Search for: All records

ABSTRACT This paper reports on a demonstration of YAMZ (Yet Another Metadata Zoo) as a mechanism for building community consensus around metadata terms. The demonstration is motivated by the complexity of the metadata standards environment and the need for more user-friendly approaches for researchers to achieve vocabulary consensus. The paper reviews a series of metadata standardization challenges, explores crowdsourcing factors that offer possible solutions, and introduces the YAMZ system. A YAMZ demonstration is presented with members of the Toberer materials science laboratory at the Colorado School of Mines, where there is a need to confirm and maintain a shared understanding for the vocabulary supporting research documentation, data management, and their larger metadata infrastructure. The demonstration involves three key steps: 1) Sampling terms for the demonstration, 2) Engaging graduate student researchers in the demonstration, and 3) Reflecting on the demonstration. The results of these steps, including examples of the dialog provenance among lab members and voting, show the ease with YAMZ can facilitate building metadata vocabulary consensus. The conclusion discusses implications and highlights next steps. 

Sentence embedding methods offer a powerful approach for working with short textual constructs or sequences of words. By representing sentences as dense numerical vectors, many natural language processing (NLP) applications have improved their performance. However, relatively little is understood about the latent structure of sentence embeddings. Specifically, research has not addressed whether the length and structure of sentences impact the sentence embedding space and topology. This paper reports research on a set of comprehensive clustering and network analyses targeting sentence and sub-sentence embedding spaces. Results show that one method generates the most clusterable embeddings. In general, the embeddings of span sub-sentences have better clustering properties than the original sentences. The results have implications for future sentence embedding models and applications. 

Scientific literature presents a wellspring of cutting-edge knowledge for materials science, including valuable data (e.g., numerical data from experiment results, material properties and structure). These data are critical for accelerating materials discovery by data-driven machine learning (ML) methods. The challenge is, it is impossible for humans to manually extract and retain this knowledge due to the extensive and growing volume of publications.To this end, we explore a fine-tuned BERT model for extracting knowledge. Our preliminary results show that our fine-tuned Bert model reaches an f-score of 85% for the materials named entity recognition task. The paper covers background, related work, methodology including tuning parameters, and our overall performance evaluation. Our discussion offers insights into our results, and points to directions for next steps.