Have you ever been sued or had to sue someone? Have you ever had to appeal against an unfair decision in the workplace? Or simply challenge a traffic fine?

If you are living in Armenia there is a big chance you have. According to the judicial statistics, in 2016 alone more than 150,000 court cases were filed in a nation of just over three million people; this amounts to around one in every 16 adults.

With the number and complexity of cases increasing, the necessity for an understandable and predictable judicial system becomes acute. More so given the importance of equal access to justice and the development of effective, accountable and transparent institutions – both of which are key targets in the implementation of SDG 16 (to promote just, peaceful and inclusive societies). Achieving this goal requires a catalogue of issues be solved, not the least of which the question of making court practice uniform, predictable and understandable.

Unfortunately, services meant to support citizens in navigating the judicial maze are overloaded. One way to overcome this is to allocate additional resources into free legal aid programs, support universities in creating free legal aid projects (such as the innocence project), or encouraging lawyers to allocate more of their time to pro-bono services. But why not tackle the problem at its root? Why not just make the court database more understandable and searchable for a non-professional? Why not help parties identify patterns of judicial conduct themselves?

Here is where technology is helpful. To tackle these issues UNDP Armenia’s Innovation Unit, Kolba Lab, in partnership with the Union of IT Enterprises of Armenia, Masys Information Systems, the Armenian judicial department and with the financial support from the Slovak Ministry of Finance launched the DATALEX Smart Search project which makes it easier for the users of the Armenian Judicial Information System to search and understand the data. Since 2008, the Armenian judiciary has accumulated data on all court cases in a single system, called DATALEX. The Datalex system contains complete information on case movement and results.

However, despite being a tool with great potential – for identifying precedents, and working out what course of action you should take – the database search capabilities was hamstrung by a few key issues:

  1. Terminology: try searching for “stabbing”, “killing” or “borrowing money” in a court decision. You will end up with random collection of results. The reason is that legal language has its own terminology: “harming health with a sharp object”, “murder”, “loan” – unless you’re familiar with the terms used in judicial decisions, you won’t find them.
  2. Search results not sufficiently refined: The cases were not technically linked to the statutory provisions applied in them, thus making search less effective even if technical terms where used. For example, a search for “murder” would return results for a different crime – “threat of murder”.
  3. Absence of prioritization: The search doesn’t give priority to any case and instead returns cases in chronological order, while it is obvious that some cases carry more weight then others. Thus, a case decided by a higher court has a higher value and is more likely to be helpful to the user. Unfortunately, the search was unable to distinguish between them.

The totality of these factors turned searching for precedents into a difficult and time consuming task. To tackle the issue UNDP Armenia’s Innovation Unit, Kolba Lab, in partnership with the Union of IT Enterprises of Armenia, Masys Information Systems, the Armenian judicial department and with the financial support from the Slovak Ministry of Finance launched the DATALEX Smart Search project to address this challenge, the DATALEX Smart Search project prototyped a smart search of judicial decisions that supported users in three key ways:

  1. Word categories rather than technical terminology are used. This allows users to type in their search queries and choosing from suggested technical synonyms if necessary.
  2. Linking statutory provisions applied by the court, thus connecting searches to only those cases which actually contain a ruling on the same issue.
  3. Prioritizing the more important cases over the less important ones, allowing the user to easily access landmark high court decisions, rather than have to carefully look sift through massive amounts of rulings.

The end result of the project is a user-friendly search window, which quickly returns information on rulings based on relevance and significance, whilst allowing citizens to search in simple, terminology-free language.

Most excitingly, the potential for further development of the tool is tremendous. By further strengthening the transparency, predictability and accountability of the judiciary, this technology could eventually be used to help forecast future court decisions, saving both citizens and legal aid providers further time and money. Moreover, replicating the smart search system could potentially alter the mechanism of searching for data across all complex government databases in Armenia, allowing citizens faster and easier access not just to the important case-law of the courts, but to licensing or traffic fines databases, among others.

Watch this space!