Literature Review and Topic Modeling

Libraries of journal articles on perennial grain sorghum and perennial pigeon pea were compiled using advanced search techniques in 4 scientific literature databases: Agricola, Web of Science, ScienceDirect, and Scopus. Search terms were created by pairing a descriptive inclusion term with a crop inclusion term through the use of a proximity operator. Search results were then collected in the reference management software Zotero. There, duplicates were identified and merged, and each article retrieved was evaluated for relevance. Then, the library was exported from Zotero so that its metadata could be evaluated in the statistical program R and the technique of topic modeling could be applied.

Topic modeling is a statistical approach to analyzing texts that searches for frequently co-occurring terms in a group of documents. These groups of terms identify topics and individual papers within the group of texts can be assigned to topics given the probability of their association with that group of terms. With sufficient text, topic modeling is an effective and objective way of evaluating a large body of literature to understand predominant themes.

Bibliographies by Perennial Crop

Pigeon pea growing in Malawi.

Pigeon Pea Bibliography

perennial rice regrowth

Rice Bibliography

Ratooned sorghum resprouting in northern Ghana.

Sorghum Bibliography

Root growth of intermediate wheat grass compared to annual wheat.

Wheat Bibliography

Full Bibliography

Sorry, but the selected Zotpress account can't be found.