgeye ulaşın.
Harita Sayfası
In GEODI, some words also have coordinates. These words can be parcels, place names, or area codes with a phone number. We see them on the map. We'll call these words "Location".
Just like in the calendar or words page, when you type Economy, we'll see the locations in documents where the economy is mentioned. When you type Economy(), you'll see locations close to the word Economy().
With Economy(), you can understand in which cities the economy is prominent in the news.
With drilling(), you'll see the drillings in the archive of a construction company, and with division(), you'll see parcels related to division in the documents of a municipality.
For more information about the Map page, please refer to the How to Use the Map Page guide.
Network Graph Page
The Network Graph is part of the GEODI Discovery Module. With the network graph, word-word, word-document, or document-word relationships are visualized.
There's a lot of detail here, please refer to the guide on how to use the network graph page for more information.
FacePro - Face Recognition
The GEODI FacePro application recognizes faces in photos and videos. You can perform queries similar to those described above for people who appear in the same photo.