Left to right, computer science and engineering (CSE) professor Shashi Shekhar, CSE doctoral student Mete Celik, Jane Goodall, and CSE professor Jaideep Srivastava
May 28, 2008 In the lush tropical forest of Tanzania's Gombe National Park, famed chimpanzee expert Jane Goodall and her research team spent decades documenting chimpanzee behavior and habitat. Researchers there are still following chimpanzees daily, recording their travel, food choice, interactions with other chimpanzees, and geographical data.
Far from the forest, the products of this work--46 years worth of paper-based maps, hand written checksheets, notes, video, and satellite images--have found a home at the Jane Goodall Institute's Center for Primate Studies (JGI-CPS), on the University of Minnesota's Twin Cities campus in St. Paul.
While Goodall's research in Gombe and her outreach efforts have given the world a better understanding and appreciation for chimpanzees, University ecologists and computer scientists have teamed up to find new ways to use the data. They are analyzing the data for patterns in everything from female grouping habits to male aggression and mating habits relating to the Simian Immune Deficiency Virus (SIV). They are also constantly seeking new research techniques and areas of study.
Department of Computer Science and Engineering (CSE) professors Shashi Shekhar and Jaideep Srivastava have worked with their students for nearly five years on two of the University center's projects. CSE professor John Carlis and his students also helped set up a database of the behavioral data for the center in the late 1990s. Anne Pusey, director of the University primate center, said interdisciplinary collaborations with CSE have been very helpful in understanding the chimpanzee data.
For one of the current projects, CSE graduate student Mete Celik created a searchable database prototype that would organize more than 600 hours of chimpanzee video footage from the Jane Goodall Institute's videographer, Bill Wallauer. The video database is housed in the U's Digital Technology Center (DTC).
Celik explained how the technology works in a presentation to Goodall when she visited the University center in March. "It's kind of a Google-like search engine," he said. A chimpanzee's name and behavior can be entered to retrieve the corresponding video clips. Eventually the researchers would like to make the database into a searchable library that allows users to add their own perceptions of the research material.
"Years and years and years of data is being computerized, so that questions that I used to ask, which entailed going back through file after file after file by hand, can now be found very quickly by pressing buttons. It's quite extraordinary," Goodall said in a campus press conference. "It makes me very jealous, because I could have done so much. We didn't have some of these technologies back then."
Sample map collected during chimpanzee research.
The other CSE-related project focuses on analyzing the data. Behavioral ecology doctoral student Carson Murray used Shekhar and Srivastava's expertise in temporal, spatial, and spatiotemporal data mining to study patterns and commonalities in female chimpanzee relationships and location behavior. CSE graduate student Sandeep Mane also worked on this project.
Srivastava said they discovered patterns in female chimpanzee association and location behavior that revealed the importance of dominance. In her presentation to Goodall, Murray said she found that a high dominance rank equals an increased loyalty to core areas in the wild. Core areas are specific territories occupied by a chimpanzee.
"[CSE researchers] helped me to come up with a way to look at these point patterns," she said, adding that she's now looking at male core areas. "They are very much driven by food. It looks like they inherit their mother's core area," Murray said.
Shekhar and Srivastava said they have been amazed at the information gleaned from the data so far. "To me personally, it's fascinating," Srivastava said. "I learned how similar chimpanzee behavior is to human behavior." Pusey said this is a common realization. "Because chimps are our closest relatives, we're always thinking about their similarities and differences," she said.
While work is still ongoing for these two projects, new areas of study are being discussed. "We're interested in the different types of social bonds you see in the [chimpanzee] community," Pusey said.
JGI funds the long-term field work at Gombe. The University center work is funded by NSF, the National Institutes of Health (NIH), and a private grant from Milton Harris. Pusey said she is continuously working on fund raising and hopes that JGI will partner with the center on future fund-raising initiatives.
On her visit to campus, Goodall touted the value in the University center's work and expressed hope for future study of the data. "It means that years and years and years of blood and toil, crawling through the forest, being scratched by thorns, having your hair caught, is put down on these bits of paper and is now amazingly is being amalgamated and will be useful to students all over the world," she said.
For more information, see:
UMN News article republished from Soundbyte, the newsletter of the Department of Computer Science and Engineering. Photo by Robyn White. Map from "A Digital Library…" slides referenced above.