European Parliament divided as decision on text and data mining edges closer

11 Jul 2017 | News
Publishers, researchers and start-ups lobby lawmakers squabbling over the policy minutiae of a largely forbidden hi-tech research tool

Publishers, researchers and start-ups are fiercely competing for MEPs’ ears as the European Parliament edges towards a decision whether to ease rules on text and data mining before a vote in the autumn.

But lawmakers inside the European Parliament remain divided over the policy intricacies of opening up the use of the sophisticated computer-combing technique, disagreeing about who should be allowed do it, and for what purpose. The policy fight is one front in a huge battle to update existing EU copyright rules, which date back to 2001.

Text and data mining turbocharges research by enabling a scientist to simultaneously scour hundreds and thousands of articles. The problem is that the material researchers want to mine is tightly controlled by journal publishers in Europe, which own the copyright. A scientist wanting to data mine papers about malaria would need to seek permission for 1,024 science journals, for example.

Researchers feel a lot of ground has been lost, with the use of mining in Europe significantly lower than in the US and Asia, due to current limitations imposed by European rules.

Last September the Commission proposed to scrap the default ban on computer scanning of tens of thousands of papers and wants to pass legislation before the end of October this year. But many consider the timetable too ambitious, with significant horse-trading between the Parliament and member states expected over the details.

“There is no breakthrough so far,” said a Commission official. “It’s particularly complicated because there’s a strong political flavour to the whole dossier.”

The proposal would improve the situation for mining in Europe overall, but not for everyone. Universities and research institutes operating “in the public interest” stand to gain in the new rules, but start-ups, citizen scientists, freelance researchers, data journalists and advocacy groups are excluded. The proposal also limits the scope of mining activity to “purposes of scientific research.”

In the Parliament, some MEPs are objecting to the narrow parameters set by the Commission, and are plotting to open up mining to a wider pool of users. Another set of MEPs is holding out against plans to broaden the exemption beyond research institutes, a position publishers back.

“The Parliament is pretty unpredictable concerning this file, as there are different opinions even within parties,” said Marie Timmermann, EU legislation and regulatory affairs officer at Science Europe, which represents public researchers funders.  

Committee positions

MEPs involved in nudging the Commission proposal into law differ on key points. German MEP and vice-chair of the Green group Julia Reda argues that excluding independent researchers, journalists, hobbyists and companies from new mining rights would have “a chilling effect on discoveries in the public interest”.

Reda has been the most consistent MEP lobbying for broader mining rights. “Restricting the scope to research institutions would deny European research- and AI-based companies a level playing field on the international market, harming EU competitiveness in these highly promising sectors,” she said.

In line with Reda’s position, the Parliament’s research committee, ITRE, on Monday voted on a proposal drafted by Polish MEP Zdzisław Krasnodębski, for a far broader mining exception, covering research institutes but also journalists, companies and citizens.

“MEP Krasnodębski seems to have understood what’s at stake and is looking for a balanced way forward,” said Lenard Koschwitz, director of European affairs at the industry group, Allied for Startups.

One feature of the Commission’s proposal is that science journals should retain the power to limit text and data mining if there is a risk to the “integrity” or “security” of the materials. Research groups have expressed the concern that publishers could unfairly use this provision to block mining.

The ITRE proposal, however, states that any measures publishers adopt to ensure the security of their systems “should not prevent or exclude the ability to develop text and data mining tools different from those offered by the right holder.”

Elsewhere, MEPs in the internal market and consumer protection committee have voted for a small extension to the scope of the Commission’s proposal, but also to restrict the amount of content researchers can mine.

The culture and education committee, which will voted on a position Monday, proposes to add extra restrictions. For example, it suggests that miners should delete much of the data collected from mining. Researchers regard this as nonsensical, since keeping the underlying data ensures that results can be replicated and verified. The proposal also suggests compensating rights holders for mining uses.

MEPs tugged from all sides

Meanwhile, advice is pouring into the Parliament from start-ups, researchers and publishers.

A letter signed in January by five research lobby groups, which together represent hundreds of universities, libraries, funders and research bodies, called on legislators to include start-ups among the beneficiaries of the new mining regime.

Start-ups argue that this option is essential to ensure a competitive market. An alliance of data companies was set up to further press their case.

Without a change of the rules, “it will be harder for text and data start-ups to grow and raise funds from investors,” says Koschwitz. “Before you invest in a company, you want absolute certainty that all mining is legal – but it is hard to ensure this under current rules.”

On the other side, publishers broadly favour the Commission’s proposal, saying that any exemption should be limited to research organisations. Companies that want to mine material should continue to pay for a licence, they argue. 

Stephen Lotinga, Chief Executive of the Publishers’ Association said, “We are concerned that the current proposals from the EU to provide a broader exception would enable very large companies such as Google, multinational pharmaceutical companies and hedge funds to unfairly avoid paying their way and undermine the nascent commercial market for text and data mining.”

Alienating publishers

The committees voting on the file will try to negotiate compromises that enjoy majority support. The legal affairs committee, as the lead penholder on the file, will adopt the Parliament’s final position. The Commission wants to cut a deal while Estonia, a country seen as an influential leader on all things tech, holds the EU presidency.

It is not clear which amendments will survive until the final assessment, but there is a feeling, particularly among officials in countries with large publishing houses, that the deal should not radically disrupt the status quo. “We need a balance between making more content available but there needs to be something big in it for the rights holders,” said an EU official. “We do not want to significantly change their situation or alienate them.”

Koschwitz said he is doubtful whether, “There is enough momentum or political will to move the needle on text and data mining. The whole dossier is extremely controversial.”

The conservative European People’s Party, the largest bloc in the Parliament and the political home of the lead rapporteur on the copyright file, German Axel Voss, is not in favour of widening the scope beyond the Commission’s proposal, arguing that this would interfere with “a well-functioning licensing market.”

But Alea López de San Román, policy officer with the League of European Research Universities, which has campaigned for a liberal regime for text and data mining said, “This licensing market is only functioning well to the huge benefit of publishers.”

Never miss an update from Science|Business:   Newsletter sign-up