Could spatial features help the matching of textual data?

Abstract

Textual data is available to an increasing extent through different media (social networks, companies data, data catalogues, etc.). New information extraction methods are needed since these new resources are highly heterogeneous. In this article, we propose a text matching process based on spatial features and assessed through heterogeneous textual data. Besides being compatible with heterogeneous data, it comprises two contributions: first, spatial information is extracted for comparison purposes and subsequently stored in a dedicated spatial textual representation (STR); and then two transformations are applied on STR to improve the spatial similarity estimation. This article outlines the proposed approach with new contributions: (i) a new geocoding methods using general co-occurrences between entities, and (ii) a thorough evaluation followed by (iii) an in-depth discussion. The results obtained on two corpora demonstrate that good spatial matches (≈ 80% precision on major criteria) can be obtained between the most similar STRs with further enhancement achieved via STR transformation.

Integrating Textual Data into Heterogeneous Data Ingestion Processing

“…For instance, many heterogeneous satellite images are currently available that require analysis [12]. Image-text matching based on spatial information [6] improves information retrieval and image annotation techniques [13]. Thus, a more global data context is provided that may be useful for experts involved in land-use planning [10].…”

GeoNLPlify: A spatial data augmentation enhancing text classification for crisis monitoring

“…Its attention to spatial entities decreases in favor of the other words in the document. At the same time, our empirical intuition is that the spatial level has an influence on the classification (by spatial level, we mean city, state or country level); people provide a better report of the local situation during a crisis by being at the right spatial level [35,36]. For example, an outbreak declaration will focus on the city or state level where the case occurs, although the consequences will be reported at a country level, as illustrated by Fig.…”

“…Furthermore, GeoNLPlify could also consider the co-occurrence of spatial entities and their relationships. Presented in [36], spatial textual representations (STRs) are spatial representation graphs of spatial entity co-occurrences found in texts. This formalization in the form of STR enables the comparison of different documents based on their respective list of spatial entities.…”

Context-Aware Text Matching Algorithm for Korean Peninsula Language Knowledge Base Based on Density Clustering

“…With the rapid development of the digital society, people's needs in the fields of artificial intelligence such as information retrieval, automatic question answering, and dialogue systems have begun to appear, and intelligent matching algorithms are needed to meet the high needs of the users [1]. In order to meet these requirements, natural language processing technology emerged, which can provide users with efficient information retrieval services [2]. Text matching algorithm is the core research area in natural language processing technology.…”