32584 research outputs found
Sort by
Evaluating Demand-Responsive Scheduling in Public Transportation Service Offerings to Retirement Communities
Get PDFThis project addresses alternatives to fixed schedule local transportation service offerings to members of retirement communities. The project design measures the effects of the contrast between fixed schedule (Fixed) service and demand-responsive services (DRT) in retirement communities in Santa Clara County that are closely matched in demographic profiles and geographic location. The research team investigates the effects of scheduling alternatives in a diary format with a within-person variable of four recent trips. The measure of public transportation usage under the Fixed schedule and DRT service offerings is supplemented by a measure of subjective well-being (STS) for each trip—in other words, the project tracks not only how much/when participants use these public transportation models but also how they feel about its use. Results indicate that the community with DRT service had significantly higher public transportation usage and STS ratings than the committee with fixed schedules. Evaluations of the service provider Valley Transportation Authority (VTA) from DRT community members trended higher but were not significantly different across each item measures in the scale. These results show the potential benefits of flexible transportation options in enhancing mobility and satisfaction among older adults and can provide informative insights for policymakers to make improvements in mobility for this group and everyone
Open Access and Bottom Lines: Finance, Tech, and Libraries
No full text“It is often stated that libraries are not neutral, but when libraries rely on others to build connections on their behalf, the lack of neutrality deepens.” – Smith, 2025, p. 13
This presentation will combine philosophical and practical research findings to illustrate the impact that technological business interests have on access to information and cultural heritage preservation. Framed within the context of original research (2024, 2025), the presentation will begin with theoretical explorations of technological impacts, both local and global, on libraries. The second portion of the presentation will provide an institutional case study that exemplifies such theories. Both paywalled and open access will be discussed, with attention drawn to findings that illustrate through both theory and in practice that in a tech-centric world financial bottom lines shape cultural heritage.
Smith, C. (2024). Lack of Collections as Data: Making Meaning Out of the Films We Cannot See. Canadian Journal of Information and Library Science, (47)3, 11-20. https://doi.org/10.5206/cjils-rcsib.v47i3.18988
Smith, C.F., ed. (2025). Platform Power and Libraries. (162) Sacramento, CA: Litwin Books (978-1-63400-156-4) https://litwinbooks.com/books/platform-power-and-libraries/ & https://spectrum.library.concordia.ca/995573/1/PPAL_final.pd
A multimodal framework for enhancing E-commerce information management using vision transformers and large language models
Get PDFIn the rapidly advancing field of visual search technology, traditional methods that rely only on visual features often struggle with accuracy and relevance. This challenge is particularly evident in e-commerce, where precise product recommendations are critical, and is further complicated by keyword stuffing in product descriptions. To address these limitations, this study introduces BiLens, a multimodal recommendation framework that integrates both visual and textual information. BiLens leverages large language models (LLMs) to generate descriptive captions from image queries, which are transformed into word embeddings, and extracts visual features using Vision Transformers (ViT). The visual and textual representations are integrated using an early fusion strategy and compared using cosine similarity, enabling deeper contextual understanding and enhancing the accuracy and relevance of product recommendations in capturing customer intent. A comprehensive evaluation was conducted using Amazon product data across five categories, testing various image captioning models and embedding methods—including BLIP-2, ViT-GPT2, BLIP-Image-Captioning-Large, Florence-2-large, GIT (microsoft/git-base-coco), Word2Vec, GloVe, BERT, and ELMo. The combination of Florence-2-large and BERT emerged as the most effective, achieving a precision of 0.81±0.14 and F1 score of 0.49±0.16. This setup was further validated on the Myntra dataset, showing generalizability with precision of 0.59±0.27, recall of 0.47±0.25, and F1 score of 0.52±0.24. Comparisons with image-only and text-only baselines confirmed the superiority of the fusion-based approach, with statistically significant improvements in F1 scores, underscoring BiLens\u27s ability to deliver more accurate, context-aware product recommendations
Machine learning-guided field site selection for river classification
Get PDFSufficient abundance and variety of field site sampling are crucial for obtaining an accurate reach-scale river classification of a regional stream network in support of scientific research and river management. However, many studies still randomly select field sites or only visit accessible streams. This leads to an inadequate exploration of stream characteristics, resulting in incomplete or inaccurate classification. Machine learning has been recognized for discovering and extracting streams’ geomorphic patterns efficiently and accurately from data, but its application in field site sampling design is still in its infancy. This study developed a general and practical field site selection framework by incorporating machine learning in a human-in-the-loop manner. This framework includes three steps: (1) initial field site selection via machine learning from prior datasets, (2) selected field site accessibility evaluation and observation, and (3) additional field site decision and selection via an iterative learning process. In an example application to the San Francisco Bay Area (California, USA), our framework extracted representative geomorphic characteristics of (i) previous known stream types from prior labeled and geospatial datasets and (ii) previously unrecognized stream types based on uncertainty information obtained by machine learning. Moreover, we propose methods for replacing inaccessible sites to ensure sufficient information is retained in the selected field sites. Results revealed clear differences in variable distributions between the 148 high‐certainty sites and the 51 high‐uncertainty sites, a pattern that was validated by our field surveys. Furthermore, the 41 newly identified high‐uncertainty sites were found under-represented in the initial surveyed sites and thus their selection for the next round of field surveys will help fill the important feature gaps left by the initial survey. The feasibility of this framework allows river scientists and land use decision-makers to better understand river patterns and manage spatial planning
Data and AI mystification: Ownership, control, and financialization in the platform
Get PDFThe social relation between the platform and its users is defined by engagement with digital infrastructures and the rendition of this engagement into data. User-data, however, are surrounded by regulatory and legal ambiguities as they are not accounted for as intangible assets and their ownership and control are not transparent. I investigate how platforms rectify these ambiguities to realize user-data value as codified capital through an intensive case study of two major platforms. I use qualitative content analysis (QCA) to analyze annual and earnings reports, terms of service (ToS) agreements, and internal documents from 2017 through 2023 with qualitative data analysis (QDA) software. The findings reveal the exploitation of user-data ambiguities by platforms on two fronts: the necessary relationship between user-data inputs and platform outputs, with a growing emphasis on artificial intelligence, and the ownership and control of user-data. I argue that these ambiguities are exploited by platforms in a process of mystification of user-data to investors and other political economic actors at one end and users at the other. Mystified assets are then transformed into codified capital through financialization in the platform, contributing to studies of corporate financialization and fictitious capital. The findings place the ownership and control of user-data and their relationship to platform outputs as essential to advancing data accumulation and platform financialization
Improving Equity, Diversity, and Inclusion in LIS Education: Insights for Curriculum Development and Professional Preparation
Get PDFObjective – To explore how the graduate-level Library and Information Studies (LIS) curriculum can be redesigned to be more socially justice focused and thus better prepare graduates to address equity, diversity, and inclusion (EDI) issues in their workplace. Design – A cross-sectional, descriptive online survey study. Setting – MA/PG Diploma program in LIS at University College, London (UCL). Subjects – 59 recent graduates from the MA/PG Diploma program in LIS at University College, London (UCL). Methods – Using the descriptive survey methodology, a 13-item online questionnaire was sent to a purposive sample of 733 alumni from the MA/PG Diploma program in LIS at University College, London (UCL). The online survey included 7 closed question and 6 open-ended questions, and was open for 6 weeks. Survey responses were analyzed using thematic coding in NVIVO software to identify key trends and insights. Main Results – Regarding effective pedagogical strategies for EDI-focused work, a notable theme was the importance of personal identity in understanding and engaging with EDI issues. Respondents mentioned that their own experiences of marginalization, promotion to management roles, and personal study, helped them recognize the significance of EDI in their professional lives and understand the broad array of protected characteristics in their EDI work. Group work and community building were also identified as crucial for effective EDI education. Respondents noted that working collaboratively, both in professional associations and with colleagues, helped them maintain motivation and deepen their understanding of EDI issues. Workshops, discussion groups, and online forums were highlighted as valuable tools for fostering these connections and promoting shared understandings. Another key theme was the need to embed an EDI ethos throughout the entire curriculum rather than isolating it in specific modules. Respondents advocated for integrating EDI principles into all aspects of LIS education, including lectures, reading lists, and course content to ensure a holistic approach. Gaps in the curriculum were also noted. First, there was a lack of training in management and leadership, particularly in areas like inclusive hiring practices and managing diverse teams. Respondents felt unprepared to address these practical challenges, which are critical for nurturing a diverse and equitable workplace. Second, fostering learner positionality needs to be strengthened in the curriculum. Positionality refers to how differences in social position, identity, and power dynamics shape individuals’ experiences and access to opportunities. Without developing an understanding of these dynamics, students may struggle to fully grasp the complexities of marginalization or may inadvertently impose their perspectives on others. Lastly, respondents highlighted the need to broaden the scope of EDI education to address all protected characteristics under UK law, not just race and ethnicity. Conclusion – Three effective pedagogical strategies and three curricular gaps were identified to help LIS graduate programs to improve their EDI-focused curriculum. Specific approaches such as embedding EDI throughout the curriculum, encouraging students to reflect on their own identities and experiences with marginalization, and promoting collaborative activities were recommended. In the process of curricular form, educators need to be mindful about the tensions related to the pressure placed on those from marginalized communities to share their experiences and lead EDI work, challenging existing structures, and performative diversity. Lessons from archival practices can be considered, such as adopting trauma-informed practices when engaging with communities that have experienced historical or ongoing harm, and shifting towards more relational and person-centered approaches to build relationships with diverse user groups
Association Between Risk Factors and Major Cancers: Explainable Machine Learning Approach
Get PDFBackground: Cancer is a life-threatening disease and a leading cause of death worldwide, with an estimated 611,000 deaths and over 2 million new cases in the United States in 2024. The rising incidence of major cancers, including among younger individuals, highlights the need for early screening and monitoring of risk factors to manage and decrease cancer risk. Objective: This study aimed to leverage explainable machine learning models to identify and analyze the key risk factors associated with breast, colorectal, lung, and prostate cancers. By uncovering significant associations between risk factors and these major cancer types, we sought to enhance the understanding of cancer diagnosis risk profiles. Our goal was to facilitate more precise screening, early detection, and personalized prevention strategies, ultimately contributing to better patient outcomes and promoting health equity. Methods: Deidentified electronic health record data from Medical Information Mart for Intensive Care (MIMIC)–III was used to identify patients with 4 types of cancer who had longitudinal hospital visits prior to their diagnosis presence. Their records were matched and combined with those of patients without cancer diagnoses using propensity scores based on demographic factors. Three advanced models, penalized logistic regression, random forest, and multilayer perceptron (MLP), were conducted to identify the rank of risk factors for each cancer type, with feature importance analysis for random forest and MLP models. The rank biased overlap was adopted to compare the similarity of ranked risk factors across cancer types. Results: Our framework evaluated the prediction performance of explainable machine learning models, with the MLP model demonstrating the best performance. It achieved an area under the receiver operating characteristic curve of 0.78 for breast cancer (n=58), 0.76 for colorectal cancer (n=140), 0.84 for lung cancer (n=398), and 0.78 for prostate cancer (n=104), outperforming other baseline models (P\u3c.001). In addition to demographic risk factors, the most prominent nontraditional risk factors overlapped across models and cancer types, including hyperlipidemia (odds ratio [OR] 1.14, 95% CI 1.11‐1.17; P\u3c.01), diabetes (OR 1.34, 95% CI 1.29‐1.39; P\u3c.01), depressive disorders (OR 1.11, 95% CI 1.06‐1.16; P\u3c.01), heart diseases (OR 1.42, 95% CI 1.32‐1.52; P\u3c.01), and anemia (OR 1.22, 95% CI 1.14‐1.30; P\u3c.01). The similarity analysis indicated the unique risk factor pattern for lung cancer from other cancer types. Conclusions: The study’s findings demonstrated the effectiveness of explainable ML models in assessing nontraditional risk factors for major cancers and highlighted the importance of considering unique risk profiles for different cancer types. Moreover, this research served as a hypothesis-generating foundation, providing preliminary results for future investigation into cancer diagnosis risk analysis and management. Furthermore, expanding collaboration with clinical experts for external validation would be essential to refine model outputs, integrate findings into practice, and enhance their impact on patient care and cancer prevention efforts
Coloring Outside the Lines: Factors Facilitating Divergent Thinking
Get PDFDivergent thinking involves producing a multitude of unique and diverse ideas. The ability to “color outside the lines” is crucial to business, education, and social domains. The flexibility, originality, and fluency dimensions of divergent thinking are commonly assessed using the Alternative Uses Task, which involves generating uses for common objects. Prior research suggests that high identification (first-person perspective) with schema-inconsistent experiences increases flexibility, while semantic density (degree of conceptual associations) impacts originality and fluency inversely. However, their combined effects across multiple divergent thinking dimensions remain unexplored. The present research investigated the interactive effects of identification (high, low), schema consistency (consistent, inconsistent), and semantic density (rich, sparse) across measures of flexibility, originality, and fluency in young adults (N = 149, mean age = 19.01). Under low or high identification, participants viewed a schema -consistent or -inconsistent video, then completed Alternative Uses Tasks for rich and sparse cues. Results revealed a significant main effect of semantic density on all measures, with rich outperforming sparse cues, and a significant three-way interaction on originality, such that the low-identification schema-consistent and high-identification schema-inconsistent conditions produced more original responses for sparse cues. These findings suggest that semantically rich prompts can lead to greater divergent thinking overall, and mentally engaging in diversified experiences can support novel ideation, offering potential targeted strategies for innovation in industry, education, and bias reduction
A Closer Look at the Relationship Between Cultural Intelligence and Organizational Citizenship Behaviors
Get PDFAs organizations—particularly in the United States—become more multicultural due to globalization, they increasingly rely on individuals\u27 ability to succeed in complex, cross?cultural environments. Cultural intelligence (CQ), or the ability to successfully adapt to unfamiliar cultural settings, has been previously related to expatriate cross-cultural adjustment, group effectiveness, and more recently, to organizational citizenship behaviors (OCB), which are voluntary extra-role workplace behaviors that lead to positive employee and organizational outcomes. This study mainly sought to theoretically expand the limited research on the relationship between CQ and OCB with some research questions. In a sample of 93 participants based in California, results found a significant relationship between CQ and OCB; a Pearson correlation analysis showed that three out of four dimensions of CQ (metacognitive, cognitive, motivational) were significantly related to three out of five dimensions of OCB (altruism, courtesy, civic virtue); a canonical correlation analysis later revealed that these same sets of dimensions were represented in one significant root. Furthermore, in the second (nonsignificant) root, behavioral CQ was related to conscientiousness and civic virtue. These findings imply that CQ might best be portrayed as having two dimensions, one measuring cognitive related constructs and one measuring a behavioral construct. Organizational implications based on the findings include HR professionals prioritizing the hiring of high-CQ individuals, as well as initiatives to improve employees’ cognitive CQ and, to a lesser extent, their behavioral CQ
Alameda Song Sparrow Habitat Use of Salt Pond Restoration Sites in South San Francisco Bay
Get PDFRestoration is acknowledged worldwide as a conservation need to return ecosystem functions. However, projects require monitoring to track system changes and assess restoration progress. In the San Francisco Bay, several large restoration projects are implementing plans to return salt-producing ponds to the historical ecosystem, tidal salt marsh. The Alameda song sparrow (Melospiza melodia pusillula), a California species of special concern, has been documented to use restored and historic marshes, but their use of restored salt ponds is not fully understood. The birds’ use of restored marshes may be a useful measure of restoration progress. This research investigated the relationships comparing Alameda song sparrow breeding and abundance with salt pond restoration age and percent cover of native plants. I also addressed whether plant communities at restoration sites of various ages met Alameda song sparrow habitat requirements. I collected transect count data on Alameda song sparrow abundance at seven sites in the South San Francisco Bay, then used quadrats to estimate plant cover within each of 3 transects per site. To assess whether sites supported breeding birds, I mist-netted and banded birds at a subset of sites where bird abundance and plant species composition data were collected. Results show little effect of time since pond breaching and native plant cover on Alameda song sparrow abundance, and a positive effect of vegetation height