A quasi-current representation for information needs inspired by Two-State Vector Formalism

Wang, Panpan; Hou, Yuexian; Li, Jingfei; Zhang, Yazhou; Song, Dawei and Li, Wenjie (2017). A quasi-current representation for information needs inspired by Two-State Vector Formalism. Physica A: Statistical Mechanics and its Applications, 482 pp. 627–637.

DOI: https://doi.org/10.1016/j.physa.2017.04.145

Abstract

Recently, a number of quantum theory (QT)-based information retrieval (IR) models have been proposed for modeling session search task that users issue queries continuously in order to describe their evolving information needs (IN). However, the standard formalism of QT cannot provide a complete description for users’ current IN in a sense that it does not take the ‘future’ information into consideration. Therefore, to seek a more proper and complete representation for users’ IN, we construct a representation of quasi-current IN inspired by an emerging Two-State Vector Formalism (TSVF). With the enlightenment of the completeness of TSVF, a “two-state vector” derived from the ‘future’ (the current query) and the ‘history’ (the previous query) is employed to describe users’ quasi-current IN in a more complete way. Extensive experiments are conducted on the session tracks of TREC 2013 & 2014, and show that our model outperforms a series of compared IR models.

Viewing alternatives

Look up in Google Scholar

Download history

Metrics

Public Attention

Altmetrics from Altmetric

Number of Citations

Citations from Dimensions

Item Actions

Export

About

CORE (COnnecting REpositories)

Recommendations