Recherche avancée

Sur d’autres sites (8494)

• ffmpeg failed to load audio file

  14 avril 2024, par Vaishnav Ghenge

  Failed to load audio: ffmpeg version 5.1.4-0+deb12u1 Copyright (c) Failed to load audio: ffmpeg version 5.1.4-0+deb12u1 Copyright (c) 2000-2023 the FFmpeg developers
  built with gcc 12 (Debian 12.2.0-14)
  configuration: --prefix=/usr --extra-version=0+deb12u1 --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --arch=amd64 --enable-gpl --disable-stripping --enable-gnutls --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libcodec2 --enable-libdav1d --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libglslang --enable-libgme --enable-libgsm --enable-libjack --enable-libmp3lame --enable-libmysofa --enable-libopenjpeg --enable-libopenmpt --enable-libopus --enable-libpulse --enable-librabbitmq --enable-librist --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libssh --enable-libsvtav1 --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx265 --enable-libxml2 --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-lv2 --enable-omx --enable-openal --enable-opencl --enable-opengl --enable-sdl2 --disable-sndio --enable-libjxl --enable-pocketsphinx --enable-librsvg --enable-libmfx --enable-libdc1394 --enable-libdrm --enable-libiec61883 --enable-chromaprint --enable-frei0r --enable-libx264 --enable-libplacebo --enable-librav1e --enable-shared
  libavutil      57. 28.100 / 57. 28.100
  libavcodec     59. 37.100 / 59. 37.100
  libavformat    59. 27.100 / 59. 27.100
  libavdevice    59.  7.100 / 59.  7.100
  libavfilter     8. 44.100 /  8. 44.100
  libswscale      6.  7.100 /  6.  7.100
  libswresample   4.  7.100 /  4.  7.100
  libpostproc    56.  6.100 / 56.  6.100
/tmp/tmpjlchcpdm.wav: Invalid data found when processing input


  


  backend :

  


  
@app.route("/transcribe", methods=["POST"])
def transcribe():
    # Check if audio file is present in the request
    if 'audio_file' not in request.files:
        return jsonify({"error": "No file part"}), 400
    
    audio_file = request.files.get('audio_file')

    # Check if audio_file is sent in files
    if not audio_file:
        return jsonify({"error": "`audio_file` is missing in request.files"}), 400

    # Check if the file is present
    if audio_file.filename == '':
        return jsonify({"error": "No selected file"}), 400

    # Save the file with a unique name
    filename = secure_filename(audio_file.filename)
    unique_filename = os.path.join("uploads", str(uuid.uuid4()) + '_' + filename)
    # audio_file.save(unique_filename)
    
    # Read the contents of the audio file
    contents = audio_file.read()

    max_file_size = 500 * 1024 * 1024
    if len(contents) > max_file_size:
        return jsonify({"error": "File is too large"}), 400

    # Check if the file extension suggests it's a WAV file
    if not filename.lower().endswith('.wav'):
        # Delete the file if it's not a WAV file
        os.remove(unique_filename)
        return jsonify({"error": "Only WAV files are supported"}), 400

    print(f"\033[92m{filename}\033[0m")

    # Call Celery task asynchronously
    result = transcribe_audio.delay(contents)

    return jsonify({
        "task_id": result.id,
        "status": "pending"
    })


@celery_app.task
def transcribe_audio(contents):
    # Transcribe the audio
    try:
        # Create a temporary file to save the audio data
        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_audio:
            temp_path = temp_audio.name
            temp_audio.write(contents)

            print(f"\033[92mFile temporary path: {temp_path}\033[0m")
            transcribe_start_time = time.time()

            # Transcribe the audio
            transcription = transcribe_with_whisper(temp_path)
            
            transcribe_end_time = time.time()
            print(f"\033[92mTranscripted text: {transcription}\033[0m")

            return transcription, transcribe_end_time - transcribe_start_time

    except Exception as e:
        print(f"\033[92mError: {e}\033[0m")
        return str(e)


  


  frontend :

  


      useEffect(() => {
        const init = () => {
            navigator.mediaDevices.getUserMedia({audio: true})
                .then((audioStream) => {
                    const recorder = new MediaRecorder(audioStream);

                    recorder.ondataavailable = e => {
                        if (e.data.size > 0) {
                            setChunks(prevChunks => [...prevChunks, e.data]);
                        }
                    };

                    recorder.onerror = (e) => {
                        console.log("error: ", e);
                    }

                    recorder.onstart = () => {
                        console.log("started");
                    }

                    recorder.start();

                    setStream(audioStream);
                    setRecorder(recorder);
                });
        }

        init();

        return () => {
            if (recorder && recorder.state === 'recording') {
                recorder.stop();
            }

            if (stream) {
                stream.getTracks().forEach(track => track.stop());
            }
        }
    }, []);

    useEffect(() => {
        // Send chunks of audio data to the backend at regular intervals
        const intervalId = setInterval(() => {
            if (recorder && recorder.state === 'recording') {
                recorder.requestData(); // Trigger data available event
            }
        }, 8000); // Adjust the interval as needed


        return () => {
            if (intervalId) {
                console.log("Interval cleared");
                clearInterval(intervalId);
            }
        };
    }, [recorder]);

    useEffect(() => {
        const processAudio = async () => {
            if (chunks.length > 0) {
                // Send the latest chunk to the server for transcription
                const latestChunk = chunks[chunks.length - 1];

                const audioBlob = new Blob([latestChunk]);
                convertBlobToAudioFile(audioBlob);
            }
        };

        void processAudio();
    }, [chunks]);

    const convertBlobToAudioFile = useCallback((blob: Blob) => {
        // Convert Blob to audio file (e.g., WAV)
        // This conversion may require using a third-party library or service
        // For example, you can use the MediaRecorder API to record audio in WAV format directly
        // Alternatively, you can use a library like recorderjs to perform the conversion
        // Here's a simplified example using recorderjs:

        const reader = new FileReader();
        reader.onload = () => {
            const audioBuffer = reader.result; // ArrayBuffer containing audio data

            // Send audioBuffer to Flask server or perform further processing
            sendAudioToFlask(audioBuffer as ArrayBuffer);
        };

        reader.readAsArrayBuffer(blob);
    }, []);

    const sendAudioToFlask = useCallback((audioBuffer: ArrayBuffer) => {
        const formData = new FormData();
        formData.append('audio_file', new Blob([audioBuffer]), `speech_audio.wav`);

        console.log(formData.get("audio_file"));

        fetch('http://34.87.75.138:8000/transcribe', {
            method: 'POST',
            body: formData
        })
            .then(response => response.json())
            .then((data: { task_id: string, status: string }) => {
                pendingTaskIdsRef.current.push(data.task_id);
            })
            .catch(error => {
                console.error('Error sending audio to Flask server:', error);
            });
    }, []);


  


  I was trying to pass the audio from frontend to whisper model which is in flask app

  


• The Guide to an Ethical Web : With Big Data Comes Big Responsibility

  13 mars, par Alex Carmona

  Roughly two-thirds of Earth’s 8 billion people use the internet for communication, education, entertainment, business and more. We are connected globally in ways previous generations could’ve never dreamed of. It’s been a wild ride, and we’re just starting.

  Many users have learned that experiences online can be a mix of good and bad. Sometimes, the bad can feel like it outweighs the good, particularly when large tech companies use our data shadily, cut corners on accessibility or act in any other way that devalues the human being behind the screen.

  As fellow internet citizens, what responsibility do we have to create a more ethical web for our customers ?

  In this article, we’ll look at ethical principles online and how to act (and not act) to build trust, reach customers regardless of ability, safeguard privacy and stay compliant while improving business outcomes.

  

  What is an “ethical web” ?

  When we talk about the ethical web, we’re talking about the use of the internet in an ethical way. Among other values, it involves transparency, consent and restraint. It applies the Golden Rule to the internet : Treat others (and their data and user experience) how you’d want yourself (and yours) to be treated. 

  With limited oversight, the internet has evolved in ways that often prioritise profit over user rights. While selling data or pushing cookies might seem logical in this context, they can undermine trust and reputation. And the tide is slowly but surely shifting as consumers and legislators push back.

  Consumers no longer want to buy from companies that will use their data in ways they don’t agree to. In 2022, 75% of UK and US consumers surveyed said they were uncomfortable purchasing from businesses with weak data ethics.

  Legislators worldwide have been taking part in this effort for nearly a decade, with laws like GDPR in the EU and LGPD in Brazil, as well as the various state laws in the US, like California’s CCPA and Virginia’s VCDPA. 

  Even tech giants are no longer above the law, like Meta, which was fined over a billion Euros for GDPR violations in 2023.

  

  These changes may make the internet feel less business-friendly at first glance, but ethical choices ultimately build a stronger digital ecosystem for both companies and consumers. 

  Likewise, all internet users alike can make this happen by shunning short-term profit and convenience for healthier, long-term choices and behaviour.

  As we dig into what it takes to build an ethical web, remember that no company or individual is free from mistakes in these areas nor is it an overnight fix. Progress is made one click at a time.

  Ethical SEO : Optimising your content and your ethics

  Content creation and search engine optimisation (SEO) require so much work that it’s hard to fault creators for not always abiding by search engine guidelines and seeking shortcuts – especially when there’s a sea of LinkedIn posts about how copying/pasting ChatGPT responses helped someone rank #1 for several keywords in one week.

  However, users turn to Google and other search engines for something of substance that will guide or entertain them.

  Content meets customer needs and is more likely to lead to sales when it’s well-written, original and optimised just enough to make it easier to find on the first page of results. This doesn’t happen when content teams dilute quality and waste a reader or viewer’s time on posts that will only yield a higher bounce rate.

  Some SEO pros do find success by building backlinks through private blog networks or crafting a million unedited posts with generative AI, but it’s short-lived. Google and other search engines always catch up, and their content plummets or gets penalised and delisted with every new update.

  Content teams can still rank at the top while sticking to ethical SEO principles. Here’s a sample list of dos and don’ts to get started :

  • Do put content quality above all else. Make content that serves the audience, not just a brand or partner ad network.
  • Do apply the E-E-A-T framework. Search engines value content written by authors who bring expertise, experience, authority and trust (E-E-A-T).
  • Don’t keyword stuff. This might have worked in the early days of SEO, but it hurts readability and now harms article performance.
  • Do use alt text as intended. While it can still help SEO, alt text should prioritise accessibility for users with screen readers.

  • Don’t steal content. Whether it’s violating copyright, copying/pasting other people’s content or simply paraphrasing without citation, companies should never steal content.
  • Don’t steal ideas. It’s okay to join in on a current conversation or trends in an industry, but content creators should be sure they have something valuable to add.
  • Do use AI tools as partners, not creators. AI can be an incredible aid in crafting content, but it should never be posted without a human’s touch.

  When we follow ethical SEO guidelines and get more clients with our content, how do we best handle their data ?

  Ethical data governance : Important principles and how to avoid data misuse

  Data governance comprises every aspect of how a company manages data, including storage, security, privacy, lifecycle management, setting policies and maintaining compliance with laws like GDPR and HIPAA.

  Applying data ethics to governance is doing it all in a transparent, restrained way that acknowledges an individual’s right to ownership over their data. 

  For organisations, this translates to getting consent to collect data and clearly spelling out how it will be stored and used — and sticking to it.

  If a user’s birth date is needed for legal reasons, it cannot be sold to a third party or later used for something else without explicit permission. Reusing data in ways that stray from its original purpose is a form of commingling, one of the data misuses that is easy for even well-intentioned teams to do accidentally.

  Ethical data governance also includes the vigilant safeguarding of users’ data and minimising potential privacy issues.

  Failing to implement and adhere to strong security measures leads to situations like the National Public Data (NPD) breach, where cyber criminals expose the addresses, phone numbers and social security numbers of hundreds of millions of people. This was due in large part to a weakness in storing login credentials and a lack of password policy enforcement.

  No one at NPD wanted this to happen, but security likely took a backseat to other business concerns, leading to the company’s filing for bankruptcy.

  More importantly, as a data broker that aggregates information from other sources, the people affected likely had no clue this organisation had been buying and selling their data. The companies originally entrusted with their information helped provide the leaked data, showing a lack of care for privacy.

  Situations like this reinforce the need for strict data protection laws and for companies to refine their data governance approach. 

  Businesses can improve their data governance posturing with managers and other higher-ups setting the right tone at the top. If leadership takes a firm and disciplined approach by setting and adhering to strong policies, the rest of the team will follow and minimise the chances of data misuse and security incidents.

  One way to start is by using tools that make the principles of data ethics easier to follow.

  Ethical web analytics : Drawing insights while respecting privacy

  Web analytics tools are designed to gather data about users and what they do while visiting a site.
  The most popular tool worldwide is Google Analytics (GA). Its brand name and feature set carry a lot of weight, but many former users have switched to alternatives due to dissatisfaction with the changes made in GA4 and reservations about the way Google handles data.

  

  Google is another tech giant that has been slapped with massive GDPR fines for issues over its data processing practices. It has run so afoul of compliance that it was banned in France and Austria for a while. Additionally, in the US Department of Justice’s ongoing antitrust lawsuit against Google, the company’s data tracking has been targeted for both how it affects users and potential rivals.

  Unlike GA, ethical web analytics tools allow websites to get the data they need while respecting user privacy.

  Matomo offers privacy protections like :

  • Providing data anonymisation and IP anonymisation
  • Allowing users the ability to not process personally identifiable information (PII)
  • Disabling the ability to track users across websites by default to make compliance easier
  • Giving users full control over their customer data without the risk of it getting into the hands of third parties
  • Much more

  We’re also fully transparent about how we handle your data on the web and in the Matomo Cloud and in how we build Matomo as an open-source tool. Our openness allows you to be more open with your customers and how you ethically use their data.

  There are other GDPR-compliant tools on the market, but some of them, like Adobe Analytics, require more setup from users for compliance, don’t grant full control over data and don’t offer on-premise options or consent-free tracking.

  Beyond tracking, there are other ways to make a user’s experience more enjoyable and ethical.

  Ethical user experience : User-friendliness, not user-hostility

  When designing a website or application, creating a positive user experience (UX) always comes first. 

  The UI should be simple to navigate, data and privacy policy information should be easy to find and customers should feel welcomed. They must never be tricked into consenting or installing. 

  When businesses resort to user-hostile tactics, the UX becomes a battle between the user and them. What may seem like a clever tactic to increase sign-ups can alienate potential customers and ruin a brand’s image. 

  Here are some best practices for creating a more ethical UX :

  Avoid dark patterns

  Dark patterns are UI designs and strategies that mislead users into paying for, agreeing to or doing something they don’t actually want. These designs are unethical because they’re manipulative and remove transparency and consent from the interaction. 

  In some cases, they’re illegal and can bring lawsuits. 

  In 2023, Italy’s Data Protection Authority (DPA) fined a digital marketing company €300,000 for alleged GDPR violations. They employed dark patterns by asking customers to accept cookies again after rejecting them and placing the option to reject cookies outside the cookie banner. 

  Despite their legality and 56% of surveyed customers losing trust in platforms that employ dark patterns, a review by the Organisation for Economic Co-operation and Development (OECD) found that 76% of the websites examined contained at least one dark pattern.

  

  If a company is worried that they may be relying on dark patterns, here are some examples of what to avoid :

  • Pre-ticking boxes to have users agree to third-party cookies, sign up for a newsletter, etc.
  • Complicated cookie banners without a one-click way to reject all unnecessary cookies
  • Hiding important text with text colour, under drop-down menus or requiring hovering over something with a mouse 
  • “Confirm shaming” users with emotionally manipulative language to delay subscription cancellations or opt out of tracking 

  Improve trust centres

  Trust centres are the sections of a website that outline how a company approaches topics like data governance, user privacy and security. 

  They should be easy to find and understand. If a user has a question about a company’s data policy, it should be one click away with language that doesn’t require a law degree to comprehend.

  Additionally, trust centres must cover all relevant details, including where data is stored and who does the subprocessing. This is an area where even some of the best-intentioned companies may miss the mark, but it’s also an easy fix and a great place to start creating a more ethical web.

  Embrace inclusivity

  People want to feel welcomed to the party — and deserve to be — regardless of their race, ethnicity, religion, gender identity, orientation or ability. 

  Inclusivity is great for customers and companies alike. 

  A study by the Unstereotype Alliance found that progressive marketing drove up short- and long-term sales, customer loyalty and purchase consideration. A Kantar study reported that 75% of surveyed customers around the world consider a company’s diversity and inclusivity when making a purchasing decision.

  An easy place to start embracing inclusivity is with a website’s blog images. The people in photos and cartoons should reflect a variety of different backgrounds.

  Another area to improve inclusivity is by making your site or app more accessible.

  Accessibility ethics : An internet for everyone

  Accessibility is designing your product in a way that everyone can enjoy or take part in, regardless of ability. Digital accessibility is applying this design to the web and applications by making accommodations like adding descriptive alt text to images for users with visual impairments.

  Just because someone has a hearing, vision, speech, mobility, neurological or other impairment doesn’t mean they have any less of a right to shop online, read silly listicles or get into arguments with strangers in the comment section.

  Beyond being the right thing to do, the Fable team shows there’s a strong business case for accessibility. People with disabilities have money to spend, and the accommodations businesses make for them often benefit people without disabilities, too – as anyone who streams with subtitles can attest.

  Despite being a win-win for greater inclusivity and business, much of the web is still inaccessible. WebAIM, a leader in web accessibility, studied a million web pages and found an average of over 55 accessibility errors per page.

  We must all play a more active role in improving the experience of our users with disabilities, and we can start with accessibility auditing and testing.

  An accessibility audit is an evaluation of how usable a site is for people with disabilities. It may be done in-house by an expert on a company’s team or, for better results, a third-party consultant who can give a fully objective audit.

  Auditing might consist of running an automated tool or manually checking your site, PDFs, emails and other materials for compliance with the Web Content Accessibility Guidelines list.

  Accessibility testing is narrower than auditing. It checks how accessibility or its absence looks in action. It can be done after a site, app, email or product is released, but it ideally starts in the development process.

  Testing should be done manually and with automated tools. Manual checks put developers in the position of their users, allowing them to get a better idea of what users are dealing with firsthand. Automated tools can save time and money, but there should always be manual testing in the process.

  Auditing gives teams an idea of where to start with improving accessibility, and testing helps make sure accommodations work as intended.

  Conclusion

  At Matomo, we strive to make the ethical web a reality, starting with web analytics.

  For our users, it means full compliance with stringent policies like GDPR and providing 100% accurate data. For their customers, it’s collecting only the data required to do the job and enabling cookieless configurations to get rid of annoying banners. 

  For both parties, it’s knowing that respect for privacy is one of our foundational values, whether it’s the ability to look under Matomo’s hood and read our open-source code, the option to store data on-premise to minimise the chances of it falling into the wrong hands or one of the other ways that we protect privacy.

  If you weren’t 100% ethical before, it’s never too late to change. You can even bring your Google Analytics data with you.

  Join us in our mission to improve the web. We can’t do it alone ! 

  Try Matomo free for 21 days

  no credit card required

• Privacy-enhancing technologies : Balancing data utility and security

  18 juillet, par Joe

  In the third quarter of 2024, data breaches exposed 422.61 million records, affecting millions of people around the world. This highlights the need for organisations to prioritise user privacy. 

  Privacy-enhancing technologies can help achieve this by protecting sensitive information and enabling safe data sharing. 

  This post explores privacy-enhancing technologies, including their types, benefits, and how our website analytics platform, Matomo, supports them by providing privacy-focused features.

  What are privacy-enhancing technologies ? 

  Privacy Enhancing Technologies (PETs) are tools that protect personal data while allowing organisations to process information responsibly. 

  In industries like healthcare, finance and marketing, businesses often need detailed analytics to improve operations and target audiences effectively. However, collecting and processing personal data can lead to privacy concerns, regulatory challenges, and reputational risks.

  PETs minimise the collection of sensitive information, enhance security and allow users to control how companies use their data. 

  Global privacy laws like the following are making PETs essential for compliance :

  • General Data Protection Regulation (GDPR) in the European Union
  • California Consumer Privacy Act (CCPA) in California
  • Personal Information Protection and Electronic Documents Act (PIPEDA) in Canada
  • Lei Geral de Proteção de Dados (LGPD) in Brazil

  Non-compliance can lead to severe penalties, including hefty fines and reputational damage. For example, under GDPR, organisations may face fines of up to €20 million or 4% of their global annual revenue for serious violations. 

  Types of PETs 

  What are the different types of technologies available for privacy protection ? Let’s take a look at some of them. 

  Homomorphic encryption

  Homomorphic encryption is a cryptographic technique in which users can perform calculations on cipher text without decrypting it first. When the results are decrypted, they match those of the same calculation on plain text. 

  This technique keeps data safe during processing, and users can analyse data without exposing private or personal data. It is most useful in financial services, where analysts need to protect sensitive customer data and secure transactions. 

  Despite these advantages, homomorphic encryption can be complex to compute and take longer than other traditional methods. 

  Secure Multi-Party Computation (SMPC)

  SMPC enables joint computations on private data without revealing the raw data. 

  In 2021, the European Data Protection Board (EDPB) issued technical guidance supporting SMPC as a technology that protects privacy requirements. This highlights the importance of SMPC in healthcare and cybersecurity, where data sharing is necessary but sensitive information must be kept safe. 

  For example, several hospitals can collaborate on research without sharing patient records. They use SMPC to analyse combined data while keeping individual records confidential. 

  Synthetic data

  Synthetic data is artificially generated to mimic real datasets without revealing actual information. It is useful for training models without compromising privacy. 

  Imagine a hospital wants to train an AI model to predict patient outcomes based on medical records. Sharing real patient data, however, poses privacy challenges, so that can be changed with synthetic data. 

  Synthetic data may fail to capture subtle nuances or anomalies in real-world datasets, leading to inaccuracies in AI model predictions.

  Pseudonymisation

  Pseudonymisation replaces personal details with fake names or codes, making it hard to determine who the information belongs to. This helps keep people’s personal information safe. Even if someone gets hold of the data, it’s not easy to connect it back to real individuals. 

  

  Pseudonymisation works differently from synthetic data, though both help protect individual privacy. 

  When we pseudonymise, we take factual information and replace the bits that could identify someone with made-up labels. Synthetic data takes an entirely different approach. It creates new, artificial information that looks and behaves like real data but doesn’t contain any details about real people.

  Differential privacy

  Differential privacy adds random noise to datasets. This noise helps protect individual entries while still allowing for overall analysis of the data. 

  It’s useful in statistical studies where trends need to be understood without accessing personal details.

  For example, imagine a survey about how many hours people watch TV each week. 

  Differential privacy would add random variation to each person’s answer, so users couldn’t tell exactly how long John or Jane watched TV. 

  However, they could still see the average number of hours everyone in the group watched, which helps researchers understand viewing habits without invading anyone’s privacy.

  Zero-Knowledge Proofs (ZKP)

  Zero-knowledge proofs help verify the truth without exposing sensitive details. This cryptographic approach lets someone prove they know something or meet certain conditions without revealing the actual information behind that proof.

  Take ZCash as a real-world example. While Bitcoin publicly displays every financial transaction detail, ZCash offers privacy through specialised proofs called Zero-Knowledge Succinct Non-Interactive Arguments of Knowledge (zk-SNARKs). These mathematical proofs confirm that a transaction follows all the rules without broadcasting who sent money, who received it, or how much changed hands.

  The technology comes with trade-offs, though. 

  Creating and checking these proofs demands substantial computing power, which slows down transactions and drives up costs. Implementing these systems requires deep expertise in advanced cryptography, which keeps many organisations from adopting them despite their benefits.

  Trusted Execution Environment (TEE)

  TEEs create special protected zones inside computer processors where sensitive code runs safely. These secure areas process valuable data while keeping it away from anyone who shouldn’t see it.

  TEEs are widely used in high-security applications, such as mobile payments, digital rights management (DRM), and cloud computing.

  Consider how companies use TEEs in the cloud : A business can run encrypted datasets within a protected area on Microsoft Azure or AWS Nitro Enclaves. Due to this setup, even the cloud provider can’t access the private data or see how the business uses it. 

  TEEs do face limitations. Their isolated design makes them struggle with large or spread-out computing tasks, so they don’t work well for complex calculations across multiple systems.

  Different TEE implementations often lack standardisation, so there can be compatibility issues and dependence on specific vendors. If the vendor stops the product or someone discovers a security flaw, switching to a new solution often proves expensive and complicated.

  Obfuscation (Data masking)

  Data masking involves replacing or obscuring sensitive data to prevent unauthorised access. 

  It replaces sensitive data with fictitious but realistic values. For example, a customer’s credit card number might be masked as “1234-XXXX-XXXX-5678.” 

  The original data is permanently altered or hidden, and the masked data can’t be reversed to reveal the original values.

  Federated learning

  Federated learning is a machine learning approach that trains algorithms across multiple devices without centralising the data. This method allows organisations to leverage insights from distributed data sources while maintaining user privacy.

  For example, NVIDIA’s Clara platform uses federated learning to train AI models for medical imaging (e.g., detecting tumours in MRI scans). 

  Hospitals worldwide contribute model updates from their local datasets to build a global model without sharing patient scans. This approach may be used to classify stroke types and improve cancer diagnosis accuracy.

  Now that we have explored the various types of PETs, it’s essential to understand the principles that guide their development and use. 

  Key principles of PET (+ How to enable them with Matomo) 

  PETs are based on several core principles that aim to balance data utility with privacy protection. These principles include :

  Data minimisation

  Data minimisation is a core PET principle focusing on collecting and retaining only essential data.

  Matomo, an open-source web analytics platform, helps organisations to gather insights about their website traffic and user behaviour while prioritising privacy and data protection. 

  Recognising the importance of data minimisation, Matomo offers several features that actively support this principle :

  • Cookieless tracking : Eliminates reliance on cookies, reducing unnecessary data collection.
  • IP anonymisation : Automatically anonymises IP addresses, preventing identification of individual users.

  

  (Image Source)

  • Custom data retention policies : Allows organisations to define how long user data is stored before automatic deletion.

  7Assets, a fintech company, was using Google Analytics and Plausible as their web analytics tools. 

  However, with Google Analytics, they faced a problem of unnecessary data tracking, which created legal work overhead. Plausible didn’t have the features for the kind of analysis they wanted. 

  They switched to Matomo to enjoy the balance of privacy yet detailed analytics. With Matomo, they had full control over their data collection while also aligning with privacy and compliance requirements.

  Transparency and User Control

  Transparency and user control are important for trust and compliance. 

  Matomo enables these principles through :

  • Consent management : Offers integration with Consent Mangers (CMPs), like Cookiebot and Osano, for collecting and managing user consent.
  • Respect for DoNotTrack settings : Honours browser-based privacy preferences by default, empowering users with control over their data.

  

  (Image Source)

  • Opt-out mechanisms : These include iframe features that allow visitors to opt out of tracking. 

  Security and Confidentiality

  Security and confidentiality protect sensitive data against inappropriate access. 

  Matomo achieves this through :

  • On-premise hosting : Gives organisations the ability to host analytics data on-site for complete data control.
  • Data security : Protects stored information through access controls, audit logs, two-factor authentication and SSL encryption.
  • Open source code : Enables community reviews for better security and transparency.

  Purpose Limitation

  Purpose limitation means organisations use data solely for the intended purpose and don’t share or sell it to third parties. 

  Matomo adheres to this principle by using first-party cookies by default, so there’s no third-party involvement. Matomo offers 100% data ownership, meaning all the data organisations get from our web analytics is of the organisation, and we don’t sell it to any external parties. 

  Compliance with Privacy Regulations

  Matomo aligns with global privacy laws such as GDPR, CCPA, HIPAA, LGPD and PECR. Its compliance features include :

  • Configurable data protection : Matomo can be configured to avoid tracking personally identifiable information (PII).
  • Data subject request tools : These provide mechanisms for handling requests like data deletion or access in accordance with legal frameworks.
  • GDPR manager : Matomo provides a GDPR Manager that helps businesses manage compliance by offering features like visitor log deletion and audit trails to support accountability.

  

  (Image Source)

  Mandarine Academy is a French-based e-learning company. It found that complying with GDPR regulations was difficult with Google Analytics and thought GA4 was hard to use. Therefore, it was searching for a web analytics solution that could help it get detailed feedback on its site’s strengths and friction points while respecting privacy and GDPR compliance. With Matomo, it checked all the boxes.

  Data collaboration : A key use case of PETs

  One specific area where PETs are quite useful is data collaboration. Data collaboration is important for organisations for research and innovation. However, data privacy is at stake. 

  This is where tools like data clean rooms and walled gardens play a significant role. These use one or more types of PETs (they aren’t PETs themselves) to allow for secure data analysis. 

  Walled gardens restrict data access but allow analysis within their platforms. Data clean rooms provide a secure space for data analysis without sharing raw data, often using PETs like encryption. 

  Tackling privacy issues with PETs 

  Amidst data breaches and privacy concerns, organisations must find ways to protect sensitive information while still getting useful insights from their data. Using PETs is a key step in solving these problems as they help protect data and build customer trust. 

  Tools like Matomo help organisations comply with privacy laws while keeping data secure. They also allow individuals to have more control over their personal information, which is why 1 million websites use Matomo.

  In addition to all the nice features, switching to Matomo is easy :

  “We just followed the help guides, and the setup was simple,” said Rob Jones. “When we needed help improving our reporting, the support team responded quickly and solved everything in one step.” 

  To experience Matomo, sign up for our 21-day free trial, no credit card details needed.