Finer Options for the Best Data Maintenance

The alternative is the use of APIs, more complex to set up, they also need to structure the data, the API acts as a communication protocol between the server holding the data and the end user who can perform queries in a language specific to the API. It is then much easier to control reading access, or even in some cases writing. Fir the 먹튀검증 this happens to be a very important matter.

This upstream study must also be that of the establishment of governance: disseminate information, even internally, is equivalent to taking the risk of finding them externally (malice, inattention). It is unthinkable (and illegal with the new RGPD regulation) to release strategic or personal data. Issues of format and quality also need to be addressed. In the context of setting up a datalake, the formats can be multiplied (text, video) without any treatment. But as part of a broadcast for reporting purposes or steering, a pre-treatment harmonization and cleaning will be necessary. In any case, the format must be suitable for large-scale industrialized use. For example, the High Authority for Transparency disseminates its data as a pdf, rendering it virtually unworkable.

The provision of data must be progressive, in order to allow the organization to adapt to the changes of culture and process that it entails (goal ultimately sought).

The typical type of a project

As a result of this upstream study, the project itself can be divided into three stages. As a first step, you need to release your data internally, in order to start spreading a culture centered on use and quality. It is also an opportunity to identify the most useful sources and to start POC (Proof of concept). Once again, security must not be neglected. This step cannot constitute, in any case, a test to determine what is strategic and what is not. Governance must be upstream as we have seen previously.

In a second step, the data can be left to external partners as part of joint projects. This step ensures that the data is understandable and usable by ‘outsiders’ i.e. non-business actors, or even the company. This will be the time to make some adjustments: metadata, accessibility of portals, usability of APIs, performance, and so on.

It is only in a third time that the data can be released in the form of open data to the public

These three steps require a progressive refinement in the understanding of the needs and data possessed internally. Each of them must have been individually carried out: the objective is to have identified the projects with strong ROI that could be carried out internally before making the data public, but also to refine the quantity made available. The entire opening project must therefore be done jointly with several entities of the company to ensure that no opportunity has been missed.

Recommended For You

About the Author: David Curry

Leave a Reply

Your email address will not be published. Required fields are marked *