A Professor Who Speaks Fluent Search
On a Wednesday morning in Portland, Oregon, a physician types a patient description into a clinical database. She is not browsing she is searching. She needs to know whether anyone in the hospital system matches a rare combination of symptoms, and she needs to know now. The query runs through a layer of structured and unstructured health data, drawing on techniques refined over decades by researchers who understood something fundamental: the body generates text, and that text needs to be found.
That layer of infrastructure search technology purpose-built for medicine is what Dr. William Hersh has spent a career constructing. He is not a household name in consumer tech, but among the physicians, informaticists, and data scientists who build the systems that make clinical knowledge navigable, his influence is broad and longstanding. From his position at Oregon Health & Science University, Hersh has trained over a thousand graduate students, authored more than 200 scientific papers, and shaped the introductory curriculum for an entire field through a program that has reached thousands more.
The story of how one professor became a central node in the network of medical search is also, implicitly, a story about how evidence-based medicine learned to retrieve itself.
Portland, 1996: A Program Takes Shape
The Biomedical Informatics Graduate Program at OHSU was founded in 1996, with Hersh as its inaugural director. In the years that followed, the program grew into one of the oldest and largest of its kind in the world. By 2023, when Hersh stepped down as director after nearly three decades, the program had awarded more than 1,000 degrees and certificates. His biography on billhersh.info describes the program's scope as a legacy of sustained institutional commitment and his own relentless focus on education.
The timing mattered. The late 1990s was a period when healthcare organizations were beginning to digitize patient records, and the question of how to organize, store, and retrieve clinical information was moving from theoretical to urgent. Hersh's background positioned him to address that moment directly. He had trained as a physician, but his intellectual home was the space where medicine meets computing specifically, the subdomain known as information retrieval, or search.
"Through my entire career, I have been an active researcher, with a focus on the field of information retrieval, also known as search," Hersh writes on his personal site. That declaration is understated, but it names the core of everything that followed: he has been studying how humans and machines find what they need in biomedical data.
The Textbook and the Framework
Educational programs require architecture, and Hersh has built much of his through books. His Information Retrieval: A Biomedical and Health Perspective reached its fourth edition with Springer in 2020. The text treats search not as a utility but as a discipline with its own theory, methods, and clinical stakes. Springer Nature's Research Communities platform notes that the book is available through SpringerLink and Amazon, with an associated website for supplementary materials.
A companion volume, Health Informatics: Practical Guide, now in its eighth edition (Lulu.com, 2022), serves a different but complementary purpose. Where the Springer text focuses on retrieval theory, the Practical Guide offers an introductory overview of the broader biomedical informatics landscape health records, standards, clinical decision support, and public health informatics. Both books are used in his own courses and in programs worldwide.
These are not reference shelf ornaments. They are working textbooks, revised across editions as the field evolves, and they reflect Hersh's commitment to dissemination over gatekeeping. He has written to teach, not merely to publish.
The 10x10 Program: Bringing Graduate-Level Training to the World
One of Hersh's most visible contributions is the American Medical Informatics Association's 10x10 program colloquially pronounced "ten by ten." Hersh conceptualized and implemented the first offering in 2005. The program provides the equivalent of an introductory graduate-level course in biomedical informatics, delivered in a format accessible to working professionals who cannot enroll in a full degree program.
By the most recent figures in the available sources, more than 3,000 individuals had completed the course. DS-I Africa's biography of Hersh describes the program as one of his signature educational innovations, noting that it has been completed by health professionals and others who needed foundational informatics knowledge without the commitment of a multi-year graduate program.
The 10x10 model was also a response to a real-world distribution problem. Informatics expertise was concentrated at a small number of research universities, but the need for foundational knowledge existed everywhere hospitals, public health agencies, pharmaceutical companies, and primary care settings. Hersh's online and in-person teaching around the world, built on the 10x10 framework, extended the program's reach well beyond Portland.
The Research Program: Search Under the Hood
Hersh's research has never been merely pedagogical. His primary research area is information retrieval applied to biomedicine a field he describes in his publications as having both theoretical depth and clinical urgency. He has led a series of challenge evaluations under the umbrella of the Text Retrieval Conference, a long-running NIST-sponsored program that uses shared tasks to advance IR research. These evaluations have provided benchmarks and datasets that the broader biomedical IR community uses to test new approaches.
His current research focus, as described in the available sources, is on applying search techniques to electronic health record data. The goal is twofold: improving patient cohort discovery helping researchers identify groups of patients who share relevant characteristics and amplifying signals of rare diseases that might otherwise go undetected in routine clinical workflows. This is search technology operating at the level of individual patient records, not journal articles or web pages.
The distinction matters. Medical literature retrieval has its own challenges, but EHR data is messier, more variable, and more clinically immediate. Records contain free-text notes, diagnostic codes, medication lists, and unstructured observations that were never designed to be queried like a library. Hersh's work on this problem connects the methods of academic IR to the operational realities of healthcare delivery.
Training the Next Generation: The NLM T15 Program
From 2002 to 2025, Hersh served as principal investigator and director of OHSU's NLM T15 Biomedical Informatics and Data Science Research Training Program, funded by the National Library of Medicine. The grant was, by the description in the sources, the largest and second-longest training grant at OHSU. It supported predoctoral and postdoctoral trainees pursuing research careers in biomedical informatics.
The T15 mechanism is significant because it represents institutional investment in human capital. Trainees funded through the program have contributed to research across the field, carrying forward methods and approaches they developed under Hersh's mentorship. The program's longevity reflects both the quality of the application and the stability of Hersh's leadership.
This infrastructure of training graduate programs, short courses, textbooks, and research grants constitutes the connective tissue of a field. Individual papers come and go; training programs build the people who write the next decade's papers.
Distance Learning and the Globalization of Informatics Education
Hersh spearheaded OHSU's early efforts in distance learning for biomedical informatics. This was not merely a logistical adaptation; it was a deliberate strategy to make specialized education available beyond the Pacific Northwest. He has taught in person and online widely around the world, bringing courses to settings where local infrastructure could not support a full graduate program.
The DS-I Africa consortium meeting in Mauritius in November 2024 illustrates this global dimension. Hersh participated as a speaker in the Computational Omics and Biomedical Informatics program track, addressing an audience drawn from across the continent. His biography for that event, published on the DS-I Africa site, describes him as serving as Director of OHSU's Biomedical Informatics Graduate Program, which it characterizes as one of the oldest and largest in the world.
This international reach is not incidental to Hersh's mission it is integral to it. Biomedical informatics is a global endeavor, and the problems it addresses disease surveillance, clinical data sharing, health equity in information access do not respect institutional boundaries.
Recognition and Fellowships
Hersh's career has been recognized by several professional bodies. He has been elected to the National Academy of Medicine, one of the highest honors in American medicine and public health. He is a Fellow of the American College of Physicians, the American College of Medical Informatics, the International Academy of Health Sciences Informatics, and the American Medical Informatics Association. He is also an elected member of the ACM Special Interest Group in Information Retrieval Academy.
These fellowships are not decorative. They represent peer recognition that Hersh's work has shaped the standards, methods, and intellectual community of multiple intersecting fields medicine, informatics, and information retrieval. The SIGIR Academy membership is particularly notable for a physician-educator, signaling that his contributions to search have been recognized by the core IR research community, not only by clinical audiences.
What This Means for WebSearches Readers
For readers researching the infrastructure of specialized search particularly those interested in how professional domains build their own information retrieval ecosystems Hersh's career offers a instructive case study. He did not work in consumer search or general web indexing. He worked at the intersection of medicine and search, building curricula, writing foundational texts, and designing evaluation frameworks for a domain where finding the right information has direct consequences for patient care.
The lessons transfer. Hersh's approach to biomedical IR treating it as a discipline with theory, methods, and practice more than as a technical afterthought has parallels in other professional domains that are now building their own search infrastructure. Legal search, scientific literature retrieval, and domain-specific answer engines all operate in the space between general IR principles and field-specific knowledge structures. Understanding how one field built that infrastructure over three decades provides a useful map for others.
The Continuing Work
Even after stepping down as department chair in 2022 and as program director in 2023, Hersh remains an active professor in the Division of Informatics, Clinical Epidemiology, and Translational Data Science at OHSU. His current research continues to focus on applying search techniques to clinical data, and he maintains an active publishing and teaching schedule.
The work of building information infrastructure for medicine is not finished. EHR systems have become more complex, clinical data has grown exponentially, and the demand for real-time, clinically relevant retrieval has intensified. Hersh's career provides one model for how to approach that problem systematically through education, research, and the patient accumulation of institutional resources over decades.
Why This Matters
Medical search is not a solved problem. It is a domain where the stakes are high, the data is messy, and the users physicians, researchers, patients need reliable answers under time pressure. Hersh's contribution is not a single algorithm or system, though he has built many. His contribution is the longer work: building the human infrastructure the trained people, the textbooks, the curricula, the research programs that makes good medical search possible.
For those working in search, discovery, and answer engines, there is something instructive in a career that has stayed consistently focused on one domain's information needs for three decades. The specific techniques change; the domain knowledge compounds. Hersh has chosen compounding.
Where to Read Further
Readers interested in exploring Hersh's work directly can start with his official biography on billhersh.info, which provides an overview of his teaching, research, and publications. The OHSU faculty profile offers institutional context for his role at the university. His Springer textbook, Information Retrieval: A Biomedical and Health Perspective, is available through Springer Nature's Research Communities platform, which also features a "Behind the Paper" post on his recent work comparing generative AI systems in introductory informatics courses. The DS-I Africa consortium biography documents his international teaching and global reach.