Abstract:For a semibounded sesquilinear form ${\mathfrak t}$ in a Hilbert space ${\mathfrak H}$ there exists a representing map $Q$ from ${\mathfrak H}$ to another Hilbert space ${\mathfrak K}$, such that ${\mathfrak t}[\varphi, \psi]-c(\varphi, \psi)=(Q\varphi,Q\psi)$, $\varphi,\psi \in {\rm dom\,}{\mathfrak t}$, with $c \in {\mathbb R}$ a lower bound of ${\mathfrak t}$. Representing maps offer a simplifying tool to study general semibounded forms. By means of representing maps closedness, closability, and singularity of ${\mathfrak t}$ are immediately translated into the corresponding properties of the operator $Q$, and vice versa. Also properties of sum decompositions ${\mathfrak t}={\mathfrak t}_1+{\mathfrak t}_2$ of a nonnegative form ${\mathfrak t}$ with two other nonnegative forms ${\mathfrak t}_1$ and ${\mathfrak t}_2$ in ${\mathfrak H}$ can be analyzed by means of associated nonnegative contractions $K\in {\mathbf B}({\mathfrak K})$. This helps, for instance, to establish an explicit operator theoretic characterization for the summands ${\mathfrak t}_1$ and ${\mathfrak t}_2$ to be, or not to be, mutually singular. Such sum decompositions are used to study characteristic properties of the so-called Lebesgue type decompositions of semibounded forms ${\mathfrak t}$, where ${\mathfrak t}_1$ is closable and ${\mathfrak t}_2$ singular; in particular, this includes the Lebesgue decomposition of a semibounded form due to B. Simon. Furthermore, for a semibounded form ${\mathfrak t}$ with its representing map $Q$ it will be shown that the corresponding semibounded selfadjoint relation $Q^*Q^{**} +c$ is uniquely determined by a limit version of the classical representation theorem for the form ${\mathfrak t}$, being studied by W. Arendt and T. ter Elst in a sectorial context. Via representing maps a full treatment is given of the convergence of monotone sequences of semibounded forms.
What problem does this paper attempt to address?
The main problems that this paper attempts to solve are about the representation of semibounded forms and their Lebesgue - type decompositions. Specifically:
1. **Representation problem**: For a semibounded bilinear form \(t\) in a Hilbert space \(H\), there exists a representation mapping \(Q\) from \(H\) to another Hilbert space \(K\) such that:
\[
t[\phi, \psi]-c(\phi, \psi)=(Q\phi, Q\psi), \quad \phi, \psi\in \text{dom}\,t,
\]
where \(c\) is the lower bound of \(t\). This representation mapping \(Q\) provides a simplified tool to study general semibounded forms.
2. **Closedness and closability problems**: Through the representation mapping \(Q\), the closedness and closability of the semibounded form \(t\) can be immediately transformed into the corresponding properties of the operator \(Q\), and vice versa. For example, the semibounded form \(t\) is closable if and only if the representation mapping \(Q\) is closable.
3. **Singularity and mutual singularity problems**: Through the representation mapping \(Q\), the properties of the sum \(t = t_{1}+t_{2}\) of two non - negative forms \(t_{1}\) and \(t_{2}\) can be analyzed, especially whether they are mutually singular. This helps to establish an explicit operator - theoretic characteristic to determine whether \(t_{1}\) and \(t_{2}\) are mutually singular.
4. **Lebesgue - type decomposition problem**: The paper studies the Lebesgue - type decomposition of the semibounded form \(t\), that is, decomposing \(t\) into a closable semibounded form \(t_{\text{reg}}\) and a singular non - negative form \(t_{\text{sing}}\):
\[
t = t_{\text{reg}}+t_{\text{sing}}, \quad \text{dom}\,t=\text{dom}\,t_{\text{reg}}=\text{dom}\,t_{\text{sing}}.
\]
This decomposition is unique, and \(t_{\text{reg}}\) is the largest closable part of \(t\), while \(t_{\text{sing}}=t - t_{\text{reg}}\).
5. **Convergence problem of monotone sequences**: Through the representation mapping \(Q\), the paper also studies the convergence of semibounded forms of monotone sequences.
In conclusion, the paper simplifies and uniformly deals with various properties of semibounded forms, especially closability, singularity and Lebesgue - type decomposition, by introducing the representation mapping \(Q\), thus providing a new theoretical framework to understand and analyze these forms.