Lexum connects and extends the work of LegisPro and GPSL by providing governments with sophisticated tools for searching, analyzing, and smartly presenting legislation and regulations on the web. Earlier this year, we announced a partnership with the Montreal-based Lexum, which has been helping legislative and regulatory bodies disseminate their material over the web since 1993. In the interview below, we catch up with Lexum’s Pierre-Paul Lemyre to discuss our shared belief in open standards, web-enabled technologies, and the benefit of interconnections between platforms.
Hudson Hollister
Pierre-Paul, thank you so much for spending some time with us today.
Pierre-Paul Lemyre
My pleasure.
Hudson Hollister
I’d like to start by asking you about the story of Lexum. Then one of the concepts that we’re eager to describe and go into more detail with is the notion of legislation and regulation as Data First, the idea that data structure can make life easier for those who draft, those who enforce, and those who use legislative and regulatory text.
Pierre-Paul Lemyre
Sounds good. Lexum is an IT firm that specializes in the management and the publication of legal information. We’ve been doing it since 1992. At that time, we were a research lab at the University of Montreal. Essentially, a professor at the law school, who was our former president, approached the Supreme Court of Canada and told them, “There’s this new thing called the internet. We’d like to take your decisions and put them out there,” and they said yes. So we started from the University, and over the years, we added contracts with different courts in Canada. Then we did the Department of Justice of Canada’s legislative website for a period of about 10 years.
A major development was around the year 2000, when we convinced the Federation of Law Societies of Canada to invest in a project called CanLII, which has become the main point of access to pretty much any kind of primary legal information produced in Canada. We now have databases from courts, boards, and tribunals. For case law, I think we have like 320 different databases. We have the 14 legislative collection systems for the provinces and territories, and more and more secondary material as well.
So that was kind of a big change for us. Around 2009, we got out of the university and created the private company, which has developed since then. Our business model has been to basically take the technology and the approaches that we developed in the various projects we have been involved with and package that into products that we market to organizations with their own internal needs in terms of management and publishing of legal information.
Hudson Hollister
I would love to ask about the evolution of Lexum’s tech stack, driven by those factors, and about the evolution of culture in the organizations that maintained bodies of law and code contributing to your growth. Which do you think you want to go into first, the evolution of the tech stack and the technologies that are used to manage legislation, or the evolution of the culture?
Pierre-Paul Lemyre
Let’s tackle technology first. I’m not an IT guy, I come from a legal background, but I can give you a broad overview. Obviously, when we started the internet was relatively new. One of our colleagues at Cornell University, who was doing pretty much the same thing, developed a web browser specifically for the purpose of accessing legal information. So you see, it started pretty early! Everything needed to be developed from scratch. We adopted an approach based on open source software, and that never changed. I think we were on the right track from the start and that helped us keep the technology in line over time. We changed, for instance, our search technology, we switched probably four or five times over the years. But a lot of the components that we still work on were selected almost 20 years ago, and we’ve kept them because they work well. Broadly, and in conclusion, we’re a Linux shop, developing in Java, and using open source technology.
Hudson Hollister
Give us an idea of how the technology manages the lifecycle of legislation. Obviously, the lawyers and drafters who work for legislatures and government bodies might use a word processor to draft new legislation. Then drafts are considered and debated and amended in parliamentary bodies. And once approved, changes flow into the code.
We’d love to hear about how Lexum views that lifecycle — ultimately resulting in codification and publication which Lexum manages — and how the technology moves around that lifecycle.
Pierre-Paul Lemyre
Totally makes sense. From our point of view, we are focusing on publishing. We are really a company that comes at the end of the cycle. So just because of that, over the years, we’ve had to face a lot of different scenarios in terms of technology, because we’re working with many different bodies producing legislation. We work with courts, as well for case law. We also work with bar associations for CLE material.
All of them work with different environments. So for instance, in the early 2000s, WordPerfect was really the product of choice for writing in the legal sector. That sounds kind of strange now, but WordPerfect was the right tool at the time. So we developed a converter to move files from WordPerfect to HTML and PDF. And we still have that in place. We still support it because some of our clients are still working in WordPerfect.
Our approach over time was to broaden the scope of file formats that we support. So we now support pretty much anything from MS Word files, WordPerfect, OpenOffice, PDF, we can work with XML, or DITA, which is just a more precise format for XML. And obviously, the more structured the information is, the better it is. From our publishing perspective there are a lot of benefits tied to that. But because of the context, our approach has always been to be really flexible on the input format, to be able to cover different types of clients and different levels of sophistication when it comes to drafting their own material.
Hudson Hollister
The last technological question I have has to do with data structure. You mentioned that structured data makes the publication easier to handle. How has the rise of standardized formats, such as Akoma Ntoso, benefited your business and benefited your customers?
Pierre-Paul Lemyre
It definitely helps. But because of the situation we are in, we need to cover all angles. And obviously, the more it is standardized, the easier it gets for publishing because we can use common code. And we definitely support that. In relation to case law, we’ve pushed a lot for the adoption of neutral citations. It’s a thing that we’ve promoted for a long time, a standardized way of citing legal documents. It can apply to legislative information, as well. Because from one jurisdiction to another they can simply use a different way of referring to their own material. So just that helps a lot because you don’t have to parse tons of different formats. It simplifies the work and obviously it reduces the burden and the effort. Eventually, it ends up bringing cost savings for the clients because it’s easier to handle.
But because of the situation we are in at the end of the cycle, at Lexum, we still have this need of covering all angles. We still have clients working in WordPerfect, in Word, some are sending us Open Office documents. So there’s all sorts of situations. When we get a nicely formatted XML supported by a standard, that’s the best scenario because then we do not need to invest time and effort in customizations, we can just work from standardized code. So it’s definitely better.
Hudson Hollister
Excellent. Let’s ask about those clients and cultural change. What changes are you seeing in legislators around increasing awareness of the need for data structure and electronic publication and what is the impact of those cultural changes on your business?
Pierre-Paul Lemyre
I think the market is trending towards using products. It’s not just legislators, it’s the overall market of the software industry. We used to be in a situation where everything needed to be customized for each client. Everything used to be based on software deployed on premise and done at high cost. What we are seeing more and more are products that are doing a large portion of the job right out-of-the-box. Often they are available in the cloud, so that’s a big change from on premise. But moving from a custom-based approach to using products is probably the major change we’ve seen over the last five years. And that also helps a company like us that comes at the end of the cycle, because then we can work with partners and develop standard solutions applying to a large set of clients. So instead of building custom solutions for each client, we can support a limited number of systems for many clients. That means that in terms of features, we can go deeper and we can develop more sophisticated systems at a lower cost. I think that is the major development of the last few years in our industry.
Hudson Hollister
Very good. Following along with that development, as you know, our email newsletter is called Data First, and we look forward to the future in which regulation and legislation will not only be published on platforms in a standardized electronic format, but drafted in the same. We look for the adoption of standardized formats, such as Akoma Ntoso, from the headwaters of legislative drafting. Do you think that a Data First approach in which the drafting itself — drafting and amending and considerations — takes place in a software environment is possible? And if so, what trends do you see leading us in that direction?
Pierre-Paul Lemyre
It is definitely possible and companies like Xcential are at the forefront of that movement. It’s really the way to go.
In terms of benefits for publishing, which is what Lexum focuses on, the main one is that you get to save time and effort, as I mentioned before. One example, is that in many older systems there was a lot of copy and pasting of content or transcribing content manually between systems. When automated, publishing systems were interpreting the data coming from the drafting environment, which can generate some errors. So either human error or software error, and in any case it implies additional time and effort.
Among the benefits of structuring content right at the drafting stage, is the capacity of reusing the structure down the road. When it comes to a table of contents for instance, it means displaying that table of contents based on the structure which is pushed by the authoring platform to the publishing platform. This is something that we can do when documents come in Akoma Ntoso format, but not necessarily when it’s a MS Word document where everything is based on the same style.
It can help with the indexing, as well, to get better searching. So let’s think of just being able to identify the titles. Usually when you search, if a keyword is present in the title, this document should rank higher than if you find the keyword, let’s say, on page seven. So it helps for putting the better results at the top of the list, as well.
You can also automatically build links between documents if you have a uniform structure and you already know the applicable naming guidelines. In the past, we’ve developed citators for that purpose. It’s basically a database that makes the relationship between a legal citation and the URL of the corresponding document. But if you get all this information structured in the first place, you don’t need to develop a citator, you can simply use the raw data to create the links between the documents. So for example, if Section 7 makes a reference to Section 13, or even another Act, you don’t need a special database to create that link, you can make it point to the right location automatically.
I’d say the last benefit is for the drafters themselves, because when we keep the tracking between two systems that are structured, and you know exactly the path and where the data is, you can actually build back and forth links between the drafting solution and the publishing solution. So the drafters can navigate between both systems and simply use the published version to jump back to authoring and change something. And the back and forth between both solutions is pretty seamless.
Hudson Hollister
Do you think that the increasing adoption of a machine-readable format like Akoma Ntoso and Data First drafting will lead us to the possibility that eventually, legislation and regulation could carry electronic structure, allowing for automated compliance or machine-executable legislation and regulation?
Pierre-Paul Lemyre
I think it’s heading in that direction. We’re not there yet. But this is definitely the first step because if you don’t take that step, it’s not a reachable goal. Some at the moment are trying to work on artificial intelligence to reverse engineer this kind of knowledge from unstructured information. But if you draft it in a structured way in the first place, you don’t need to invest in solutions like artificial intelligence, because you can just work from what is there and build knowledge on top of the raw data. So yes. That’s definitely something in development and a step that we need to take to get there. And I think more and more jurisdictions are taking that step. That’s something that we see developing. It goes hand in hand with what was discussed earlier, the development of products. It cannot work if every jurisdiction uses its own approach and its own custom code. If every system is unique, it’s almost impossible to develop a standardized way of re-using the data. With standardization you can support large sets of vendors and softwares. So I think that’s definitely the way forward.
Hudson Hollister
Fantastic. As you know, our Data First audience is legislative and regulatory drafters and other professionals in the United States, Canada, and around the world. Many of the legislative and regulatory drafters today must face intense manual processes, as they draft amendments in prose, or as they compare the impact of proposed amendments. What encouragement might you offer to our Data First audience on how these technological changes and cultural changes might improve their lives, and those of the constituencies that they serve?
Pierre-Paul Lemyre
I think the move toward using products built on data standards will bring the technology to enable Data First authoring, drafting, and ultimately publishing. It will enable benefits for smaller jurisdictions potentially with less budget and staff, who are often over burdened by these manual tasks, because they could not invest in those large custom-made software solutions that used to be the only ones available 15 or 20 years ago.
I think what is changing at the moment is that even smaller jurisdictions have the capacity to adapt the right approach to draft in a structured format right from the start. And this will lead to more efficient publishing, and better publishing involving more features, as we’ve said earlier. I think what is changing right now is that the market of advanced legislative solutions is becoming more accessible for smaller jurisdictions, and that’s a good thing.
Hudson Hollister
Pierre-Paul, thank you so much for spending some time with us. I think our audience will be edified by the observations and the expertise that you brought. I hope we have a chance to do this again.
Pierre-Paul Lemyre
That would be great. Thank you.
