書誌計量指標の（隠れた）多様性に関する考察：破壊的指標が明らかにするもの
本稿では、書誌計量指標、特に破壊的指標の算出に伴う分析の柔軟性と、それがもたらす結果の不確実性について論じ、マルチバース的手法の適用による透明性と頑健性の向上を提唱する。
The Impact of Analytical Flexibility on the Disruption Index and the Need for Multiverse-Style Methods in Bibliometrics
The disruption index (DI1), while widely used to measure disruptive research, suffers from significant analytical flexibility, leading to a multiverse of potential results and threatening the credibility of bibliometric research. The authors advocate for the adoption of multiverse-style methods to enhance transparency and robustness in the field.
利用學術論文推薦系統促進跨學科知識轉移
學術論文推薦系統應重視推薦結果的新穎性和多樣性，以促進跨學科知識轉移，打破學科壁壘，並鼓勵跨領域研究。
연구 논문 추천 시스템을 활용한 학제 간 지식 전달 촉진: 인용 네트워크 분석 및 콘텐츠 기반 접근 방식
본 논문은 연구 논문 추천 시스템에서 적절한 추천의 중요성을 넘어, 새롭고 다양한 추천을 제공함으로써 연구자들의 학제 간 연구를 촉진하고 지식 전달을 향상시킬 수 있다는 것을 주장한다.
研究論文レコメンデーションシステムを用いた学際的な知識移転の促進
本稿では、サイロ化された論文閲覧を減らし、フィルターバブルを打破し、学際的な研究を促進するために、研究者に対して新規性と多様性に富んだ論文レコメンデーションを提供することの重要性を論じています。
The Impact of Paper Embedding Methods on the Novelty and Diversity of Research Paper Recommendations
The choice of research paper embedding method significantly influences the novelty and diversity of recommendations provided by research paper recommender systems, potentially impacting interdisciplinary knowledge transfer.
The Relationship Between Unemployment and High School Graduation Rates in Maryland, Virginia, and Washington D.C.
Unemployment rates are a stronger predictor of high school graduation rates than per-pupil school spending in Maryland, Virginia, and Washington D.C.
Alternate Learning and Compression Approaching the Rate-Distortion Bound
This paper explores the connection between backward-adaptive lossy compression and online learning, specifically focusing on the Natural Type Selection (NTS) algorithm and its ability to approach the rate-distortion bound by balancing exploration and exploitation.
HtmlRAG：在 RAG 系統中，HTML 比純文字更適合建模檢索到的知識
HtmlRAG 是一種新的 RAG 方法，它使用 HTML 作為外部知識的格式，透過保留 HTML 文件中的結構和語義資訊來提高 RAG 系統的效能。
HTMLRAG: RAG 시스템에서 검색된 지식 모델링을 위해 HTML이 일반 텍스트보다 효과적인 이유
HTML 문서는 제목, 표 구조와 같은 구조적 및 의미적 정보를 풍부하게 담고 있어 일반 텍스트보다 RAG 시스템의 검색 지식 모델링에 더 효과적입니다.
© 2024 by Linnk AI