Radio remains a pervasive medium for mass information dissemination, with AM/FM stations reaching more Americans than either smartphone-based social networking or live television. Increasingly, radio broadcasts are also streamed online and accessed over the Internet. We present WavePulse, a framework that records, documents, and analyzes radio content in real-time. While our framework is generally applicable, we showcase the efficacy of WavePulse in a collaborative project with a team of political scientists focusing on the 2024 Presidential Elections. We use WavePulse to monitor livestreams of 396 news radio stations over a period of three months, processing close to 500,000 hours of audio streams. These streams were converted into time-stamped, diarized transcripts and analyzed to track answer key political science questions at both the national and state levels. Our analysis revealed how local issues interacted with national trends, providing insights into information flow. Our results demonstrate WavePulse's efficacy in capturing and analyzing content from radio livestreams sourced from the Web. Code and dataset can be accessed at \url{https://wave-pulse.io}.

Indexing is an important step towards strong performance in retrieval-augmented generation (RAG) systems. However, existing methods organize data based on either semantic similarity (similarity) or related information (relatedness), but do not cover both perspectives comprehensively. Our analysis reveals that modeling only one perspective results in insufficient knowledge synthesis, leading to suboptimal performance on complex tasks requiring multihop reasoning. In this paper, we propose SiReRAG, a novel RAG indexing approach that explicitly considers both similar and related information. On the similarity side, we follow existing work and explore some variances to construct a similarity tree based on recursive summarization. On the relatedness side, SiReRAG extracts propositions and entities from texts, groups propositions via shared entities, and generates recursive summaries to construct a relatedness tree. We index and flatten both similarity and relatedness trees into a unified retrieval pool. Our experiments demonstrate that SiReRAG consistently outperforms state-of-the-art indexing methods on three multihop datasets (MuSiQue, 2WikiMultiHopQA, and HotpotQA), with an average 1.9% improvement in F1 scores. As a reasonably efficient solution, SiReRAG enhances existing reranking methods significantly, with up to 7.8% improvement in average F1 scores.

Knowledge graphs (KGs) generated by large language models (LLMs) are becoming increasingly valuable for Retrieval-Augmented Generation (RAG) applications that require knowledge-intensive reasoning. However, existing KG extraction methods predominantly rely on prompt-based approaches, which are inefficient for processing large-scale corpora. These approaches often suffer from information loss, particularly with long documents, due to the lack of specialized design for KG construction. Additionally, there is a gap in evaluation datasets and methodologies for ontology-free KG construction. To overcome these limitations, we propose SynthKG, a multi-step, document-level ontology-free KG synthesis workflow based on LLMs. By fine-tuning a smaller LLM on the synthesized document-KG pairs, we streamline the multi-step process into a single-step KG generation approach called Distill-SynthKG, substantially reducing the number of LLM inference calls. Furthermore, we re-purpose existing question-answering datasets to establish KG evaluation datasets and introduce new evaluation metrics. Using KGs produced by Distill-SynthKG, we also design a novel graph-based retrieval framework for RAG. Experimental results demonstrate that Distill-SynthKG not only surpasses all baseline models in KG quality -- including models up to eight times larger -- but also consistently excels in retrieval and question-answering tasks. Our proposed graph retrieval framework also outperforms all KG-retrieval methods across multiple benchmark datasets. We release the SynthKG dataset and Distill-SynthKG model publicly to support further research and development.

On July 14, Key Technology debuted its new FM Alert software driven by artificial intelligence (AI). The new AI alert system can help processors control foreign materials entering product streams, as well as improving documentation and overall food safety. It will be a part of the company's exhibit at Pack Expo in October at booth S-3547. The AI system captures and saves images of foreign materials (FMs) that a sorter detects and rejects from its stream, with data available immediately to alert operators. "Thanks to the application of advanced artificial intelligence, our new FM Alert software achieves uniquely accurate results -- identifying, recording and acting on true FM findings on the line," said Marco Azzaretti, director of marketing at Key. "The food processing industry continues to focus more and more on elevating food safety. By making product safer, this effective FM-fighting tool helps customers protect their brand's reputation and avoid costly recalls. Every food processor wants to prevent contamination, making FM Alert universally beneficial across all applications."

Key Technology introduces AI-driven FM alert software for its digital sorting systems. This powerful tool captures and saves digital images of critical foreign material (FM) contaminants that the sorter detects and rejects from the product stream. Data outputs from the software can be utilized to immediately alert operators and/or signal a downstream device. AI-enhanced FM Alert helps processors better control FM and improve documentation to protect food safety. "Thanks to the application of advanced artificial intelligence, our new FM Alert software achieves uniquely accurate results – identifying, recording, and acting on true FM findings on the line," said Marco Azzaretti, director of marketing at Key. "The food processing industry continues to focus more and more on elevating food safety. By making product safer, this effective FM-fighting tool helps customers protect their brand's reputation and avoid costly recalls. Every food processor wants to prevent contamination, making FM Alert universally beneficial across all applications."

HAYWARD, California – Robots that can cook -- from flipping burgers to baking bread -- are in growing demand as virus-wary kitchens try to put some distance between workers and customers. Starting this fall, the White Castle burger chain will test a robot arm that can cook french fries and other foods. The robot, dubbed Flippy, is made by Pasadena, California-based Miso Robotics. White Castle and Miso have been discussing a partnership for about a year. Those talks accelerated when COVID-19 struck, said White Castle Vice President Jamie Richardson.

Twenty-five different startups were funded in January cumulatively raising $784 million; a great start for the new year. Four acquisitions were reported during the month while the IPO front lay waiting for something to happen. According to Silicon Valley Bank's annual survey of 1,000 executives, startup founders are confident 2018 will be a good year for funding and for business conditions – except hiring and retaining foreign talent. More than half of startups surveyed (51%) reported that at least one of their founders is an immigrant. One-third of startups said laws and regulations prompted them to locate facilities (or move non-sales operations) offshore.

Masters of Illusion Xavier Mortimer, Scott Pepper, Chris Randall, Joshua Jay, Murray SawChuck, Billy Kidd and Greg Gleason are the featured magicians in this new episode hosted by Dean Cain. Killjoys As the Killjoys plan to execute a high-risk theft on a well-armored convoy, Aneela (Hannah John-Kamen) is distracted by her own to find Delle Seyah (Mayko Nguyen). Guy's Family Road Trip The Fieris depart Flagstaff, Ariz., and enjoy a flight through the Grand Canyon before heading east on Route 66 and making a few stops before getting to Albuquerque. Dark Matter Former Raza crew member and now emperor Ryo (Alex Mallari Jr.) captures Two (Melissa O'Neil) and holds her hostage in a bid to force the crew of the ship to give up the Blink Drive. Three, Five and Six (Anthony Lemke, Jodelle Ferland, Roger Cross) are ready to deal, but Ryo's negotiating position is undermined by treachery among his own ranks in this new episode of the science-fiction series.

John Markoff talks about his book, "What The Dormouse Said: How the 60s Counterculture Shaped the Personal Computer Industry." In the 60s, John McCarthy was working on replacing human intelligence using artificial intelligence and Doug Engelbart was working on augmenting human intelligence using computers. Both of them profoundly influenced several other engineers including those at Xerox PARC and the Stanford Artifical Intelligence Laboratory (SAIL). John Talks about their contribution and how the PC revolution eventually unfolded. This is Part 1 of a two-part presentation.

UBIQUITY: Congratulations on "What the Dormouse Said" it's a fascinating book. MARKOFF: Well, I guess I'd call it a revisionist history. It about things that happened around Stanford University between roughly 1960 and 1975, and is a kind of pre-history of personal computing and the personal computer industry. What I was trying to do was to get at some of the culture through which the technology was developed. MARKOFF: Because technology never happens in a vacuum. The book was an effort to try to pin down how personal computing first emerged around the Stanford campus at two laboratories in the 1960's: one was run by John McCarthy, and was called the Stanford Artificial Intelligence Laboratory; and the other was run by Doug Engelbart and known as the Augmentation Research Center or the Augmented Human Intellect Research Center.