Thomas Gossner

Conlator is a mobile application that translates any contract into an accessible easy-to-understand format. The user takes images of the contract, which are then analysed and extracted for relevant information. This content is then communicated back to the user in simplified language and with supporting visual aid.

Typically, we consult lawyers for contracts that are considered to include decision-making that is low in frequency but could potentially result in high impact, like for example, buying a house or getting married. In contrast, when it comes to contracts that are considered to be higher in frequency but result in lower impact - like insurances or terms and conditions of websites - we typically do not invest time and money in consulting a lawyer. A study by North Carolina State University found that the difficulty in understanding legal text led to 38% of the participants having signed contracts without ever reading them at all. It is no wonder that the UK’s Money Advice Service found that the lack of understanding of financial contracts alone costs the average adult in the UK 428 GBP every single year.

The reason for this goes beyond legal jargon. Where the language of  legalese often builds on long and complicated words that are specific to the legal realm, the ways contracts are presented to us additionally inhibits us from understanding contracts thoroughly. And when an employer hands us our employment contract, we have little choice but to accept it as it was provided by the drafter of the contract.

Conlator is designed to ensure that any contract is understood through three layers. In the first layer, Conlator facilitates optical character recognition and natural language processing (NLP) to digitalise and analyse paper-based contracts. Their content is then simplified using recent advancements in NLP and suitable visuals are matched to further increase understanding and engagement. Alongside the simplified version, the original contract is stored within Conlator and presented to the user if needed to ensure liability in the second layer. Finally, a direct connection to a lawyer can be made through Conlator, who can help the user to answer any further questions regarding the contract.

In the early experimentation (EXP01), various layout changes and visuals were tested on a standardised contract. Visuals showed to be highly effective in conveying the contract‘s contents quickly. To ensure the correct understanding, the related paragraph was aligned with the visual shown to prevent potential misunderstanding.

In further experimentation (EXP02) eight different standardised employment contracts were analysed, clustered, and a set of concurrent terms was extracted. The following terms were explained to participants with the prompt to visualise the content in the form of hand sketches. Based on previous experiments (EXP03), and the collected sketches, a visual database was set up and commonly used motives were extracted. These were rebuilt, tested, and refined in multiple iterations. 

Finally, the developed visuals were transferred back to a normal A4 contract layout (EXP04), to be tested and validated with multiple stakeholders. The refined visuals and the employment contract’s content were transferred to a commonly used A4 format. The aim was to allow for comparability between current contract formats. 

The case study on visualising the standardised employment contract resulted in representations for a set of paragraphs that were presented alongside the contract’s text. In tests comparing the adapted contract with the original contract, the retention rate after one day increased by 63%, and the time to find information decreased by 83%.