PhraseMap: Attention-based Keyphrases Recommendation for Information Seeking

Yamei Tu, Rui Qiu, Yu-Shuen Wang, Po-Yin Yen, Han-Wei Shen

Room: 106

2023-10-24T23:57:00ZGMT-0600Change your timezone on the schedule page
Exemplar figure, described by caption below
In this study, we introduce a visual analytics approach for efficient information retrieval from extensive document collections. We transform large corpora into an attention-based phrase graph, known as PhraseMap, where nodes represent keyphrases and edges illustrate their relationships. Our method employs the PhraseMap as a backend knowledge base and offers multiple user interactions. To enhance the visualization of the PhraseMap, we employ a resource-controlled self-organizing map (RC-SOM) to evenly display phrases on a grid while ensuring semantically similar phrases are located close together (A). We also devise a navigation algorithm to support information retrieval, consisting of (1) a question-answering (QA) model that identifies phrases relevant to user queries (A') and (2) the option to accept/decline top-recommended phrases (B), followed by further exploration through documents (C).
Fast forward
Full Video

Machine learning;natural language processing;textual data;user-in-the-loop;visual analytics


Many Information Retrieval (IR) approaches have been proposed to extract relevant information from a large corpus. Among these methods, phrase-based retrieval methods have been proven to capture more concrete and concise information than word-based and paragraph-based methods. However, due to the complex relationship among phrases and a lack of proper visual guidance, achieving user-driven interactive information-seeking and retrieval remains challenging. In this study, we present a visual analytic approach for users to seek information from an extensive collection of documents efficiently. The main component of our approach is a PhraseMap, where nodes and edges represent the extracted keyphrases and their relationships, respectively, from a large corpus. To build the PhraseMap, we extract keyphrases from each document and link the phrases according to word attention determined using modern language models, i.e., BERT. As can be imagined, the graph is complex due to the extensive volume of information and the massive amount of relationships. Therefore, we develop a navigation algorithm to facilitate information seeking. It includes (1) a question-answering (QA) model to identify phrases related to users’ queries and (2) updating relevant phrases based on users’ feedback. To better present the PhraseMap, we introduce a resource-controlled self-organizing map (RC-SOM) to evenly and regularly display phrases on grid cells while expecting phrases with similar semantics to stay close in the visualization. To evaluate our approach, we conducted case studies with three domain experts in diverse literature. The results and feedback demonstrate its effectiveness, usability, and intelligence.