Codreanu Dana

The library that integrates the #data processing library #polars with the database system #duckdb is here: https://lnkd.in/d6ZMab-H Now that #polars and #duckdb have comparable performances, the integrated #python library is a good news. The #data orchestrator platform that integrated the two is Dagster: ✅ https://dagster.io/ Comparison of #polars with other tools: https://lnkd.in/dZncufK5

3

1 Comentariu

The aim of our #dataviz track is to share knowledge about data visualization methods, tools, best practices and applications. Topics include Chart Design, Dashboarding, Data Storytelling, Case studies and Best practices. The talks can be delivered either in person in Budapest or remotely via Zoom. More details about the track: https://lnkd.in/dMv7i2eV #budapestbi24 #callforspeakers #dataviz

3

🌱🌐 Harnessing AI for Biodiversity Monitoring: A Non-Invasive Approach Thrilled to share our latest research exploring how advanced AI techniques can revolutionize ecological studies! Our team (Ninoslava Jankovic, Branislav Radomirovic, Luka Jovanovic, Stefan Ivovic, Smiljana Tedic, Dr.-Ing. Miodrag Zivkovic, Nebojsa Bacanin Dzakula) has developed a method that uses audio analysis to estimate biodiversity without disturbing fragile ecosystems. 🔍 The Challenge: Monitoring biodiversity is crucial for understanding ecosystem health, but traditional methods often disrupt the very environments we're trying to protect. 💡 Our Solution: We combined Mel Frequency Cepstral Coefficients (MFCC) and Convolutional Neural Networks (CNNs) to identify avian species from sound recordings. To ensure high accuracy, we optimized our model's performance using the Chimp Optimization Algorithm (ChOA), achieving an impressive accuracy of 83.87%. 🌳 Why It Matters: This research offers a non-invasive and efficient way to track species and monitor ecological health. As we face growing environmental challenges, leveraging AI for conservation efforts has never been more critical. Excited to see how this work can inspire future innovations in environmental monitoring and wildlife protection! 🐦✨ Find more: https://lnkd.in/dn5CJZ4D #AI #Biodiversity #Ecology #Sustainability #MachineLearning #EnvironmentalScience

10

What is funny, reading this excellent article, is that so many of the researchers are committed to a paradigm which is premised on regional isolated language development. It would be fun to run an ML or DL comparative analysis of code languages proliferation over the last 60 years and nuanced 'similarities', and look for similar groupings and vectors to an assessment of 'language coincidences' as detailed by the article. . I think they are missing what is right under their noses :)

3

2 Comentarii

"We know AI is a risk but we have to learn to live with it" 😕 Over the past year Barsali Bhattacharyya and I have spoken to dozens of people who are being leveraging AI in their jobs, and the above is the common refrain. There's a grudging acknowledgement that things can no longer be undone now as the economic benefits outweigh the harms. That's something this article by The Economist and the short story it refers to challenges. Are we all being very short sighted and just accelerating in the hope that AI will help solve our and its own problems? The economy might grow but will climate change, energy shortage and economic inequality be exacerbated by AI in the short term? Heady questions for sure but ones we hope those working with AI and being impacted by it are thinking about. Share your thoughts if you are? 🧏 https://lnkd.in/gmBzrrVn

6

Is it possible for AI to automate the intricate hierarchical tasks involved in taxonomy? Read Firehead's latest blog post, Taxonomies, Ontologies, and Semantics in the Technical Communicator’s World of AI https://lnkd.in/d6kuguiv #techcomm #taxonomy #ontology #semantics #AI #technicalwriting

1

Discovering Global Trends with Semantic Visions' Insights! Our colleague, Tatiana Galitskaya, Ontology Information Scientist, has dug into our data and here we present some of the insights. The following chart represents the quantity of articles for key topics, across 12 languages, for the month of February 2024. Semantic Visions offers unique insights into topics that resonate worldwide. We're excited to share these trends, understanding that they reflect a complex tapestry of global discourse without stereotyping any community or language. Our unique tool, leveraging global media trends, monitors global media across 12 languages and processes over 1m articles daily. Stay tuned for more #SVInsights and discover what the world is talking about! #DataAnalytics #Onthology #Sustainability #FinancialData #Data #AI

13

More than a year on, this remains one of our favorite Heather Hedden articles. Why choose a #taxonomy? Why choose an #ontology? What matters ultimately is what you intend to do with your knowledge organization system. Effectively, do you want to retrieve data OBJECTS or retrieve ATTRIBUTES about data objects? To be specific: "taxonomies [are optimized to] retrieve a spreadsheet FILE, whereas ontologies can retrieve DATA FROM INDIVIDUAL CELLS IN THE SPREADSHEET." A dozen use cases immediately spring to mind from this example! What knowledge organization system do you need? Do you need both, but in different contexts? Has you company, unfortunately, chosen the wrong one?

10

New #AIFS blog: Forecast performance is multifaceted. Zied Ben Bouallègue and the AIFS team discuss how accuracy and activity can be ‘competing’ attributes for #MachineLearning-based deterministic forecasts. ➡️ https://lnkd.in/gCDJxKN4

96

1 Comentariu

🚀 Excited to share our latest blog article: « A Friendly Guide to Retrieved Augmented Generation Systems Evaluation »! 🌟 In this article, we explore the core concepts of RAG and provide detailed insights into evaluation metrics as well as state-of-the-art practices on how to effectively evaluate and optimize RAG models for your projects. Whether you're a beginner or an expert in the field of Generative AI, this blog aims to provide valuable knowledge and practical tips to enhance your understanding of RAG evaluation. Check out the blog here: https://lnkd.in/eAVZiVCx As always, we’re eager to hear your thoughts and feedback. 🌐💡 #GenerativeAI #RetrievedAugmentedGeneration #AI #MachineLearning #DataScience #AIResearch #TechInnovation #llm --- Follow us at Katanemo we’re committed to making AI and Generative AI accessible to everyone!

10

Slides of the presentation for the ODISSEI - Open Data Infrastructure for Social Science and Economic Innovations conference on the next #AI generation applications of #data, including http://ask.dataverse.org and Ghostwriters to chat with the dead people! https://lnkd.in/egHV6Rch

8

Our very own Adam Wangrat shared some of our research and product development at Neural Alpha combining #LLMs with #knowledgegraphs for automating ESG assessments at the recent Connected Data World London meetup. The talk illustrates how we are using Retrieval Augmented Generation #RAG architectures to improve Q&A capabilities across disclosures, spatial, news, supply chains and many other types of ESG data. This kind of AI powered automation will become increasingly widespread as ESG becomes more data intensive and requiring automation.

11

2 Comentarii

EK's Heather Hedden recently delivered a guest-webinar for Progress Semaphore. Taxonomies at their core enable effective tagging and retrieval of content, and combined with ontologies they extend to the management and understanding of related data. There are even greater benefits of taxonomies and ontologies to enhance your enterprise information architecture when applying them to a semantic layer. Extending taxonomies to a semantic layer can be a game-changing solution, allowing you to connect information silos, alleviate knowledge gaps, and derive new insights. https://lnkd.in/e9WUpyAq

51

2 Comentarii

NEW! Part two of a #KempnerInstitute blog series: the authors propose a simple and solvable model where many of the aspects of #LLMs are already present. Read more: https://bit.ly/3RqYMhX #neuralnetworks #AI

5

Great position paper that I strongly recommend to read from Gael Varoquaux et al. entitled: Hype, Sustainability, and the Price of the Bigger-is-Better Paradigm in AI. The bigger is better assumption is not only scientifically fragile but also unsustainable. The drive for larger models has significant costs: rising compute demands, increased environmental impact, and a concentration of power. Smaller, more focused models often perform just as well or better for specific tasks, proving that scaling isn't the universal solution. In many cases, it's not about the size of the model but how it's applied. Focusing on efficiency, adaptability, and purpose-driven AI can lead to more sustainable, equitable, and impactful outcomes across sectors like healthcare, education, and climate action. Let’s move away from the "bigger is better" mindset and start embracing smarter, more purposeful AI innovations. #AI #Sustainability #MachineLearning #Innovation Read the paper: https://lnkd.in/g_HiKUrz

38

Understanding the North American Industry Classification System (NAICS) Codes is essential for accurately categorizing and analyzing industries. These codes provide a standardized way to classify businesses based on their primary economic activities, enabling better data comparison and analysis across different sectors.  https://applied-dev.com/ #NAICS #IndustryClassification #BusinessCategorization #AppliedDevelopment 

1

🚀 AI's Transformative Potential: A Vision from Anthropic's CEO Recently, Dario Amodei, CEO of Anthropic (Claude.ai), published an interesting article about how AI could change society within 5-10 years of reaching human-level abilities. He talks about longer lives, better politics, new ways of working, and a stronger economy. Definately worth to read full version: https://lnkd.in/eHQ5Jmv4 Some key points: 🧬 Speeding Up Science • AI could do 50-100 years of science in just 5-10 years, especially in medicine. • This might mean: Curing most diseases, even tough ones like cancer Helping people live to 150 years old Giving people more control over their bodies and minds 💰 Boosting Developing Countries • AI could help poor countries grow very fast. • Amodei thinks: Developing countries could grow 20% each year Africa could become as rich as China is now, in just 10 years • This growth could come from spreading new health treatments and technologies quickly 🏛️ Improving How Countries Run • AI might make democracies stronger by: Helping courts make fairer decisions Making government services work better Fighting against fake news from dictators 🌍 Solving Big Problems • Amodei also talks about challenges we need to solve: Finding new ways to run the economy when AI can do most jobs Making sure everyone benefits from AI, not just rich countries Helping people trust new AI technologies 🤔 Being Hopeful, But Careful • Amodei is excited about the future, but he says: We all need to work together to make good things happen We must be careful about AI risks too Some of these ideas might not come true, so we need to be ready for anything This view of how AI could change the world is both exciting and challenging. It makes us think about the future in new ways. What do you think about this AI-powered future? What excites or worries you? #AIFuture #Innovation #TechForGood #GlobalDevelopment #anthropic #ai #artificialintelligence

7

At EAI, we are committed to fostering discussions on these innovations and advancing AI-driven solutions that benefit the global research community. AI is undoubtedly transforming research and publishing services. 👓 read the whole article here: https://lnkd.in/eJxxD5nH By automating routine tasks, enhancing data analysis, and improving the efficiency of publishing workflows, AI is enabling researchers to focus on groundbreaking discoveries and academic excellence. 🔍 AI in Research: Accelerating Discovery and Analysis In the research world, the sheer volume of data is staggering. Whether in healthcare, climate science, or engineering, the ability to process and analyze large datasets is crucial for making new discoveries. Traditional methods, while effective, are often too slow to handle the scale of data involved today. This is where AI shines.   AI is also enhancing predictive modeling, which helps researchers forecast trends or outcomes based on historical data. AI tools can run simulations and refine models much faster than traditional methods, offering researchers an invaluable tool to address some of the world's most pressing challenges.   🔍 Streamlining the Publishing Process  Publishing has traditionally been a time-consuming and resource-intensive process. AI is now streamlining many of these steps, significantly reducing the time spent on routine tasks and allowing researchers and publishers to focus on higher-value work. 🔍 AI-Assisted Literature/Paper Reviews One of the most labor-intensive tasks for researchers is conducting literature reviews— searching through hundreds or even thousands of published papers to identify relevant work. AI is making this process far more efficient. Tools like Connected Papers and Semantic Scholar leverage AI to help researchers quickly identify and explore academic papers related to their work.   🔍 The Future of AI in Publishing  AI has the potential to improve the accuracy and accessibility of research. AI-powered tools can help ensure that academic publishing remains unbiased, objective, and transparent, addressing some of the ongoing concerns about fairness and equality in research evaluation.   👓 Ethical Considerations and Challenges  While AI offers immense benefits, its application in research and publishing raises ethical concerns that need to be addressed. Issues like algorithmic bias, transparency in decision-making, and data privacy are critical when deploying AI in academic settings. Researchers and publishers must ensure that AI tools are used responsibly and ethically, maintaining the integrity of the scientific process. 👓 read the whole article here: https://lnkd.in/eJxxD5nH #eai #eaipublishing #publishingservices #publications #ai #aitools #innovation #research #submityourpaper #publishwithus  

3

1 Comentariu