Interpreting the Fuzzy Semantics of Natural-Language Spatial Relation Terms with the Fuzzy Random Forest Algorithm

Abstract

Naïve Geography, intelligent geographical information systems (GIS), and spatial data mining especially from social media all rely on natural-language spatial relations (NLSR) terms to incorporate commonsense spatial knowledge into conventional GIS and to enhance the semantic interoperability of spatial information in social media data. Yet, the inherent fuzziness of NLSR terms makes them challenging to interpret. This study proposes to interpret the fuzzy semantics of NLSR terms using the fuzzy random forest (FRF) algorithm. Based on a large number of fuzzy samples acquired by transforming a set of crisp samples with the random forest algorithm, two FRF models with different membership assembling strategies are trained to obtain the fuzzy interpretation of three line-region geometric representations using 69 NLSR terms. Experimental results demonstrate that the two FRF models achieve good accuracy in interpreting line-region geometric representations using fuzzy NLSR terms. In addition, fuzzy classification of FRF can interpret the fuzzy semantics of NLSR terms more fully than their crisp counterparts.

A fuzzy approach for determining the cognitive spatial location of an object in geographical information system

“…In [22], a model is presented that provides a complete description in natural language of the inner and outer parts of clear or fuzzy lines and corresponds to the cognitive habits of a person. In [23], an approach to the interpretation of fuzzy semantics of terms of spatial relations in a natural language using a fuzzy random forest algorithm is proposed. These models can handle fuzzy spatial queries.…”

Modeling Words for Qualitative Distance Based on Interval Type-2 Fuzzy Sets

“…In addition, another key issue is to determine the semantics of each qualitative distance variable, which is important for natural language understanding, scene reconstruction and many GIS-based DSSs [2,3]. Although natural language is gradually becoming an important data source for GISs [9] and some studies interpreted fuzzy semantics of natural language spatial relation (NLSR) terms using the fuzzy random forest (FRF) algorithm and fuzzy sets [5,10], most GISs currently cannot handle natural language effectively, and most linguistic descriptions about spatial relations cannot be used in current GISs. This issue has not yet been solved; therefore, how to integrate the spatial relationships described in natural language into a GIS is a challenge in the field of GIS [11].…”

Qualitative topological relations between interval type‐2 fuzzy spatial objects

“…Over the past 20 years, many studies in the field of geographic information systems (GIS) and spatial database domains have focused on models of fuzzy geographical objects and their relationships, because the shapes of many spatial objects are inherently vague (e.g. residential centers, polluted rivers, forests) (Bloch, ; Dilo, ; Du et al, ; Guo, & Cui, ; Liu & Shi, ; Schneider, ; Schneider, ; Wang, Du, Feng, Zhang, & Zhang, ; Zhan, ). Molenaar () introduced three definition levels of fuzzy objects, and extensional uncertainty plays a dominant role in the spatial representation of fuzzy objects.…”